Research Article
JPEG Lifting Algorithm Based on Adaptive Block Compressed Sensing
Algorithm 1
JPEG-ABCS algorithm based on OMP (orthogonal matching pursuit).
(1) | Input: | | Original image I; rate control bpp; | | Subimage dimension n = 64; | (2) | Initialization: | | ; N = R × C; | | ;// | | ;// | | //step1: adaptive blocking and vectorization | (3) | for j = 1, …, T2 do | (4) | ; | (5) | | (6) | ; | (7) | ; | (8) | end for | (9) | ; | (10) | ; | (11) | ; | (12) | | (13) | if ; | (14) | if //Condition = 1 | (15) | ; | | //: horizontal scanning and vertical linking | (16) | else if //Condition = 2 | (17) | ; | | //: vertical scanning and horizontal linking | (18) | end if | (19) | else if | (20) | if //Condition = 3 | (21) | ; | | //: zigzag along the main diagonal direction | (22) | else if //Condition = 4 | (23) | ; | | //: zigzag along the counter-diagonal direction | (24) | end if | (25) | end if | | //step2: adaptive observing and bit rate control | (26) | for i = 1, …, T1 do | (27) | ; ; | (28) | ; | | //synthetic feature (J) | (29) | end for | (30) | ; | (31) | ; | (32) | ; | | //-- sampling ratio of subimages | (33) | ; | (34) | | | //prevent undersampling and oversampling | (35) | ; | (36) | ; | (37) | ; | | //: noiseless; : noise | (38) | ; | (39) | | | //step3: codec and antiquantization | (40) | | (41) | | (42) | | | //step4: reconstruction and denoising | (43) | if Condition = = 1 | (44) | | (45) | else if Condition = = 2 | (46) | | (47) | else | (48) | end if | (49) | | (50) | for i = 1, …, T1 do | (51) | ; | | //-- column vector of | (52) | ; | | //calculate optimal iterative of subimages | (53) | for do | (54) | ; | (55) | ; | (56) | ; | (57) | end for | (58) | ; | | //: reconstruction sparse representation | (59) | ; | | //: reconstruction original signal of subimages | (60) | end for | (61) | ; | | //step5: antivectorization and jointing | (62) | ; | | //: recovered image with JPEG-ABCS |
|