Maize Leaf Disease Severity Analysis using Integrated Color Filtering and Threshold Masking

Abstract

As the color is a dominant factor for isolation of the diseased part of corn maize, the detachment of disease affected parts in maize plant leaf is achieved using integrated Color Filtering followed by a threshold masking. The Particular HSV from a color image of maize plant is extracted .Major steps involved here is to initially convert an RGB image of disease affected maize plant to HSV and second is to apply a threshold mask to filter out the green color of healthy maize plant and detach the brown and yellow diseased area thereof. This method is applied and tested with around 30 maize leaves, and the results found that the proposed methodology performs well with overlapped healthy maize leaf compared to K-means Clustering algorithm. False Positive is produced in K-means method and this Proposed system as integrates with Color Filtering and thresholding works well with overlapped images so that it increases True negative as the Accuracy of the proposed method increases. This proposed methodology identifies well with perfect maize leaf images and misclassifications occur only with images with dark shadows, light illuminations and sanded background.

Customized Semantic Segmentation for Enhanced Disease Detection of Maize Leaf Images

“…[9] Maize leaf RGB images are converted to HSV color space and then using color filtering Technique non-green color pixels are filtered out and a threshold mask is applied for segmenting out disease affected regions ROI which is extracted from the original image and their Entropy, Mean and IDM values of around 50 leafs are extracted and compared. [10]…”