Computer Vision-Based Biomass Estimation for Invasive Plants

Our Master utilize UAV mono sensing and computer reason to efficiently estimate in waste plant biomass, overcoming the inefficiencies and subjectivity of traditional manual surveys and data collection in complex environments. Recent advancement in UAV remote sensing for plant biomass estimation include high resolution imaging and multimodel data integration, these developments enhance scientific research, environmental monitoring, and agricultural management by offering more precise data support. For advancements in UAV remote sensing for plant biomass estimation encumbers high resolution sensors, multispectral/hyperspectral imaging, GBS/INS, specialized data processing software, AI and deep learning, improve batteries, enhanced data storaging, and cloud computing.

Relate to traditional brand farmers is automatization methods, our proposed map doesn't require it's speedy mirror service and data cutting, and for other modeling special, hyperspectral is dimension method, we don't need a special requirement. Our method can agricultural and accurately is to be plant primaries without experience requirement, which is our advantage. Our laboratory will continue to study UAV remoting sensors, plant pharmacy is amazing.

In the future we will focus on firmware operating, deep running organisms, applying in multi saw data filting, right, continue in deep results on plant preference and the sensor terminology innovations. To begin, employ a random sampling approach with enough biomass samples to determine soil sampling locations in the study area. Position the sample frame horizontally over the vegetation, fully encompassing the plants with a minimum separation distance of two meters between each plant.

Use a drone and a camera to build an unmanned aerial vehicle or UAV remote sensing filming system. To plot the root within the specified study area using the UAV, first establish a heading inside overlap rate of 70%maintain the camera perpendicular to the ground and at an altitude of 30 meters to capture photos at uniform time intervals of two seconds. Run the script file to store the aerial imagery for subsequent processing with Python software for biomass estimation.

For creating the dataset using Python programming, proceed to segment the raw image data into smaller images of size the same as the sample images. Use the sliding window method for segmentation, setting the horizontal and vertical steps to 280 pixels. From the segmented small images randomly select 880 invasive plant images and 1500 background images to create a dataset, then split this data set into training, validation, and test sets in a six to two to two ratio.

To begin, collect the above ground biomass of chamomile manually within each sample plot after completion of drone data collection. While collecting, cut the chamomile along the inside edge of the sample plot, then cut the rhizome of the chamomile from the bottom, remove any dirt, rocks, or other plants that are mixed in, finally, bag and label the samples. Next, bring the collected samples of invasive plants to the laboratory.

Air dry all samples to evaporate most of the moisture. Dry the samples in an oven at 55 degrees Celsius for 72 hours to remove any residual moisture. Place a bag of chamomile on an electronic balance and record the stabilized readings in grams.

Once done, calibrate again and continue weighing. Weigh the chamomile every hour until the mass no longer changes. Subtract the weight of the bag from the reading to obtain the measured mass of that sample.

Finally, calculate the above ground biomass of the invasive plant using the formula displayed. To prepare the software, download and install Anaconda from its official website. Launch the PyCharm IDE program, then open the Anaconda prompt command line and 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. Open the Anaconda prompt and activate the PyTorch environment by entering conda activate pytorch, type nvidia smi to check the current compute unified device architecture or CUDA version. Then install PyTorch version 1.8.1 by running the command conda install pytorch=1.8.1 torchvison=0.9.1 torchaudio=0.8.1 cudatoolkit=11.0 c pytorch.

To run the model recognition, pre-process the images to prepare them for model input. Using the displayed code, resize the images from 280 by 280 pixels to 224 by 224 pixels and normalize them to ensure they meet the model size requirements. Train a multi-class recognition model with the already created dataset by setting 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. Right click and press run script, then employ a meticulously trained recognition model and systematically traverse the original image for identification purposes.

Configure horizontal and vertical steps precisely at 280 pixels to result 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. Perform simple data augmentation with the random resized crop and random horizontal flip functions to extend the image set and extract the six vegetation indices.

To ensure precise estimation of the biomass of invasive plants, create a KNN nearest neighbor regression model using the output and the extracted vegetation indices as inputs. Mikania micrantha, could be observed climbing atop the plant adorned with white flowers, the other plants, as well as the road and accompanying elements, were uniformly depicted in the background. The model recognized the red part as mikania micrantha, demonstrating robust detection in complex backgrounds.

The regression analysis demonstrated strong predictive performance with an R square value of 0.62 and an RMSE of 10.56 grams per square meter. The model enhanced the accuracy of chamomile biomass estimation and the spatial distribution map effectively captured the distribution of chamomile biomass.