Research Article
Extreme Image Classification Algorithm Based on Multicore Dense Connection Network
Algorithm 1
Extreme image classification and restoration algorithm.
| | Inputs: OriginalImg: original image | | | Dehaze: AOD-Net | | | DeRainStreak: DID-MDN | | | DeRaindrop:Attentive GAN for raindrop | | | DeLowIllumination: HDR | | | DeBlur: DeblurGAN-v2 | | | Outputs:ProcessedImage | | (1) | i = 2 | | (2) | def ImageRestoration (OriginalImg): | | (3) | ClassNum = MDCNet (OriginalImg) | | (4) | if ClassNum = = 0: | | (5) | OutImg = OriginalImg | | (6) | elif ClassNum = = 1: | | (7) | OutImg = DeHaze (OriginalImg) | | (8) | elif ClassNum = = 2: | | (9) | OutImg = DeRainStreak (OriginalImg) | | (10) | elif ClassNum = = 3: | | (11) | OutImg = DeRaindrop (OriginalImg) | | (12) | elif ClassNum = = 4: | | (13) | OutImg = DeLowIllumination (OriginalImg) | | (14) | elif ClassNum = = 5: | | (15) | OutImg = DeBlur (OriginalImg) | | (16) | return ClassNum, OutImage | | (17) | for epoch in range (i): | | (18) | Num, Image = ImageRestoration (OriginalImg) | | (19) | if Num = = 0: | | (20) | ReImage = Image | | (21) | break | | (22) | else: | | (23) | ReImage = Image | | (24) | ProcessedImage = ReImage |
|
