Food Quality Inspection and Grading Using Efficient Image Segmentation and Machine Learning-Based System

V Hemamalini,S Nachiyappan,T Devi,S Rajarajeswari,Bhupesh Kumar Singh,Abhishek Raghuvanshi,M Sambath,Rijwan Khan

doi:10.1155/2022/5262294

Abstract

One of the most critical aspects of quality assurance is inspecting products for defects before they are sold or shipped. A good product is more vital than having more of the same item for a customer’s enjoyment. The client has a significant role in determining the quality of a product. Another way to think about quality is as the total of all the characteristics that contribute to the creation of items that the client enjoys. Recently, the application of machine vision and image processing technology to improve the surface quality of fruits and other foods has increased significantly. This is primarily because these technologies make significant advancements in areas where the human eye falls short. This means that, by utilizing computer vision and image processing techniques, time-consuming and subjective industrial quality control processes can be eliminated. This article discusses how to check and assess food using picture segmentation and machine learning. It is capable of classifying fruits and determining whether a piece of fruit is rotten. To begin, Gaussian elimination is used to remove noise from images. Then, photos are subjected to histogram equalization in order to improve their quality. Segmentation of the image is carried out using the K-means clustering technique. Then, fruit photos are classified using machine learning methods such as KNN, SVM, and C4.5. These algorithms determine if a fruit is damaged or not.

Highlights

Quality is more crucial to the customer’s delight than supplying more items of the same kind [1]. e client is an important factor in determining product quality
The quality of an already existing set of images is made better by the histogram equalization method. is improves the accuracy of the classifiers
Image segmentation is performed using the K-means clustering algorithm. en machine learning methods such as KNN, support vector machine (SVM), and C4.5 are used for classifying fruit images. ese algorithms classify fruits as damaged or good

Summary

Introduction

Quality is more crucial to the customer’s delight than supplying more items of the same kind [1]. e client is an important factor in determining product quality. E color, size, and form of fruits might vary even if they were harvested from the same tree on the same day [10] Depending on their ripeness and storage circumstances, food goods naturally change color or texture after harvest (humidity and temperature, fungal infections, presence of volatile substances, storage duration, etc.). It is critical to identify any foreign material on quality control lines, such as stems, leaves, dirt, or blemishes on the skin, and not mistake them for the real thing [11, 12] After harvest, these traits are determined by the fruit’s ripeness and the storage circumstances under which it was kept (humidity and temperature, fungal infections, presence of volatile substances, storage duration, etc.). It is capable of grading fruit into different categories and identifying defective fruits

Literature Survey

Methodology and Results

Conclusion