Deep learning-based image dehazing methods have made great progress, but there are still many problems such as inaccurate model parameter estimation and preserving spatial information in the U-Net-based architecture. To address these problems, we propose an image dehazing network based on the high-resolution network, called DeHRNet. The high-resolution network originally used for human pose estimation. In this paper, we make a simple yet effective modification to the network and apply it to image dehazing. We add a new stage to the original network to make it better for image dehazing. The newly added stage collects the feature map representations of all branches of the network by up-sampling to enhance the high-resolution representations instead of only taking the feature maps of the high-resolution branches, which makes the restored clean images more natural. The final experimental results show that DeHRNet achieves superior performance over existing dehazing methods in synthesized and natural hazy images.

No abstract

Abstract. In this paper, the author designed an area-scan CCD reflecting mirror splicing image-forming system with area-scan CCD ICX415AL as its transducer module and this system can be used for tracking dynamic targets. By analyzing the theory of vignetting generating, the author made mathematic model of vignetting and confirmed the splicing and overlapping pixel number of the optical system. What's more, the sequential circuit and driving power circuit of ICX415ALwas designed and the correlated noise in video signals was strained with CDS technology. Therefore, the signal-to-noise ratio (SNR) of the system was elevated. With FPGA as its core controlling module, this system postponed the splicing image-forming system to a period during which a line of CCD data are read, thus the need of real-time tracking was completely met.