The Comparative Analysis of The Filtering Schemes In Crop Disease Detection

Abstract

The rice plants are affected by various rice diseases such as rice impression, Bacterial scourge, Black kernel, Brown spot, Leaf smut, Seeding blight in different growth stages of the rice crops. The emerging trend in the agriculture is Computer Aided Detection (CAD) based on the digital imaging of the rice crops provides the detailed analysis about the crop disease by applying the image mining process. In order to detect the crop disease, CAD system applies the image segmentation and image mining process by acquiring the image of the crop during the transplantation stage. The captured image is applied for the noise removal and segmentation process to characterize the domain of interest from the crop image. The noise removal process is applied by using the Cross Central Filter which combines the features of Gaussian filter and Hybrid Median Filter operations. Core Fuzzy C-Means algorithm is applied to Image segmentation process for the object division. The Cross Central Filter produces the least Mean Square Error approximation value as of 4.71 and eminent Peak Signal to Noise Ratio value as 62.615 dB. Core Fuzzy C-Means achieves precision value as 95% and specificity as 96% values.