The Cardamom plant diseases cause significant production and economic losses. To increase crop yield, it is crucial to detect plant diseases early and take effective control measures. To overcome the shortcomings of traditional visual observation techniques, digital image processing techniques have been used to detect plant diseases quickly and accurately. In this article, we presented an advanced feature selection technique for detecting diseases related to cardamom plant leaf. This involves various steps such as image loading, pre-processing, segmentation, extraction and classification. The method describes various bacterial/fungal diseases and uses image processing techniques to identify and classify these diseases. The proposed method uses Fuzzy K-means clustering method to extract disease affected regions of cardamom plants. The system is tested on healthy or diseased cardamom leaves and plants. The proposed method tested on 300 images of healthy and diseased leaves of cardamom plants. The Support Vector Mechanism (SVM) classifies these images into normal and abnormal images. A comprehensive set of experiments was conducted to check the performance of the proposed approach and compare it with other leaf disease detection models. Experimental results showed that the proposed approach achieves recognition accuracy of over 96%.