The author thanks the Chinese Academy of Agricultural Sciences and Guangxi University for supporting this work. The work was supported by the National Key R&#38;D Program of China (2022YFC2601500 &#38; 2022YFC2601504), the National Natural Science Foundation of China (32272633), Shenzhen Science and Technology Program (KCXFZ20230731093259009)

1. Preparation of datasets
<ol>
	<li>Selecting the research object
	<ol>
		<li>Select test samples based on the focus of the experimental study, considering options like Mikania micrantha or other invasive plants.</li>
	</ol>
	</li>
	<li>Collecting UAV images
	<ol>
		<li>Prepare square plastic frames of size 0.5 m*0.5 m and quantity 25-50, depending on the size of the area studied.</li>
		<li>Employ a random sampling approach to determine soil sampling locations in the study area using a sufficient number of biomass samples. Position the sample frame horizontally over the vegetation, fully encompassing the plants with a minimum separation distance of 2 m between each plant.</li>
		<li>Use a drone and a camera to form a UAV remote-sensing filming system, as shown in Figure 1.</li>
		<li>Use the UAV to plot the route within the specified study area. The route planning setup is shown in Figure 2.
		<ol>
			<li>Establish a heading and side overlap rate of 70%, capture photos at uniform time intervals of 2 s, maintain the camera angle perpendicular to the ground at 90&#176;, and position the camera altitude at 30 m. This resulted in continuous visible image data of the study area with a single image resolution measuring 8256 x 5504 pixels, as depicted in Figure 3.</li>
		</ol>
		</li>
		<li>Store aerial imagery for subsequent processing with Python software for biomass estimation.</li>
	</ol>
	</li>
	<li>Collecting the aboveground biomass
	<ol>
		<li>Collect the aboveground biomass of Mikania micrantha manually within each sample plot after completing the drone data collection. Bag them and label each bag accordingly.
		<ol>
			<li>When collecting Mikania micrantha, prevent sample plots from moving. First, cut the Mikania micrantha along the inside edge of the sample plot.</li>
			<li>Then, cut the rhizome of the Mikania micrantha from the bottom. Remove any dirt, rocks, or other plants that are mixed in. Finally, bag and label the samples.</li>
		</ol>
		</li>
		<li>Bring the collected samples of invasive plants from step 1.3.1 to the laboratory. Air-dry all the collected samples to evaporate most of the moisture.</li>
		<li>To further remove moisture from air-dried samples, use an oven. Set the temperature to 55 &#176;C. Dry the samples for 72 h, then weigh each sample on an electronic balance and record the biomass data in grams (g).
		<ol>
			<li>Place the electronic balance in an undisturbed environment, weigh, calibrate, and continue weighing. Place bags of Mikania micrantha on the electronic balance, wait for the readings to stabilize, and record the readings.</li>
			<li>Weigh the Mikania micrantha every hour until the mass no longer changes and record the reading minus the weight of the bag as the measured mass of that sample. Calculate the aboveground biomass of the invasive plant using the formula below: 
			<img alt="Equation 1" src="/files/ftp_upload/66067/66067eq01.jpg" style="margin: auto;" /> 
			where B represents the biomass of Mikania micrantha in grams per square meter (g/m2), M is the weight of the measured Mikania micrantha, measured in grams (g), S corresponds to the area of the sample plot in square meters (m2).</li>
		</ol>
		</li>
	</ol>
	</li>
	<li>Creating a dataset
	<ol>
		<li>Extract the RGB image corresponding to the sample image from the original UAV image. Divide it into a grid of 280 &#215; 280 pixels using Python programming (Supplementray Figure 1).</li>
		<li>Segment the raw image data into smaller images of the same size as the sample images using Python programming. Use the sliding window method for segmentation, setting the horizontal and vertical steps to 280 pixels.</li>
		<li>From the small images segmented in step 1.4.2, randomly select 880 invasive plant images and 1500 background images to create a dataset. Then, split this dataset into training, validation, and test sets in a 6:2:2 ratio (Supplementray Figure 2).</li>
	</ol>
	</li>
</ol>
2. Identification of Mikania micrantha
<ol>
	<li>Preparing the software
	<ol>
		<li>Go to Anaconda&#39;s official website (https://www.anaconda.com/) and download and install Anaconda. Then, head to PyCharm&#39;s website (https://www.jetbrains.com/pycharm/) and download the PyCharm IDE.</li>
	</ol>
	</li>
	<li>Creating a Conda environment.
	<ol>
		<li>Open the Anaconda Prompt command line after installing Anaconda, then type conda create -n pytorch python==3.8 to create a new Conda environment. After the environment is created, enter conda info --envs to confirm that the pytorch environment exists.</li>
		<li>Open the Anaconda Prompt and activate the pytorch environment by entering conda activate pytorch. Check the current (Compute Unified Device Architecture)CUDA version by typing nvidia-smi. Then, install PyTorch version 1.8.1 by running the command conda install pytorch==1.8.1 torchvision==0.9.1 torchaudio==0.8.1 cudatoolkit=11.0 -c pytorch.</li>
	</ol>
	</li>
	<li>Runs for model recognition 
	NOTE: Use PyTorch to build the Mikania micrantha recognition model used in this paper. The network model employed is ResNet10121, which remains consistent with the original paper in its architecture. Modifications are made to the network&#39;s output section to meet the requirements for chamomile recognition.
	<ol>
		<li>Preprocess the images to prepare them for model input. Resize the images from 280 x 280 pixels to 224 x 224 pixels and normalize them to ensure they meet the model&#39;s size requirements using the following code: 
		transform = transforms.Compose([ 
		transforms.Resize((224, 224)), 
		transforms.ToTensor(), 
		&#8203;transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])])</li>
		<li>Perform image feature extraction and reduce dimensionality using a convolutional neural network.
		<ol>
			<li>First, initialize the convolutional layer for the initial feature extraction through self.conv1. With this convolutional layer, the original image is convolved into a feature map with self.in_channel channels for extracting initial features (Supplementary Figure 3A). 
			NOTE: Advanced features are extracted in a convolution operation on the residual passes. These layers are produced by invoking the _make_layer function, which comprises a sequence of residual blocks. Each residual block consists of convolution, batch normalization, and activation functions to gradually extract sophisticated features (Supplementary Figure 3B).</li>
			<li>Use the layer&#39;s function to modify the channel number for dimensionality reduction via 1x1 convolution. This operation decreases the computational load while preserving significant features (Supplementary Figure 3C). 
			NOTE: Overall, ResNet101 performs feature extraction by using various convolutional layers, and dimensionality reduction is achieved through 1x1 convolutional layers within the residual block. This approach allows the network to learn features more deeply and avoid the issue of gradient vanishing, thus enabling more efficient learning and representation of image features for complex tasks.</li>
			<li>After convolutions and pooling operations, input the high-quality features into a fully connected layer. 
			NOTE: In the ResNet architecture, feature extraction takes place in the convolutional layer. These features are subsequently sent to the fully-connected (FC) layer for classification (Supplementary Figure 4). The operation self.avgpool(x) performs adaptive average pooling to reshape the tensor to a fixed size. The operation torch.flatten(x, 1) spreads the tensor into a one-dimensional vector, and self.fc(x) applies the fully connected layer to the flattened vector, ultimately serving as the final step for classification. This process effectively passes the extracted features through the convolutional layer, transforming them into a format suitable for classification via the fully connected layer.</li>
			<li>Use the Softmax function to obtain the final output based on the three classification requirements.</li>
		</ol>
		</li>
		<li>Train a multi-class recognition model with the dataset from step 1.4. Set the number of iterations to 200 and an initial learning rate of 0.0001. Reduce the learning rate by a third every 10 iterations with a batch size of 64. Save the optimal model parameters automatically after each iteration (Supplementary Figure 5).</li>
		<li>Employ a meticulously trained recognition model and systematically traverse the original image from step 1.2.2 for identification purposes.
		<ol>
			<li>Configure horizontal and vertical steps precisely at 280 pixels, resulting in the generation of a comprehensive distribution map highlighting the presence of invasive flora within the boundaries of the study area. Present the selected results visually as shown in Figure 4. 
			NOTE: The initial image is preprocessed by segmenting it into smaller chunks, classifying each chunk using a trained deep learning model, and combining the results into an output image. If a chunk is classified as an invasive plant, the corresponding location on the output image is set to 255. The resulting output image is saved as a grayscale image file. The specific implementation code is shown in Supplementary Figure 6.</li>
		</ol>
		</li>
	</ol>
	</li>
</ol>
3. Estimation of invasive plant biomass
<ol>
	<li>Perform simple data augmentation with the RandomResizedCrop and RandomHorizontalFlip functions (Supplementary Figure 7) to extend the image set created in step 1.2 and extract the six vegetation indices commonly used for estimating biomass, which are RBRI, GBRI, GRRI, RGRI, NGBDI, and NGRDI. Refer to Table 1 for the calculation formulas for these indices.</li>
	<li>Create a K-nearest neighbor regression (KNNR)22 model using the output of the model to ensure precise estimation of the biomass of invasive plants. Use the extracted vegetation indices as inputs for the estimation model.</li>
	<li>Use the coefficient of determination R2 and Root Mean Square Error (RMSE)23 to assess the model&#39;s accuracy, which is calculated as follows: 
	<img alt="Equation 2" src="/files/ftp_upload/66067/66067eq02.jpg" style="margin: auto;" /> 
	<img alt="Equation 3" src="/files/ftp_upload/66067/66067eq03.jpg" style="margin: auto;" /> 
	NOTE: The K-Nearest Neighbor Regression (KNNR) algorithm is a nonparametric machine learning technique used to solve regression problems. Its fundamental concept is to predict outcomes by determining the closest K neighbors in feature space based on input sample distances. Key benefits of using KNNR include its simplicity and ease of understanding, and it requires no training phase. Additionally, KNNR does not make excessive assumptions about the data&#39;s distribution. KNNR can be applied in regression issues to anticipate continuous objective variables and precisely evaluate the biomass of invasive plants.</li>
	<li>Employ the aboveground biomass estimation model chosen in step 3.2 and scan through the invasive plant distribution map from step 2.3.4 with horizontal and vertical strides of 280 pixels.</li>
</ol>

Invasive Plant Biomass Assessment by Unmanned Aerial Vehicle Remote Sensing Data

<ol>
	<li>Lian, J. Y., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Influence+of+obligate+parasite+Cuscuta+campestris+on+the+community+of+its+host+Mikania+micrantha.">Influence of obligate parasite Cuscuta campestris on the community of its host Mikania micrantha.</a> Weed Research. 46, 441-443 (2006).</li><li>Yu, X., Bao, Q. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Aboveground+biomass+estimation+of+potato+from+UAV+multispectral+imagery.">Aboveground biomass estimation of potato from UAV multispectral imagery.</a> Remote Sensing. 54 (4), 96-99 (2023).</li><li>Guo, T. C., Wang, Y. H., Feng, W. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Research+on+wheat+yield+estimation+based+on+multimodal+data+fusion+from+unmanned+aircraft+platforms.">Research on wheat yield estimation based on multimodal data fusion from unmanned aircraft platforms.</a> Acta Agronomica Sinica. 48 (7), 15 (2022).</li><li>Shao, G. M., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Estimation+of+transpiration+coefficient+and+aboveground+biomass+in+maize+using+time-series+UAV+multispectral+imagery.">Estimation of transpiration coefficient and aboveground biomass in maize using time-series UAV multispectral imagery.</a> The Crop Journal. 10 (5), 1376-1385 (2022).</li><li>Jiang, Q., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=UAV-based+biomass+estimation+for+rice-combining+spectral,+TIN-based+structural+and+meteorological+features.">UAV-based biomass estimation for rice-combining spectral, TIN-based structural and meteorological features.</a> Remote Sensing. 11 (7), 890 (2019).</li><li>Fei, S. P., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=UAV-based+multi-sensor+data+fusion+and+machine+learning+algorithm+for+yield+prediction+in+wheat.">UAV-based multi-sensor data fusion and machine learning algorithm for yield prediction in wheat.</a> Precision Agriculture. 24, 187-212 (2023).</li><li>Shu, S., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Aboveground+biomass+estimation+of+rice+based+on+unmanned+aerial+vehicle+imagery.">Aboveground biomass estimation of rice based on unmanned aerial vehicle imagery.</a> Fujian Journal of Agricultural Sciences. 37 (7), 9 (2022).</li><li>Wu, X., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=UAV+LiDAR-based+biomass+estimation+of+individual+trees.">UAV LiDAR-based biomass estimation of individual trees.</a> Fujian Journal of Agricultural Sciences. 22 (34), 15028-15035 (2022).</li><li>Yang, X., Zan, M., Munire, M. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Estimation+of+above+ground+biomass+of+Populus+euphratica+forest+using+UAV+and+satellite+remote+sensing.">Estimation of above ground biomass of Populus euphratica forest using UAV and satellite remote sensing.</a> Transactions of the Chinese Society of Agricultural Engineering. 37 (1), 7 (2021).</li><li>Li, B., Liu, K. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Forest+biomass+estimation+based+on+UAV+optical+remote+sensing.">Forest biomass estimation based on UAV optical remote sensing.</a> Forest Engineering. 5, 38 (2022).</li><li>Li, Z., Zan, Q., Yang, Q., Zhu, D., Chen, Y., Yu, S. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Remote+estimation+of+mangrove+aboveground+carbon+stock+at+the+species+level+using+a+low-cost+unmanned+aerial+vehicle+system.">Remote estimation of mangrove aboveground carbon stock at the species level using a low-cost unmanned aerial vehicle system.</a> Remote Sensing. 11 (9), 1018 (2019).</li><li>Luo, S., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Fusion+of+airborne+LiDAR+data+and+hyperspectral+imagery+for+aboveground+and+belowground+forest+biomass+estimation.">Fusion of airborne LiDAR data and hyperspectral imagery for aboveground and belowground forest biomass estimation.</a> Ecological Indicators. 73, 378-387 (2017).</li><li>Li, S., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Research+of+grassland+aboveground+biomass+inversion+based+on+UAV+and+satellite+remoting+sensing.">Research of grassland aboveground biomass inversion based on UAV and satellite remoting sensing.</a> Remote Sensing Technology and Application. 1, 037 (2022).</li><li>Wengert, M., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Multisite+and+multitemporal+grassland+yield+estimation+using+UAV-borne+hyperspectral+data.">Multisite and multitemporal grassland yield estimation using UAV-borne hyperspectral data.</a> Remote Sensing. 14 (9), 2068 (2022).</li><li>Li, Y., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=The+effect+of+season+on+Spartina+alterniflora+identification+and+monitoring.">The effect of season on Spartina alterniflora identification and monitoring.</a> Frontiers in Environmental Science. 10, 1044839 (2022).</li><li>Wang, F., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Estimation+of+above-ground+biomass+of+winter+wheat+based+on+consumer-grade+multi-spectral+UAV.">Estimation of above-ground biomass of winter wheat based on consumer-grade multi-spectral UAV.</a> Remote Sensing. 14 (5), 1251 (2022).</li><li>Lu, N., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Improved+estimation+of+aboveground+biomass+in+wheat+from+RGB+imagery+and+point+cloud+data+acquired+with+a+low-cost+unmanned+aerial+vehicle+system.">Improved estimation of aboveground biomass in wheat from RGB imagery and point cloud data acquired with a low-cost unmanned aerial vehicle system.</a> Plant Methods. 15 (1), 17 (2019).</li><li>Chen, H., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Mapping+forest+aboveground+biomass+with+MODIS+and+Fengyun-3C+VIRR+imageries+in+Yunnan+Province,+Southwest+China+using+linear+regression,+K-nearest+neighbor+and+random.">Mapping forest aboveground biomass with MODIS and Fengyun-3C VIRR imageries in Yunnan Province, Southwest China using linear regression, K-nearest neighbor and random.</a> Remote Sensing. 14 (21), 5456 (2022).</li><li>Yan, M., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Biomass+estimation+of+subtropical+arboreal+forest+at+single+tree+scale+based+on+feature+fusion+of+airborne+LiDAR+data+and+aerial+images.">Biomass estimation of subtropical arboreal forest at single tree scale based on feature fusion of airborne LiDAR data and aerial images.</a> Sustainability. 15 (2), 1676 (2023).</li><li>Tian, Y. C., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Aboveground+mangrove+biomass+estimation+in+Beibu+Gulf+using+machine+learning+and+UAV+remote+sensing.">Aboveground mangrove biomass estimation in Beibu Gulf using machine learning and UAV remote sensing.</a> Science of the Total Environment. 781, 146816 (2021).</li><li>Shrivastava, A., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Beyond+skip+connections:+Top-down+modulation+for+object+detection.">Beyond skip connections: Top-down modulation for object detection.</a> arXiv. , (2016).</li><li>Belkasim, S. O., Shridhar, M., Ahmadi, M. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Pattern+classification+using+an+efficient+KNNR.">Pattern classification using an efficient KNNR.</a> Pattern Recognition. 25 (10), 1269-1274 (1992).</li><li>Joel, S., Jose Luis, A., Shawn, C. K. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Farming+and+earth+observation:+Sentinel-2+data+to+estimate+within-field+wheat+grain+yield.">Farming and earth observation: Sentinel-2 data to estimate within-field wheat grain yield.</a> International Journal of Applied Earth Observation and Geoinformation. 107, 102697 (2022).</li><li>Tian, L., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Review+of+remote+sensing-based+methods+for+forest+aboveground+biomass+estimation:+Progress,+challenges,+and+prospects.">Review of remote sensing-based methods for forest aboveground biomass estimation: Progress, challenges, and prospects.</a> Forests. 14 (6), 1086 (2023).</li><li>Wei, X. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Biomass+estimation:+A+remote+sensing+approach.">Biomass estimation: A remote sensing approach.</a> Geography Compass. 4 (11), 1635-1647 (2010).</li><li>D&#233;bora, B., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=New+methodology+for+intertidal+seaweed+biomass+estimation+using+multispectral+data+obtained+with+unoccupied+aerial+vehicles.">New methodology for intertidal seaweed biomass estimation using multispectral data obtained with unoccupied aerial vehicles.</a> Remote Sensing. 15 (13), 3359 (2023).</li><li>Zhang, J. Y., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Unmanned+aerial+system-based+wheat+biomass+estimation+using+multispectral,+structural+and+meteorological+data.">Unmanned aerial system-based wheat biomass estimation using multispectral, structural and meteorological data.</a> Agriculture. 13 (8), 1621 (2023).</li><li>Shen, H., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Influence+of+the+obligate+parasite+Cuscuta+campestris+on+growth+and+biomass+allocation+of+its+host+Mikania+micrantha.">Influence of the obligate parasite Cuscuta campestris on growth and biomass allocation of its host Mikania micrantha.</a> Journal of Experimental Botany. 56 (415), 1277-1284 (2005).</li></ol>

The authors have nothing to disclose.

We present the detailed steps of an experiment on estimating the biomass of invasive plants using UAV remote sensing and computer vision. The main process and steps of this agreement are shown in Figure 7. Proper sample quality is one of the most crucial and challenging aspects of the program. This importance holds true for all invasive plants as well as&#160;any other plant biomass estimation experiments24.
To identify the distribution of ...

In this protocol, the proposed method of invasive biomass estimation based on UAV remote sensing and computer vision can reflect the distribution of invasive organisms and predict the degree of invasive biohazard. Estimates of the distribution and biomass of invasive organisms are critical to the prevention and control of these organisms. Once invasive plants invade, they can damage the ecosystem and cause huge economic losses. Quickly and accurately identifying invasive plants and estimating key invasive plant biomass are major challenges in invasive plant monitoring and control. In this protocol, we take Mikania micrantha as an example to explore an invasive plant biomass estimation method based on unmanned aerial remote sensing and computer vision, which provides a new approach and method for the ecological research of invasive plants and promotes the ecological research and management of invasive plants.
At present, the biomass measurement of Mikania micrantha is mainly done by manual sampling1. The traditional methods of biomass measurement need a lot of workforce and material resources, which are inefficient and limited by the terrain; it is difficult to meet the needs of regional biomass estimation of Mikania micrantha. The major advantage of using this protocol is that it provides a method for quantifying regional invasive plant biomass and spatial distribution of invasive plants in a way that does not take into account the sampling limitations of the area and eliminates the need for manual surveys.
UAV remote sensing technology has achieved certain results in plant biomass estimation and has been widely used in agriculture2,3,4,5,6,7, forestry8,9,10,11, and grassland12,13,14. UAV remote sensing technology has the advantages of low cost, high efficiency, high precision, and flexible operation15,16, which can efficiently obtain remote sensing image data in the study area; then, the texture feature and vegetation index of remote sensing image are extracted to provide data support for the estimation of plant biomass in large area. Current plant biomass estimation methods are mainly categorized into parametric and nonparametric models17. With the development of machine learning algorithms, nonparametric machine learning models with higher accuracy have been widely used in remote sensing estimation of plant biomass. Chen et al.18 utilized mixed logistic regression (MLR), KNNR, and random forest regression (RFR) to estimate the aboveground biomass of forests in Yunnan Province. They concluded that the machine learning models, specifically KNNR and RFR, resulted in superior outcomes compared to MLR. Yan et al.19 employed RFR and extreme gradient boosting (XGBR) regression models to assess the accuracy of estimating subtropical forest biomass using various sets of variables. Tian et al.20 utilized eleven machine-learning models to estimate the aboveground biomass of varying mangrove forest species in Beibuwan Bay. The researchers discovered that the XGBR method was more effective in determining the aboveground biomass of mangrove forests. Plant biomass estimation using man-machine remote sensing is a well-established practice, however, the use of UAV for biomass estimation of the invasive plant Mikania micrantha has yet to be reported both domestically and internationally. This approach is fundamentally different from all previous methods of biomass estimation for invasive plants, especially Mikania micrantha.
To sum up, UAV remote sensing has the advantages of high resolution, high efficiency, and low cost. In the feature variable extraction of remote sensing images, texture features combined with vegetation indexes can obtain better regression prediction performance. Nonparametric models can obtain more accurate regression models than parametric ones in plant biomass estimation. Therefore, to calculate the null distribution of invasive plants and their biomass precisely, we suggest the following outlined procedures for the invasive plant biomass experiment that relies on remote sensing using UAVs and computer vision.

<table border="1" fo:keep-together.within-page="1" fo:keep-with-next.within-page="always">
	<tbody>
		<tr>
			<td>DSLR camera</td>
			<td>Nikon</td>
			<td>D850</td>
			<td>Sensor type: CMOS; Maximum number of pixels: 46.89 million; Effective number of pixels: 45.75 million; Maximum resolution 8256 x 5504.</td>
		</tr>
		<tr>
			<td>GPU - Graphics Processing Unit</td>
			<td>NVIDIA&#160;</td>
			<td>RTX3090</td>
			<td></td>
		</tr>
		<tr>
			<td>Hexacopter</td>
			<td>DJI&#160;</td>
			<td>M600PRO</td>
			<td>Horizontal flight: 65 km/h (no wind environment); Maximum flight load: 6000 g</td>
		</tr>
		<tr>
			<td>PyCharm</td>
			<td>Python IDE</td>
			<td>2023.1</td>
			<td></td>
		</tr>
		<tr>
			<td>Python</td>
			<td>Python</td>
			<td>3.8.0</td>
			<td></td>
		</tr>
		<tr>
			<td>Pytorch</td>
			<td>Pytorch</td>
			<td>1.8.1</td>
			<td></td>
		</tr>
	</tbody>
</table>

computer vision-based biomass estimation for invasive plants

We report on the detailed steps of a method to estimate the biomass of invasive plants based on UAV remote sensing and computer vision. To collect samples from the study area, we prepared a sample square assembly to randomize the sampling points. An unmanned aerial camera system was constructed using a drone and camera to acquire continuous RGB images of the study area through automated navigation. After completing the shooting, the aboveground biomass in the sample frame was collected, and all correspondences were labeled and packaged. The sample data was processed, and the aerial images were segmented into small images of 280 x 280 pixels to create an image dataset. A deep convolutional neural network was used to map the distribution of Mikania micrantha in the study area, and its vegetation index was obtained. The organisms collected were dried, and the dry weight was recorded as the ground truth biomass. The invasive plant biomass regression model was constructed using the K-nearest neighbor regression (KNNR)&#160;by extracting the vegetation index from the sample images as an independent variable and integrating it with the ground truth biomass as a dependent variable. The results showed that it was possible to predict the biomass of invasive plants accurately. An accurate spatial distribution map of invasive plant biomass was generated by image traversal, allowing precise identification of high-risk areas affected by invasive plants. In summary, this study demonstrates the potential of combining unmanned aerial vehicle remote sensing with&#160;machine learning techniques to estimate invasive plant biomass. It contributes significantly to the research of new technologies and methods for real-time monitoring of invasive plants and provides technical support for intelligent monitoring and hazard assessment at the regional scale.

We show representative results of a computer vision-based method for the estimation of invasive plants, which is implemented in a programmatic way on a computer. In this experiment, we evaluated the spatial distribution and estimated the biomass of invasive plants in their natural habitats, using Mikania micrantha as a research subject. We utilized a drone camera system to acquire images of the research site, a portion of which is exhibited in Figure 3. We utilized the ResNet101 con...

Read this Scientific Article Protocol about Computer Vision-Based Biomass Estimation for Invasive Plants at JoVE.com

Author Spotlight: UAV Remote Sensing for Efficient Invasive Plant Biomass Estimation

We report detailed procedures for an invasive plant biomass estimation method that utilizes data obtained from unmanned aerial vehicle (UAV) remote sensing to assess biomass and capture the spatial distribution of invasive species. This approach proves highly beneficial for conducting hazard assessment and early warning of invasive plants.

Computer Vision-Based Biomass Estimation for Invasive Plants

We report on the detailed steps of a method to estimate the biomass of invasive plants based on UAV remote sensing and computer vision. To collect samples ...

computer-vision-based-biomass-estimation-for-invasive-plants

Research

JoVE Journal

Engineering

1.2K Views.  Guangxi University. Onze methode maakt gebruik van UAV-teledetectie en computervisie om invasieve plantenbiomassa efficiënt te schatten, waardoor de inefficiënties en subjectiviteit van traditionele handmatige onderzoeken en gegevensverzameling in complexe omgevingen worden overwonnen. Recente ontwikkelingen op het gebied van UAV-teledetectie voor het schatten van de biomassa van planten omvatten beeldvorming met hoge resolutie en multimodale gegevensintegratie. Deze ontwikkelingen verbeteren wetenschappelijk ...

Video: Auteur in de schijnwerpers: UAV-teledetectie voor efficiënte schatting van invasieve plantbiomassa

Implementation of a Reference Interferometer for Nanodetection

A thermally and mechanically stabilized fiber interferometer suited for examining ultra-high quality factor microcavities is fashioned. After assessing its free spectral range (FSR), the module is put in parallel with a fiber taper-microcavity system and then calibrated through isolating and eliminating random shifts in the laser frequency (i.e. laser jitter noise). To realize the taper-microcavity junction and to maximize the optical power that is transferred to the resonator, a single-mode optical fiber waveguide is pulled. Solutions containing polystyrene nanobeads are then prepared and flown to the microcavity in order to demonstrate the system&#8217;s ability to sense binding to the surface of the microcavity. Data is post-processed via adaptive curve fitting, which allows for high-resolution measurements of the quality factor as well as the plotting of time-dependent parameters, such as resonant wavelength and split frequency shifts. By carefully inspecting steps in the time-domain response and shifting in the frequency-domain response, this instrument can quantify discrete binding events.

A reference interferometer technique, which is designed to remove undesirable laser jitter noise for nanodetection, is utilized for probing an ultra-high quality factor microcavity. Instructions for assembly, setup, and data acquisition are provided, alongside the measurement process for specifying the cavity quality factor.

A thermally and mechanically stabilized fiber interferometer suited for examining ultra-high quality factor microcavities is fashioned. After assessing ...

Real-Time DC-dynamic Biasing Method for Switching Time Improvement in Severely Underdamped Fringing-field Electrostatic MEMS Actuators

Mechanically underdamped electrostatic fringing-field MEMS actuators are well known for their fast switching operation in response to a unit step input bias voltage. However, the tradeoff for the improved switching performance is a relatively long settling time to reach each gap height in response to various applied voltages. Transient applied bias waveforms are employed to facilitate reduced switching times for electrostatic fringing-field MEMS actuators with high mechanical quality factors. Removing the underlying substrate of the fringing-field actuator creates the low mechanical damping environment necessary to effectively test the concept. The removal of the underlying substrate also a has substantial improvement on the reliability performance of the device in regards to failure due to stiction. Although DC-dynamic biasing is useful in improving settling time, the required slew rates for typical MEMS devices may place aggressive requirements on the charge pumps for fully-integrated on-chip designs. Additionally, there may be challenges integrating the substrate removal step into the back-end-of-line commercial CMOS processing steps. Experimental validation of fabricated actuators demonstrates an improvement of 50x in switching time when compared to conventional step biasing results. Compared to theoretical calculations, the experimental results are in good agreement.

The robust device design of fringing-field electrostatic MEMS actuators results in inherently low squeeze-film damping conditions and long settling times when performing switching operations using conventional step biasing. Real-time switching time improvement with DC-dynamic waveforms reduces the settling time of fringing-field MEMS actuators when transitioning between up-to-down and down-to-up states.

Mechanically underdamped electrostatic fringing-field MEMS actuators are well known for their fast switching operation in response to a unit step input ...

Scalable Nanohelices for Predictive Studies and Enhanced 3D Visualization

Spring-like materials are ubiquitous in nature and of interest in nanotechnology for energy harvesting, hydrogen storage, and biological sensing applications.&#160; For predictive simulations, it has become increasingly important to be able to model the structure of nanohelices accurately.&#160; To study the effect of local structure on the properties of these complex geometries one must develop realistic models.&#160; To date, software packages are rather limited in creating atomistic helical models.&#160; This work focuses on producing atomistic models of silica glass (SiO2) nanoribbons and nanosprings for molecular dynamics (MD) simulations. Using an MD model of &#8220;bulk&#8221; silica glass, two computational procedures to precisely create the shape of nanoribbons and nanosprings are presented.&#160; The first method employs the AWK programming language and open-source software to effectively carve various shapes of silica nanoribbons from the initial bulk model, using desired dimensions and parametric equations to define a helix.&#160; With this method, accurate atomistic silica nanoribbons can be generated for a range of pitch values and dimensions.&#160; The second method involves a more robust code which allows flexibility in modeling nanohelical structures.&#160; This approach utilizes a C++ code particularly written to implement pre-screening methods as well as the mathematical equations for a helix, resulting in greater precision and efficiency when creating nanospring models.&#160; Using these codes, well-defined and scalable nanoribbons and nanosprings suited for atomistic simulations can be effectively created.&#160; An added value in both open-source codes is that they can be adapted to reproduce different helical structures, independent of material.&#160; In addition, a MATLAB graphical user interface (GUI) is used to enhance learning through visualization and interaction for a general user with the atomistic helical structures.&#160; One application of these methods is the recent study of nanohelices via MD simulations for mechanical energy harvesting purposes.

Accurate modeling of nanohelical structures is important for predictive simulation studies leading to novel nanotechnology applications.&#160; Currently, software packages and codes are limited in creating atomistic helical models.&#160; We present two procedures designed to create atomistic nanohelical models for simulations, and a graphical interface to enhance research through visualization.

Spring-like materials are ubiquitous in nature and of interest in nanotechnology for energy harvesting, hydrogen storage, and biological sensing ...

Surface Enhanced Raman Spectroscopy Detection of Biomolecules Using EBL Fabricated Nanostructured Substrates

Fabrication and characterization of conjugate nano-biological systems interfacing metallic nanostructures on solid supports with immobilized biomolecules is reported. The entire sequence of relevant experimental steps is described, involving the fabrication of nanostructured substrates using electron beam lithography, immobilization of biomolecules on the substrates, and their characterization utilizing surface-enhanced Raman spectroscopy (SERS). Three different designs of nano-biological systems are employed, including protein A, glucose binding protein, and a dopamine binding DNA aptamer. In the latter two cases, the binding of respective ligands, D-glucose and dopamine, is also included. The three kinds of biomolecules are immobilized on nanostructured substrates by different methods, and the results of SERS imaging are reported. The capabilities of SERS to detect vibrational modes from surface-immobilized proteins, as well as to capture the protein-ligand and aptamer-ligand binding are demonstrated. The results also illustrate the influence of the surface nanostructure geometry, biomolecules immobilization strategy, Raman activity of the molecules and presence or absence of the ligand binding on the SERS spectra acquired.

We describe the fabrication and characterization of nano-biological systems interfacing nanostructured substrates with immobilized proteins and aptamers. The relevant experimental steps involving lithographic fabrication of nanostructured substrates, bio-functionalization, and surface-enhanced Raman spectroscopy (SERS) characterization, are reported. SERS detection of surface-immobilized proteins, and probing of protein-ligand and aptamer-ligand binding is demonstrated.

Fabrication and characterization of conjugate nano-biological systems interfacing metallic nanostructures on solid supports with immobilized biomolecules ...

SSVEP-based Experimental Procedure for Brain-Robot Interaction with Humanoid Robots

Brain-Robot Interaction (BRI), which provides an innovative communication pathway between human and a robotic device via brain signals, is prospective in helping the disabled in their daily lives. The overall goal of our method is to establish an SSVEP-based experimental procedure by integrating multiple software programs, such as OpenViBE, Choregraph, and Central software as well as user developed programs written in C++ and MATLAB, to enable the study of brain-robot interaction with humanoid robots.
This is achieved by first placing EEG electrodes on a human subject to measure the brain responses through an EEG data acquisition system. A user interface is used to elicit SSVEP responses and to display video feedback in the closed-loop control experiments. The second step is to record the EEG signals of first-time subjects, to analyze their SSVEP features offline, and to train the classifier for each subject. Next, the Online Signal Processor and the Robot Controller are configured for the online control of a humanoid robot. As the final step, the subject completes three specific closed-loop control experiments within different environments to evaluate the brain-robot interaction performance.
The advantage of this approach is its reliability and flexibility because it is developed by integrating multiple software programs. The results show that using this approach, the subject is capable of interacting with the humanoid robot via brain signals. This allows the mind-controlled humanoid robot to perform typical tasks that are popular in robotic research and are helpful in assisting the disabled.

The overall goal of this method is to establish an SSVEP-based experimental procedure by integrating multiple software programs to enable the study of brain-robot interaction with humanoid robots, which is prospective in assisting the sick and elderly as well as performing unsanitary or dangerous jobs.

Brain-Robot Interaction (BRI), which provides an innovative communication pathway between human and a robotic device via brain signals, is prospective in ...

Colloidal Synthesis of Nanopatch Antennas for Applications in Plasmonics and Nanophotonics

We present a method for colloidal synthesis of silver nanocubes and the use of these in combination with a smooth gold film, to fabricate plasmonic nanoscale patch antennas. This includes a detailed procedure for the fabrication of thin films with a well-controlled thickness over macroscopic areas using layer-by-layer deposition of polyelectrolyte polymers, namely poly(allylamine) hydrochloride (PAH) and polystyrene sulfonate (PSS). These polyelectrolyte spacer layers serve as a dielectric gap in between silver nanocubes and a gold film. By controlling the size of the nanocubes or the gap thickness, the plasmon resonance can be tuned from about 500 nm to 700 nm. Next, we demonstrate how to incorporate organic sulfo-cyanine5 carboxylic acid (Cy5) dye molecules into the dielectric polymer gap region of the nanopatch antennas. Finally, we show greatly enhanced fluorescence of the Cy5 dyes by spectrally matching the plasmon resonance with the excitation energy and the Cy5 absorption peak. The method presented here enables the fabrication of plasmonic nanopatch antennas with well-controlled dimensions utilizing colloidal synthesis and a layer-by-layer dip-coating process with the potential for low cost and large-scale production. These nanopatch antennas hold great promise for practical applications, for example in sensing, ultrafast optoelectronic devices and for high-efficiency photodetectors.

A protocol for the colloidal synthesis of silver nanocubes and fabrication of plasmonic nanoscale patch antennas with sub-10 nm gaps is presented.

We present a method for colloidal synthesis of silver nanocubes and the use of these in combination with a smooth gold film, to fabricate plasmonic ...

Experimental Methodology for Estimation of Local Heat Fluxes and Burning Rates in Steady Laminar Boundary Layer Diffusion Flames

Modeling the realistic burning behavior of condensed-phase fuels has remained out of reach, in part because of an inability to resolve the complex interactions occurring at the interface between gas-phase flames and condensed-phase fuels. The current research provides a technique to explore the dynamic relationship between a combustible condensed fuel surface and gas-phase flames in laminar boundary layers. Experiments have previously been conducted in both forced and free convective environments over both solid and liquid fuels. A unique methodology, based on the Reynolds Analogy, was used to estimate local mass burning rates and flame heat fluxes for these laminar boundary layer diffusion flames utilizing local temperature gradients at the fuel surface. Local mass burning rates and convective and radiative heat feedback from the flames were measured in both the pyrolysis and plume regions by using temperature gradients mapped near the wall by a two-axis traverse system. These experiments are time-consuming and can be challenging to design as the condensed fuel surface burns steadily for only a limited period of time following ignition. The temperature profiles near the fuel surface need to be mapped during steady burning of a condensed fuel surface at a very high spatial resolution in order to capture reasonable estimates of local temperature gradients. Careful corrections for radiative heat losses from the thermocouples are also essential for accurate measurements. For these reasons, the whole experimental setup needs to be automated with a computer-controlled traverse mechanism, eliminating most errors due to positioning of a micro-thermocouple. An outline of steps to reproducibly capture near-wall temperature gradients and use them to assess local burning rates and heat fluxes is provided.

We describe the use of micro-thermocouples to estimate local temperature gradients in steady laminar boundary layer diffusion flames. By extension of the Reynolds Analogy, local temperature gradients can be further used to estimate the local mass burning rates and heat fluxes in such flames with high accuracy.

Modeling the realistic burning behavior of condensed-phase fuels has remained out of reach, in part because of an inability to resolve the complex ...

Fabrication of Nanopillar-Based Split Ring Resonators for Displacement Current Mediated Resonances in Terahertz Metamaterials

Terahertz (THz) split ring resonator (SRR) metamaterials (MMs) has been studied for gas, chemical, and biomolecular sensing applications because the SRR is not affected by environmental characteristics such as the temperature and pressure surrounding the resonator. Electromagnetic radiation in THz frequencies is biocompatible, which is a critical condition especially for the application of the biomolecular sensing. However, the quality factor (Q-factor) and frequency responses of traditional thin-film based split ring resonator (SRR) MMs are very low, which limits their sensitivities and selectivity as sensors. In this work, novel nanopillar-based SRR MMs, utilizing displacement current, are designed to enhance the Q-factor up to 450, which is around 45 times higher than that of traditional thin-film-based MMs. In addition to the enhanced Q-factor, the nanopillar-based MMs induce a larger frequency shifts (17 times compared to the shift obtained by the traditional thin-film based MMs). Because of the significantly enhanced Q-factors and frequency shifts as well as the property of biocompatible radiation, the THz nanopillar-based SRR are ideal MMs for the development of biomolecular sensors with high sensitivity and selectivity without inducing damage or distortion to biomaterials. A novel fabrication process has been demonstrated to build the nanopillar-based SRRs for displacement current mediated THz MMs. A two-step gold (Au) electroplating process and an atomic layer deposition (ALD) process are used to create sub-10 nm scale gaps between Au nanopillars. Since the ALD process is a conformal coating process, a uniform aluminum oxide (Al2O3) layer with nanometer-scale thickness can be achieved. By sequentially electroplating another Au thin film to fill the spaces between Al2O3 and Au, a close-packed Au-Al2O3-Au structure with nano-scale Al2O3 gaps can be fabricated. The size of the nano-gaps can be well defined by precisely controlling the deposition cycles of the ALD process, which has an accuracy of 0.1 nm.

A protocol for the design and fabrication of a novel nanopillar-based split ring resonator (SRR) is presented.

Terahertz (THz) split ring resonator (SRR) metamaterials (MMs) has been studied for gas, chemical, and biomolecular sensing applications because the SRR ...

Plasma-Assisted Molecular Beam Epitaxy Growth of Mg3N2 and Zn3N2 Thin Films

This article describes a procedure for growing Mg3N2 and Zn3N2 films by plasma-assisted molecular beam epitaxy (MBE). The films are grown on 100 oriented MgO substrates with N2 gas as the nitrogen source. The method for preparing the substrates and the MBE growth process are described. The orientation and crystalline order of the substrate and film surface are monitored by the reflection high energy electron diffraction (RHEED) before and during growth. The specular reflectivity of the sample surface is measured during growth with an Ar-ion laser with a wavelength of 488 nm. By fitting the time dependence of the reflectivity to a mathematical model, the refractive index, optical extinction coefficient, and growth rate of the film are determined. The metal fluxes are measured independently as a function of the effusion cell temperatures using a quartz crystal monitor. Typical growth rates are 0.028 nm/s at growth temperatures of 150 &#176;C and 330 &#176;C for Mg3N2 and Zn3N2 films, respectively.

Plasma-Assisted Molecular Beam Epitaxy Growth of Mg3N2 and Zn3N2 Thin Films

This article describes the growth of epitaxial films of Mg3N2 and Zn3N2 on MgO substrates by plasma-assisted molecular beam epitaxy with N2 gas as the nitrogen source and optical growth monitoring.

This article describes a procedure for growing Mg3N2 and Zn3N2 films by plasma-assisted molecular beam epitaxy (MBE). The films are grown on 100 oriented ...

Robotic Sensing and Stimuli Provision for Guided Plant Growth

Robot systems are actively researched for manipulation of natural plants, typically restricted to agricultural automation activities such as harvest, irrigation, and mechanical weed control. Extending this research, we introduce here a novel methodology to manipulate the directional growth of plants via their natural mechanisms for signaling and hormone distribution. An effective methodology of robotic stimuli provision can open up possibilities for new experimentation with later developmental phases in plants, or for new biotechnology applications such as shaping plants for green walls. Interaction with plants presents several robotic challenges, including short-range sensing of small and variable plant organs, and the controlled actuation of plant responses that are impacted by the environment in addition to the provided stimuli. In order to steer plant growth, we develop a group of immobile robots with sensors to detect the proximity of growing tips, and with diodes to provide light stimuli that actuate phototropism. The robots are tested with the climbing common bean, Phaseolus vulgaris, in experiments having durations up to five weeks in a controlled environment. With robots sequentially emitting blue light-peak emission at wavelength 465 nm-plant growth is successfully steered through successive binary decisions along mechanical supports to reach target positions. Growth patterns are tested in a setup up to 180 cm in height, with plant stems grown up to roughly 250 cm in cumulative length over a period of approximately seven weeks. The robots coordinate themselves and operate fully autonomously. They detect approaching plant tips by infrared proximity sensors and communicate via radio to switch between blue light stimuli and dormant status, as required. Overall, the obtained results support the effectiveness of combining robot and plant experiment methodologies, for the study of potentially complex interactions between natural and engineered autonomous systems.

Distributed robot nodes provide sequences of blue light stimuli to steer the growth trajectories of climbing plants. By activating natural phototropism, the robots guide the plants through binary left-right decisions, growing them into predefined patterns that by contrast are not possible when the robots are dormant.

Robot systems are actively researched for manipulation of natural plants, typically restricted to agricultural automation activities such as harvest, ...

Computer Vision-Based Biomass Estimation for Invasive Plants

Samenvatting

Meer video's verkennen

Hoofdstukken in deze video