Cucumber disease recognition using machine learning and transfer learning

Cucumber is grown, as a cash crop besides it one of the main and popular vegetables in Bangladesh. As Bangladesh's economy largely dependent on agricultural sector, cucumber farming could make economic productivity growth more sustainable. But many diseases diminish situation cucumber. Early detection disease can help to stop from spreading other healthy plants also accurate identifying will reduce losses through specific treatments. In this paper, we have presented two approaches namely traditional machine learning (ML) CNN-based transfer learning. Then compared performance applied techniques find out most appropriate for recognizing diseases. our ML approach, system involves five steps. After collecting image, pre-processing done by resizing, filtering, contrast-enhancing. various algorithms using k-means based image segmentation after extracted 10 relevant features. Random forest gives best accuracy with 89.93% approach. We studied investigate further improvement recognition performance. Lastly, comparison among models such InceptionV3, MobileNetV2, VGG16 has been performed. Between these approaches, MobileNetV2 achieves highest 93.23%.