Simulating multimedia IoT

Abstract

While working with IOT projects with camera sensor,the cost of hardware becomes prohibitive as it is not possible to duplicate the hardware for a large number of sensors. Here we need to have lots of multimedia sensors which send image data. The image data so generated is to be compressed at the client node suitably using some transform. To make a cost effective client we simulate it as a client process in linux. Here a random number generator is used to generate an image sample. In this implementation we use a wavelet transform to compress the generated image. This is then sent over a socket to the server process. On the server side a stream socket receives the image data which is then suitably decompressed. Here a broad range of image processing algorithms are applied to enhance the image, threshold and segment it. Here different morphological operations are applied to isolate the region of interest. There could be multiple clients generated in several windows. This way a cost effective client and server is implemented to simulate the IOT architecture.