A STUDY OF THE GRADING PROCESS FOR AGRICULTURAL FOODS USING ARTIFICIAL INTELLIGENCE, MACHIN LEARNING TECHNOLOGY

Abstract

Fruit anomalies can be detected using a multi-spectral image sensor in a computer vision system. As a starting point, the suggested method relies entirely on Near-Infrared (NIR) photographs to segment the fruit. Second, specific adaptive pre-processing methods are applied on the segmented RGB and NIR fruit images. By using thresholding techniques, the fruit's seven different colour components may be inspected for defects. Finally, the seven threshold colour component photos are put to a vote to see if the fruit image is defective. In the recent year, companies that provide services and deliver packages have changed. Online shopping provides various advantages for the postal and courier industry. Cartons or wooden boxes of various sizes are used to package the goods sold by the merchant. It is possible to depict the shape of an object by using a contour-based object detection algorithm. A contour is a representation of geometric principles in the form of edge or curve components. To determine the object's surface area, one must first determine its dimensions. Dimensions are widely used to measure the length, width, and height of an object.