Low complexity block tree coding for hyperspectral image sensors

Curvelet Transform Based Compression Algorithm for Low Resource Hyperspectral Image Sensors

“…Bjøntegaard delta peak signal-to-noise rate (BD-PSNR), Structural Similarity (SSIM) index, and Feature Similarity (FSIM) index are unit less parameter. Te coding memory is calculated as kilobyte (kB) or megabyte (MB) and the coding complexity (encoding time and decoding time) is calculated in second (sec) [22].…”

“…Te proposed HSICA is compared with the other transformbased HSICA [16][17][18][19][20][21][22][23] [16] 3D-SPIHT [17] 3D-WBTC [18] 3D-LSK [19] 3D-NLS [20] 3D-LMBTC [21] 3D-LCBTC [22] 3D-ZM-SPECK [23] 3D-LEZSPC 3D-SPIHT [17] 3D-WBTC [18] 3D-LSK [19] 3D-NLS [20] 3D-LMBTC [21] 3D-LCBTC [22] 3D-ZM-SPECK [23] 3D-LEZSPC (f ) Figure 1: PSNR (in db) comparison of 3D-LEZSPC with 3D-SPECK [16], 3D-SPIHT [17], 3D-WBTC [18], 3D-LSK [19], 3D-NLS [20], 3D-LMBTC [21], 3D-LCBTC [22], and 3D-ZM-SPECK [23] at various bit rates for the four diferent hyperspectral images. [32,33].…”

“…Tese algorithms have low-coding complexity, bit stream embeddedness, low coding memory demand, and high-coding gain than other transform-based HSICAs. Te insignifcant coefcients are represented by the block cube (3D-SPECK, 3D-LSK, and 3D-ZM-SPECK) or tree (3D-SPIHT and 3D-NLS) or block cube tree fashion (3D-WBTC, 3D-LMBTC, and 3D-LCBTC) [16][17][18][19][20][21][22][23].…”

“…In the same way, 3D-LEZSPC and 3D-NLS exist in the range of −0.69 dB to −0.26 dB for Washington DC Mall HS image, −0.48 dB to −0.24 dB for the Yellowstone Scene 0, −0.49 dB to −0.22 dB for the Yellowstone Scene 3, −0.49 dB to −0.06 dB for the Yellowstone Scene 18, −0.51 dB to −1.88 dB for Pavia Centre, and −0.74 dB to −1.42 dB for Pavia University hyperspectral image. Bjontegaard metric calculation or 3D-SPIHT [17] 3D-WBTC [18] 3D-LSK [19] 3D-NLS [20] 3D-LMBTC [21] 3D-LCBTC [22] 3D-ZM-SPECK […”

“…Continued.Bit rate (bpppb) Computation time 3D-SPECK[16] 3D-SPIHT[17] 3D-WBTC[18] 3D-LSK[19] 3D-NLS[20] 3D-LMBTC[21] 3D-LCBTC[22] 3D-ZM-SPECK[23] …”