ASABE Technical Library - Abstract

Abstract. Computer vision system has been proven to be successful for papaya measurement in agro-industry sorting. Sixty papayas from one cultivar were sampled to capture their images on four sides by a webcam imaging system and their 1,680 data points were analyzed from visually classified fruits at harvest in seven maturity stages according to the farmers‘ standard. The image processing has been used to determine color and extract shape features and assess the quality of papaya. MATLAB programming was used to extract the color and shape features by expressing colors in CIE L*a*b* color space, to calculate color index and extracting shape descriptors to determine weight and shape. The relationship between color index and total soluble solid (TSS) were determined. Determination of weight was estimated by ellipsoid volume estimation. The results showed that color index and TSS had a strong relationship with R² of 0.97 (sigmoid function), while the weight estimation was highly determined with R² of 0.971 (linear regression) and determination of papaya shape determined by shape index and elongation. These results suggest that with sigmoid function can correlate better for color index and TSS. This nonlinear relation is similar to the trend of respiration rate in ripen process.