Soil pH Prediction Using Machine Learning Classifiers and Color Spaces

Abstract

The Indian economy is primarily dependent on agriculture. Successful production of crops is a need to ensure whether a particular crop will yield in a specific soil. Ph value, alkalinity, basicity directly affect the growth of the plant. Soil preparation is the most crucial process before plantation. All these factors of the soil can be determined by using the color image processing techniques. All farmers are interested in knowing how much yield can be expected. In the past days, yield prediction was performed by considering farmer’s experience for a particular field. The crop yield prediction is a major issue that is unsolved based on available data with some limitations. If the crop is not yielding correctly, that means it must have some drawbacks. The proposed pH value prediction of 40 soil images is carried out. Using these color models, pH factor of each soil image is calculated. Different classifiers are applied to each color space model, and accuracy and RMSE values are obtained. So, the system primarily focuses on predicting the appropriate pH of a soil so that the crop will be predicted by using the pH values of the soil. The soil images are processed, and the pH values are gained.