Agricultural water use needs to be optimized, especially in regions where the water supply is limited. In this study, an innovative approach to soil surface texture classification and cocopeat recommendation is presented through the usage of RGB images, which can efficiently provide a cost-effective alternative in contrast to conventional methods. Cocopeat, a product made from coconut husks, is being increasingly used in agriculture as a sustainable alternative for a range of plant species because it has superior water-retention properties. How much cocopeat is necessary to maintain just the right amounts of water for different soil in a variety of environmental conditions is still hard to determine. Our results show how RGB imagery successfully captures the textural properties of special interest in accurate classification and customization of cocopeat recommendations to enhance crop production and soil fertility. It entails scalability, non-invasive sampling, and rapid data collection by the inherent use of drone or satellite imagery for this large-scale application.