Tomato Grading System by Image Processing and Machine Learning

Goh, Chun Hou (2020) Tomato Grading System by Image Processing and Machine Learning. Final Year Project (Bachelor), Tunku Abdul Rahman University College.

Text Goh Chun Hou.pdf Restricted to Registered users only Download (2MB)

Abstract

Nowadays, consumer demand more on good quality agricultural product and the grading system is used to determine the product quality. Manual grading system is commonly used in different agricultural product but manual grading system is time consuming, low efficient and labour intensive. To overcome the limitations of manual grading system, automatic grading system based on image processing technique and machine vision technique have been developed and gradually phase out the manual grading system. Besides, machine learning techniques have been combined together with automatic grading system to increase the accuracy in classification. Tomato is one of the most familiar agricultural products in Malaysia to its high market value in local and global market. The quality of tomato can be determined by the colour, appearance and size and the colour is the most considered factor for the consumer. The colour of tomato can be classified in to 6 classes which are green, breakers, turning, pink, light red and red. There are different type of automatic tomato grading system have been proposed by the researchers and the method or technology used are different. The objective of this research is to design and develop an automatic tomato grading system with most effective image processing technique and the most suitable machine learning algorithm by using MATLAB. The tomato images captured will undergo a series of image processing and used as inputs for the machine learning classifier. Then the classifier will classify the tomato into different group according to the surface colour. The results obtained are expected to display on MATLAB GUI.