Abstract
Efficient image data compression algorithms are required to minimize the cost of data transmission and storage as the quality and file sizes of images keep increasing. With the advancements in image sensors and processing units of mobile devices, use of complex but more effective compression algorithms is becoming more prevalent on a wide variety of devices. There are two main ways, lossy and lossless, to compress data. In lossy compression, while the compression gain may be substantial, the quality of the image may not be preserved. When any loss on data is not tolerable, such as in medical images, lossless compression algorithms are the only choice. The focus of our present study is on enhancing the performance of some of the existing algorithms and proposing new ones to better compression gains of lossless image compression. More specifically, we proposed two pre-processing techniques, the pseudo-distance technique (PDT) and the hierarchical coding technique. Then we compressed the pre-processed image data using the block-sorting transformations, and inversion coding technique along with an entropy coder. On various image data sets, our proposed data compression techniques performed better than GIF, PNG, and JPEG. In addition, we parallelized PDT to use in multi-processors. We also implemented some of the lossless data compression algorithms and transformations in microcontrollers (Arduino Uno, TI MSP432), and developed techniques to asses power consumption during the data compression process.