Biological Systems Engineering



Rahman 0000-0001-9520-2753

Date of this Version



Journal of Agricultural, Environmental and Consumer Sciences (2022) 22: 5–10



Postharvest processing of agricultural produce is still done the conventional way in Bangladesh. Manual grading of agricultural produce, especially fruits and vegetables, is laborious and costly due to acute shortage of labor during the peak season, as well as difficulty maintaining the product quality. Machine vision system (MVS) applications are widely used nowadays as a non-destructive and cost-effective technology for automatically grading and sorting large volumes of produce in the packing house according to size, shape, color, texture, and surface defects. In this study, a simple MVS was constructed measuring different color features of mango fruit surface as a part of developing an automatic grading system. A CCD camera with a fluorescent lighting system was incorporated for acquiring images of mangos. Different color properties were extracted from the acquired images and analyzed. The best-suited color information (HSI model) was found so that the fruits can be separated from their background easily and differentiated. The measured color information will be further used for developing a grading algorithm based on different features of mango, further aimed to develop an automatic mango grading system.


In this study, it was determined that the developed image acquisition system was suitable to recognize the target objects (e.g., green and ripe mangos) from its background by using image color properties. Moreover, the HSI color threshold information was found more suitable than RGB to identify the green and ripe mangos. Therefore, the color information could be used for developing a mango grading algorithm. The color information can be combined with the physical dimension and defect information for more accurate grading of fruits which is a crucial need in the Bangladesh agriculture sector.