Detection of Defected Maize Leaf using Image Processing Techniques
Arabinda Dash, Prabira Kumar Sethy
Abstract
One of the major obstacles to producing Maize is the various types of maize leaf diseases. Early detection using computer vision and image processing techniques can be one of the most powerful solutions to overcome this problem. This paper proposes a detection method based on image processing techniques to identify the defected maize leaf infected by the diseases like Gray-leaf spot, Common as well as southern rust, and northern leaf blight. In our proposed method, the subtracted channel image is calculated by subtracting the extracted green channel from the red channel of the corresponding processed image. After then, the image segmentation is done by converting the subtracted channel image to a binary image. Finally, the detection of infected leaves is done by taking the decision based on calculating the number of white pixels present on the leaf images. The processed images that include both infected and healthy leaves have only been considered for the experimental analysis. The results show that our proposed method successfully detects the defected maize leaves from the healthy leaf images.