The The Crop Disease Detection on Vertical Farming using Image Processing

Abstract

Presently, the indoor farming technology developed to accommodate the food demand. But, there still having food loss mainly due to crop infected by disease, which reflexively reduces the production rate. The main challenge is to improve the crop disease control and increases the crop quality and quantity in indoor farming mostly in vertical structure. The purpose of this research is to build crop disease detection system for vertical farming using image processing. Image processing method is one of computer vision technology use to identify and process the object in image. Hence, the image processing method is applied for the plant disease’s detection. Disease detection involves the steps like image acquisition, image pre- processing, image segmentation, feature extraction and disease classification. Initially, the input images were retrieved from vertical farming using digital camera. Then, in image segmentation steps, k-mean clustering used to predict the infected area of leaves. K-mean clustering is a color based segmentation model to segment the infected region and placing it to its relevant classes. The feature of disease spot was extracted based on its texture and colour. Multi support vector machine technique (MultiSVM) were used to categorize feature and finally, classify the disease. Experimental analyses were done on four samples of Choy Sum plant to detect disease name, percentage area of infection and accuracy. Among the four plants tested, plant number 4 show the highest accuracy of downy mildew disease with 98.3871% with area of infection 11.0154% while the plant number 1 show the highest accuracy of healthy leaf also with 98.3871%. Lastly, this project successfully builds the GUI for whole disease detection system.