Hello, my dehazing results obtained with the netG_epoch_8 test are not ideal, and even obvious problems have arisen. Then I analyzed the atmospheric light value A according to the output, mainly because the predicted A is generally too small, only 0.05. So, the final output graph is too bright, what is the cause of this?
Hello, my dehazing results obtained with the netG_epoch_8 test are not ideal, and even obvious problems have arisen. Then I analyzed the atmospheric light value A according to the output, mainly because the predicted A is generally too small, only 0.05. So, the final output graph is too bright, what is the cause of this?