Our goal is to successfully classify plant diseases by image-based identification of the plant leaves. We aim to detect and provide the crop-disease pair for the crops or indicate if they are healthy otherwise. Machine learning would help identify these crop diseases quicker, identify newer diseases in crops the farmers may not have seen in this part of the world which will greatly benefit the Canadian agricultural industry. Additionally, this tool could be used in smaller farms where inexperienced farmers could identify the disease and treat it accordingly.
The following image provides a qualitative analysis of our model on unencontered data.