Adaptive Image Compressive Sensing Using Texture Contrast

Abstract

The traditional image Compressive Sensing (CS) conducts block-wise sampling with the same sampling rate. However, some blocking artifacts often occur due to the varying block sparsity, leading to a low rate-distortion performance. To suppress these blocking artifacts, we propose to adaptively sample each block according to texture features in this paper. With the maximum gradient in 8-connected region of each pixel, we measure the texture variation of each pixel and then compute the texture contrast of each block. According to the distribution of texture contrast, we adaptively set the sampling rate of each block and finally build an image reconstruction model using these block texture contrasts. Experimental results show that our adaptive sampling scheme improves the rate-distortion performance of image CS compared with the existing adaptive schemes and the reconstructed images by our method achieve better visual quality.

Fuzzy Adaptive-Sampling Block Compressed Sensing for Wireless Multimedia Sensor Networks

“…Recently, many ABCS algorithms have been proposed [9,13,14,16,[21][22][23] to address the various issues that arise with BCS. These ABCS algorithms assign a different sampling to each block based on its features (e.g., textures, edges, standard deviation, or saliency).…”

“…Traditional image compression techniques (e.g., JPEG, JPEG2000, and SPIHT), which are based on the Nyquist sampling theorem, can substantially reduce the size of an image while ensuring satisfactory image quality; however, these methods are unsuitable for implementation in single SNs because they are complicated and easily affected by channel errors during transmission. For instance, losing even a few bits during image transmission can threaten the success of the reconstruction process [9].…”

“…To mitigate this problem, compressed sensing (CS) was recently presented. Developed in 2004, CS has recently gained considerable attention from researchers because it can be used to compress multimedia data such as images and videos effectively [9][10][11][12][13][14][15][16][17]. Moreover, in the fields of data compression and communication, CS is one of the best theories due to its performance and nonadaptive coding, and its encoding and decoding operations are independent [18].…”

“…Therefore, adaptive block compressed sensing (ABCS) algorithms have been proposed [9,14,16,21,22] to address the issues of BCS. These ABCS algorithms adaptively allocate the sampling to each block based on an image feature, i.e., either saliency, standard deviation, edge, or texture.…”

“…These ABCS algorithms adaptively allocate the sampling to each block based on an image feature, i.e., either saliency, standard deviation, edge, or texture. Sun et al [9] exploited texture features to adaptively sample each block in their ABCS algorithm. The texture contrast in each block was computed by measuring the texture variations of each pixel in the block.…”