Image compression enhancement for WSN application using AHAAR wavelet transform and classification

Abstract

prolonging the lifetime of wireless sensor networks (WSNs) is an essential requirement due to limited energy storage capability of sensor node. Battery lifetime can be extended by reducing the amount of data transmitted. Thus, this paper proposed a new image compression of grayscale technique called Adaptive Haar wavelet transform theory to by providing a lossy compression. This method was introduced to overcome the drawback of the original theory by improving the compression capability. It takes into consideration the visual effect on the output image by preserving the image details. The exposure fuzzy logic classifier is utilized in this paper to improve the process of classifying the output of the compressed image into over, under or well-exposed images. Multi scale Retinex (MSR) technique was introduced to enhance the compressed classified images from over or under-expose image contrast. This work aims to increase the long lifetime of sensor by reducing the energy consumption to transfer images in WSN. A universal gray scale image database images had been applied to test the compression ratio. The output is evaluated by comparing the image size before and after compression in KB, the energy of the images before and after and also the energy consumption after the image being compressed. 81.19% energy consumption improvement in the output result of the proposed method.