Fractal video sequences coding with region-based functionality

Fast Quadtree Based Normalized Cross Correlation Method for Fractal Video Compression using FFT

“…In Table 2 even though the PSNR of the proposed algorithm for 'Paris' video is lower than that for the region based algorithm [20] but SSIM is the second largest among all videos. In this particular video sequence entire frame is moving in the left direction from 6 to 7 pixels, so almost all the range blocks are modified structurally with s and o parameters.…”

“…The average PSNR for 'Bus' video sequence is low, whereas the SSIM index is better. However, for the remaining videos CR and PSNR using the proposed method are better than region based [20] method. It is also observed that compression is very high for low bit rate videos with good quality of output video.…”

“…The result shows that the proposed method improves the compression ratio (CR) 2.5-9 times, PSNR 4 -8 dB more and encoding time 16-34 times in comparison with traditional CPM/NCIM. Similarly, CR is increased by 6-28, The proposed algorithm is also compared with Zhu's regionbased algorithm [20] in terms of compression ratio and PSNR of the first 15 frames of each video sequence as shown in Table 2.…”

“…NHEXS employs halfway stop technique and modified partial distortion criterion (MPDC) to minimize encoding time. Video sequences are also encoded by region-based approach [20] using NHEXS. The regions can be defined according to the previously computed segmentation map and are encoded independently of each other.…”

A Gradient Vector Flow Snake Model using Novel Coefficients Setting for Infrared Image Segmentation

“…The active contour model can be classified into the parameter-based type [8][9][10][11][12][13][14][15][16][17][18][19][20] and the geometrybased type [21][22][23][24][25]. This paper focuses on the former type and proposes a novel model to segment infrared images more accurately.…”