Estimation of size and shape of citrus fruits using image processing for automatic grading

摘要

Size is one of the important parameters in grading of fruits. Also quantifying the fruit's shape gives value addition to the fruits. This paper describes analytical methods to estimate the size and shape of citrus fruits to grade them based on single view fruit images. Sweet-lime and orange fruits are taken for case study of size and shape determination respectively. The size of the sweet-lime fruits were estimated and graded into three categories using simple methods like radius signature method, area method and perimeter method. Also the existing method based on Light Detection and Ranging (LIDAR) sensor for citrus fruits size determination was improved through a method employing image processing. The shapes of the orange fruits were estimated using Heuristic Shape separator method and shape numbers were obtained for varying shaped orange fruits. The results were found to be reasonably in good agreement with the human assessment.