Fast Joint Image Compression-Encryption Algorithm used for 3D Reconstruction

This paper presents a JPEG-improved algorithm for compression and encryption of 2D structured light images applied in 3D reconstruction. The proposed method incorporates compression and encryption simultaneously to achieve a high compression ratio while adding security without affecting 3D reconstruction. The 2D structured light image is divided into 8×8 blocks, then DCT is applied for each block followed by an optimized JPEG-based quantization matrix. The resulting blocks are merged in a single matrix, and the DC coefficients in each column are subjected to first-order difference to reduce the Min-Max span for fast decoding. The decompression is accomplished reversely, and a fast sequential search method is proposed based on a conventional searching approach. The results showed achieving up to a 98% compression ratio of 2D structured light images while maintaining successful and accurate 3D reconstruction. Moreover, the execution time decreased significantly by (61%-80%).