A novel algorithm for soil image segmentation using color and region based system

Abstract

In the recent research era data mining is a very essential research domain. The data mining techniques are used to extract significant knowledge in agriculture management. These techniques are time consuming and less expensive than the statistical techniques. Many Researchers develop efficient techniques to improve the productivity of agriculture. This paper developed a new segment method to segment the soil region from other information. This research introduced Color and Region Based segment method to separate the soil region from its background. To evaluate the proposed segmentation the five metrics are used dice coefficient, jaccard index, Sensitivity, Specificity and Precision. The new approach produced 98% accuracy, 98% Sensitivity and 98% Specificity.