Modified JPEG Image Compression With Region-dependent Quantization

In many useful applications, such as in the simulations industry and the biomedical field, certain regions and/or objects in an image are of greater interest to the user than others such as the terrain or the background information. This paper presents, for these types of applications, a region-based variable quantization scheme where the quantization granularity in different preselected regions of the image is varied at the discretion of the user. The techniques developed in this work are compatible with the popular JPEG Still Image Standard for compression of continuous-tone grey-scale and color images. Further, region selection techniques and algorithms that complement variable quantization techniques are presented. We introduce three masks: step, linear interpolated, and raised cosine interpolated, that control the transition in the quantization granularity between regions of different compression ratios in an image. The paper also includes a detailed discussion of simulation results using the proposed approach. When the techniques suggested here are used, the subjective visual quality considerably improves for comparable compression ratios of standard JPEG implementation, or equivalently, a higher compression ratio is achieved for the same subjective visual quality.