Automatic Soil Classification by using Gabor Wavelet & Support Vector Machine in Digital Image Processing

Soil is an imperative element of agriculture where there are various types of soil available for the same. Every soil has its distinct type of features and its harvesting nature for the development of agriculture in various manners. It is highly required to get familiar with this knowledge that signifies the quality of the soil and which kind of soil is required for any particular cultivation that better fulfill the desire. Here the system has been implemented using machine learning approach for recognizing the soil type and its features. In this field; a lots of progression has been done but accuracy still suffers. Machine learning is a best approach for analyzing the harvesting data as per the experience. This paper has developed a system to classify the soil at its best attempt by extracting its features and nature. The system uses Support Vector Machine (SVM) and Gabor Wavelet for recognizing soil type and classify for better recommendation. The objective of this examination is to deal with the soil pictures to deliver a high level soil classification system for rustic ranchers for minimal cost. Soil surface is the rule factor to be considered proceeding to attaining an improvement. It impacts the reap assurance and deals with the water transmission property. The normal hydrometer procedure chooses the degree of sand, slit and clay present in a soil test. Framework achieved 97.12 % of accuracy with minimal error rate.