The application of IoT technology along with the machinery learning implementation process is a breakthrough opportunity to deal with the problems of modern agriculture. This paper presents a wide-ranging system that utilizes IoT for data capturing from sensors and machine learning models for analysis of remote sensing images. The proposed system is envisaged to enhance crop health check, yield estimation as well as the resource utilization. The IoT sensors recorded the soil moisture and temperature, and environmental conditions, and the satellite and drone images were used to produce the vegetation indices and crop stress identification. Similarly, the designed models that incorporated convolutional neural networks scored high accuracy in image-based assessments, attaining an accuracy rate of 92 percent for crop health classification. The use of the system was demonstrated to combine principles of sensorial feedback and image analysis to enhance precision farming. Experimental results show that the proposed framework can increase the yield rates in agricultural activities, minimize wastage of resources and allow timely intervention. This work also defines how to create efficient methods of smart farming so that sustainable agriculture becomes a reality and how to develop new solutions to feed the growing population in the world.