Non-destructive Image Processing Analysis for Defect Identification and Maturity Detection on Avocado Fruit

Image processing techniques and computer intervention has been widely accepted in smart agriculture and smart farming applications such as classification, quality grading, sorting, detection, evaluation in the agricultural field. Defect identification and maturity detection on avocado fruit give a challenging task by integrating a deep learning system nearer to a human-level recognition. The proposed research will benefit local markets and traders of purple avocados as it will provide automatic sorting and grading evaluation based on local standards. This paper aims to develop a non-destructive image processing analysis for defect identification and maturity detection on avocado fruit capable of classifying the type of defect and its maturity based on a convolutional neural network (CNN) classifier. This study had an overall accuracy of 93% for defect identification and 98.97% for maturity detection.