A Smart Hybrid Model for Fruit Quality Grading: Merging Canny Edge Detection and K-means Clustering Algorithm

Fruit recognition and grading are essential steps in fruit quality screening. In this research, a novel approach that combines the power of Canny edge detection and K-means clustering algorithms for fruit quality grading is proposed. The application of multi-channel clustering to process fruit images enables effective separation of the main body and background, followed by feature extraction through contour detection and filtering. Moreover, a Transfer Learning model utilizing an 18-layer residual neural network is trained to perform fruit classification, while digital image processing techniques are employed to extract relevant image features for fruit grading. To evaluate the accuracy and efficiency of our method, a dataset of 2,000 images was utilize to conduct the fruit recognition and grading experiments using Matlab software. The experimental results using apple samples show that our approach achieves a recognition accuracy of 95% and reduces processing time to less than one second, significantly improving efficiency in real-time applications.