Feature Detection and Analysis of Pigeon Peas Using Image Processing

Abstract

Image processing is one of the challenging areas in computer science which finds its application and place in almost every area of research. Pigeon peas, which are popularly called as Toor dal in India, is widely used as one of the main ingredients in food globally, as it is rich in proteins. One of the challenges faced by the food industry is in assessing the quality and the type of the grains and pulses. The quality detection by manual inspection in analysing the grains and pulses may not give accurate results compared to an automated system. For a common man, the colour, size and texture of dals are important parameters in deciding their quality. But in some cases, more the colour lesser the quality as the dals may be polished. This paper focuses on developing a model which can extract the geometrical features and analyses in classifying pigeon peas into different groups. Digital samples of both polished and unpolished pigeon peas are taken for analysing and classifying them based on three parameters: size, colour and texture. The images are processed using matlab, a software tool.