Plant Leaf Detection by Multi-Class SVM Model Using CCM and Histogram Features

Abstract:-Agriculture, serving as the cornerstone of civilizations, plays a pivotal role in providing sustenance and essential resources. Given its paramount significance in human life as a primary source of food, the detection of plant diseases has emerged as a critical concern. Although traditional methods exist for identifying such diseases, agriculture professionals and plant pathologists have traditionally relied on visual inspection alone to detect leaf diseases. However, this conventional approach to identifying plant leaf diseases can be subjective, time-intensive, financially burdensome, and necessitates a substantial workforce equipped with extensive knowledge about various plant diseases. This paper has proposed a model that classify the plant unhealthy leaf and identify the type of infected image. Paper has found that proposed model has increases the work detection accuracy by the use of Co-occurrence matrix and histogram feature. Extracted features were used for the training of multi class support vector machine. Experiment was done on real tomato plant leaf dataset and result shows that proposed model has increases the detection accuracy of multi class leaf diseases as well.

---