Digital Image Processing for Detecting and Classifying Plant Diseases

Abstract

Agriculture is the backbone of human sustenance in this world. With growing population, there is need for increased productivity in agriculture to be able to meet the demands. Diseases can occur on any part of a plant, but in this paper only the symptoms in the fruits of a plant is considered using segmentation algorithm and edge/ sizing detectors. We also looked at image processing using fuzzy logic controller. The system was designed using object oriented analysis and design methodology. It was implemented using MySQL for the database, and PHP programming language. This system will be of great benefit to farmers and will encourage them in investing their resources since crop diseases can be detected and eliminated early.

Convolutional Neural Networks for Classification Motives and the Effect of Image Dimensions

“…p represents the p-th layer, kij represents the convolutional kernel, Nj represents the bias, and Mj represents a set of input mappings. Unsupervised learning is commonly used to train different architectural elements, such as bias and kernel weights [19].…”

“…Before stochastic merging, it is necessary to compute the probability p for every region. The calculation formula is shown in Equation 2 [19].…”

“…Sj represents the amalgamation of region j, where F denotes the attributes, and i indicates the location of each element within region j. The stochastic variable St is utilized in the pooling procedure for each successive map F. Equation 3 represents the mathematical notation for the stochastic variable St [19].…”