Classification of Soil Image using Feature Extraction

There are various methods for soil classification using different algorithms. In the image processing to describe the texture of image using statistical and geometric feature. Signal processing techniques are used to filters images and calculate features of transformed images. The purpose of proposed algorithm is to perform classification ofby making use of SVM (Support Vector Machine).We can replace SVM with ANN technique for classification. Goal is to classify pictures of soil sample with high accuracy as well as cost effective. In this paper, color moment, Hsv , wavelet transform and gabor filter feature extraction methods applying to original images and extracting texture features of soil images it for classification. Results on a database of soil images belonging different types of Soils to show that proposed method performs soil classification effectively.