External Features-Based Approach to Date Grading and Analysis with Image Processing

The analysis and classification of dates is based on their external features: size, appearance, shape, and colour. The process is currently performed manually after harvesting as part of the post-harvesting process. Grading manually is tedious because it usually results in time delays, product quality risks, and it is associated with time and cost delays as well. Although the use of computers and information technology has seen tremendous growth in many small and large sectors, it has been in its infancy in the cultivation of fruit and dates. Using image processing algorithms, we can enhance human vision capabilities through analysis and make images easier to comprehend. A major objective of computer vision-based algorithms for classifying and sorting of dates is to make the procedure fully automated by minimizing the manual component involved in the process. This paper presents an image processing-based algorithm that uses machine learning techniques to extract the characteristics of colour intensity and colour homogeneity, allowing us to grade images in a more timely and automated manner. In order to obtain the results, we extracted the appearance of the date images based on an image processing algorithm. It is used as a validation element for the results that the quality of dates-fruit images can be evaluated through the prior selection process in both separate and in groups. This study has managed to achieve a rate of 95% accuracy in data classification.