Detection of Cotton Plants Using Deep Learning Based Object Detection Model

Object recognition in factual surroundings is a challenging duty because of erratic ambient lighting and varying properties of objects such as shape, size, and color. This study describes the deep learning (DL)-based object recognition model to identify cotton plants. DL-based YOLOv5 model was developed to recognize cotton plants due to its high interference speed. The cotton recognition model employs a technique based on convolutional neural networks (CNNs). The developed cotton recognition model was deployed in a microprocessor (Raspberry Pi) to identify the cotton plants from input images. Cotton recognition is useful to increase mechanization or automation of cotton farming processes such as specific site applications, spraying, dosses optimization, insect control, robotic harvester, and weeding. Precision, recall, F1, and mAP were determined to be 1.00 at 0.696 confidence, 1 at 0.000 confidence, 0.98 at 0.589 confidence, and mAP 0.879 at 0.5 confidence, respectively.