An Integrated Semi-Supervised Learning Framework For Image Compression Using Dct, Huffman Encoding, and Lzw Coding

Abstract

The application of compression of the data to digital files is called image compression. The encoding of images tackles the question of reducing the volume of data needed for digital image representation. The storage space for image loading is also limited. The compression of the picture can be lossy or lossless. This text aims to use different strategies including Discrete Cosine Transformation, LZW Coding, and Huffman Encoding to implement simple JPEG compression. Digital picture data is transformed from spatial domain to frequency domain by the Digested Cosine Transformations (DCT). The compression methods used in this article do not impact data loss in the field of picture transparency. We use JPEG for this reason. JPEG is a format for still compression frames based on the Discrete Cosine Transform and is also appropriate for most compression applications. In a python framework, the project has been tested and a compressed picture was finally created. We were able to achieve 96.11%, 46.61%, and 67.91% compression in DCT, LZW, and Huffman, respectively. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.