Apple is one of the most popular plants in the Indian-origin Kashmir valley. Every year, the apples from Kashmir are exported to other areas of the globe, creating a substantial amount of revenue. However, apple trees are prone to several diseases which devastate apple yields and cause major losses for apple growers. Disease in apple plants mostly originates in the plant leaves. The prompt detection and prediction of such diseases is thus essential in a country like India, where half of the population does farming. The conventional methods of apple plant disease prediction are time-consuming and laborious, involving lab assistance to diagnose the diseases. With the advent of machine learning and deep learning, it is now possible to quickly determine if a plant is infected or not with reliable accuracy. In this article, we introduce D-KAP, a deep learningbased Kashmiri apple plant disease prediction framework capable of detecting several apple plant diseases. For feature extraction and prediction, our model employs the advanced deep learning capabilities of Convolutional Neural Networks (CNN). Our framework produces state-of-the-art results in identifying apple plant diseases with an accuracy of 92 percent over testing samples. In addition, we also introduce a novel Kashmiri apple plant leaf dataset containing samples of three distinct diseases along with healthy leaves.

No abstract