In recent years, Internet of Things (IoT) has gained more importance in various fields due to the advancement in technologies. Generally, IoT is a technology where based on sensor data, the computer can access the data from any aspect without any human interaction. IoT concept came into the picture in order to overcome the problem like data accuracy and related parameters when data were collected through human interference. Traditionally, the radio-frequency identification (RFID) and the wireless sensor networks were considered as the major techniques used in IoT. In recent researches, while designing the network, path selection remains to be major drawback or issue which results in less efficiency for the whole system. The research can be extended with the idea of developing a technique where we combine WSN along with image processing concepts.Here we employ certain advancement in path selection where we select the clusters through which data can be transmitted. This will improve the overall efficiency of the system. Data loss and computational time can be improved if an optimal path is selected for transmission. These techniques can be employed in various IoT applications like field monitoring, quality prediction etc. The research will be carried on MATLAB / NS2 interface.