Development of Citrus Grading System Using Image Processing

Abstract

Citrus production in Indonesia has been increasing to a yield of J 7-25 tons/ha. However, the quality is still a serious problem. Most of the citrus produced from the production centers such as Pontianak and Medan enter domestic market with unattractive appearance with big varieties in size, color, and taste. For that reason, the quality improvement of citrus through the development of quality assurance system using new technologie's is necessary. Many postharvest handling technologies such as waxing and modified atmosphere packaging can be applied, but uniformity in size, color, and taste is very important in marketing the products. One of the advanced technologies that can be used for sorting and grading fruits is an automatic grading system with image processing for quality measurement. Image processing technology is a technology developed to obtain information from image by modifying the image into a desired and more informative one and analyzing it, or translate the meaning into an action in machine vision. Image processing technology has been applied to detect cherry tomato in a bunch of cherry tomato plants by recognizing object with different color. The method was integrated in the harvesting robot for cherry tomato cultivated in a greenhouse [1]. Another example of image processing application is in mushroom harvesting robot, to detect the mature mushroom to harvest [2] and in watermelon harvester to do the same task [3]. The objectives of this study were to develop software using image processing to grade citrus based on their weight and color, and to develop a real-time grading machine prototype.