Enhancing Apple Disease Detection with Pixel Pyramid Net through Pixel-Level Segmentation and Spatial Pyramid Pooling Techniques

Background: Apple fruit is prone to various diseases that can significantly affect fruit quality. Methods: This paper proposes a novel approach for apple disease detection utilizing PPN-Pixel Pyramid Net, a neural network architecture designed for pixel-level segmentation tasks. The methodology incorporates spatial pyramid pooling techniques to effectively address input images of diverse sizes and aspect ratios. Result: Experimental results showcase the model’s performance metrics across different classes, providing an in-depth analysis of Accuracy, Intersection over unionand Mean BF Score.