Regularized Decompression of Kirlian Images by Optimized JPEG Dequantization Array

This paper proposes an optimized dequantization array for JPEG compression scheme. The Joint Photographers Expert Group (JPEG) algorithm is a powerful lossy compression strategy for still image. The existing JPEG block compression approach results in poor visual quality because of creating discontinuities at low bit rates. Moreover, if the decompressed image is processed to detect edges or to extract some spatial features, false edges appear. In Order to overcome these difficulties, an optimized dequantization array is proposed for the Joint Photographers Expert Group (JPEG) scheme. In this scheme, all the details retrieved from the JPEG still picture which corresponds to the JPEG standard that can be used for various types of compression and digital image processing. Instead of specifying the Huffman code values statically, they are generated dynamically from the retrieved Huffman table specification that can be used for the purpose of compression. Here, the proposed decompression scheme minimizes the reconstruction error subjected to the regularity of the decoded image blocks by using an optimized dequantization array. .The experimental result shows very good visual quality and reduces discontinuities in post processing such as edge detection.