NEW APPROACHES FOR SIZE DETERMINATION OF APPLE FRUITS FOR AUTOMATIC SORTING AND GRADING

Nondestructive quality evaluation of fruits is important and very vital for the food and agricultural industry. The fruits in the market should satisfy the consumer preferences. Traditionally grading of fruits is performed primarily by visual inspection using size as a particular quality attribute. Image processing offers solution for automated fruit size grading to provide accurate, reliable, consistent and quantitative information apart from handling large volumes, which may not be achieved by employing human graders. A fruit grading system based on machine vision technology to assess the fruits quality would bring economical benefits to the fruit industry by increasing consumer/buyer confidence in the quality of the fruit. Conventional methods of size determination are not suitable for objects of irregular shape. Majority of the apples have good degree of symmetry about the stem and calyx axis. The paper discusses three different methods to determine the size of this category of apples by applying the known geometrical models. Three more methods are presented in the paper to determine the size of the other category of apples. The Apple fruit images in different postures are captured to cover the entire fruit, using a CCD camera. The analysis is carried out using MATLAB ® package on Windows ® platform. It was found that though all the methods outlined in the paper could be used, some methods achieve higher performance. RUITS and vegetables are graded based on their external factors like size, shape, color, external damage, etc. Classification of fruits based on size is one of the most important tasks performed in the packinghouses. Machine vision technology has been proved to be an excellent tool that could be effectively used to replace human sorters for consistent and reliable judgment in estimating and comparing sizes of the fruits. Many methods like area measurement, linear regression relations between the areas taken from two different points of view, comparing measurement of length, width, perimeter etc., may be employed to estimate and compare fruit sizes. Ratio of length of major and minor axes, contour of random views of the fruits can be used for size comparison. Fruits, vegetables, and cereals size determination using