Enhancing apple disease detection with pixel pyramid net through pixel-level segmentation and spatial pyramid pooling techniques

Apple fruit is prone to various diseases that can significantly affect fruit quality.

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.

Experimental results showcase the model’s performance metrics across different classes, providing an in-depth analysis of Accuracy, Intersection over unionand Mean BF Score.