Conventional snow melting system used in the current winter season is operated by previously set temperature and humidity conditions but this snow melting system is inefficient. Because that even though the humidity and temperature condition is satisfied with the set conditions but sometimes actual snowfall does not occur on the road. As a result, unnecessary power consumption is generated. also, conventional snowmelting system dosen't analysis snowfall condition on the road surface. In this paper, optimal snow melting system is suggested to improve the conventional system’s inefficient problem using IoT technology. To solve this problem, presented snow melting system apply wind speed data with arduino micro controller and snow fall image analysis methods with raspberry pi micro controller for setting the optimum determination condition and then operating reasonable power consumption system. The proposed system obtain road surface image and subtract between background image and road surface image to get the difference image using opencv technique. From this image information, we can determine snow falling condition of a road and analyzes difference image histogram to calculate snowfall surface area and distribution map. Therefore the suggested system can control and handle snow melting system more efficiently and optimally. This paper proposes new optimal snow melting system based on the IoT technology and image processing with good results.