Detection of diseases in fruits using Image Processing Techniques

India is one of the leading producers of fresh fruits in the world but its contribution to the global market in terms of exports is very little. One of the reasons for this huge difference is the significantly high wastage of the produce due to the unavailability of systems for the detection of diseases in fruits efficiently, during the harvest and in the post-harvest period. In this paper, two approaches have been applied for the detection of disease in Apple, namely, Multi Support Vector Machine and Convolutional Neural Networks. A comparative analysis has been carried out on the results obtained using the aforementioned approaches. In the Multi Support Vector Machine (Multi SVM) approach, K-means clustering has been used for segmentation and Gray Level Co-Occurrence Matrix (GLCM) has been used for Feature Extraction whereas in the Convolutional Neural Network approach, transfer learning using ResNet-50 has been used for the detection of diseases in Apples. The results obtained from the two approaches can be summarized as 86.3% average classification accuracy in the Multi SVM approach and 95.79% accuracy in the CNN approach