A New Singular Vector Sparse Representation Technique for Crop Image Compression

Nowadays, the application of crop images for sharing crop information is perpetually increasing. As a result, image datasets need more storage space and channel bandwidth, leading to higher costs. Therefore, reducing image data size is essential. This article, therefore, introduces a compression method based on the discrete wavelet transform (DWT) and the modified singular vector sparse reconstruction (MSVSR) approaches. It gives good reconstruction quality and compression characteristics. In the first stage, input images are decomposed using DWT into frequency subbands. In addition, a modified sparse representation of singular vectors based on the singular value decomposition (SVD) approach is applied in detailed subbands to improve the compression efficiency. At the reconstruction stage, piecewise linear interpolation (PLI) and inverse DWT are used to retrieve a good-quality image. The performance of the proposed method has been evaluated based on various fidelity parameters, including bit-per-pixel (BPP), peak signal-to-noise ratio (PSNR), mean square error, and structural-similarity index. Moreover, the experimental results illustrate that the proposed DWT-MSVSR technique with Daubechies 4 wavelet has achieved significantly higher compression (67.27%), and structural similarity index measure (SSIM) (36.27%), as compared with SVSR with similar image quality, as well as other SVD-based existing methods. From the evaluated results, it is observed that this method has proven to be efficient in compressing different types of crop images with acceptable reconstruction quality.