Image enhancement (IE) is a common thing we use to get better results from previous imagery. This image enhancement is not only used by us, but it is implemented in many fields. Such as implementation in the military ...
详细信息
Image enhancement (IE) is a common thing we use to get better results from previous imagery. This image enhancement is not only used by us, but it is implemented in many fields. Such as implementation in the military field, medical field, legal field, industry field, entertainment field, and much more. The main use of IE in each field is to get clear information. Pedestrian detection is an essential way of support in current traffic management. Traditional pedestrian detection error & miss detection rates are high owing to irregular lighting, dim tunnel atmosphere, and blurred controlled picture, making subsequent identifying hard. Arapid image enhancement (FIE) algorithm founded on picture model restriction is therefore suggested in this document and reduced to the pedestrian region of interest (ROI) in the pavement close the road under highway tunnel (HT) scene. First, the technique used to assess the local atmospheric light (LAL) by combining global atmospheric light (GAL) and partitioned atmospheric light (AL). Second, the transmission is predicted to be founded on the plan obtained as of the image model's constraints. The third is for balancing tunnel illumination, the technique utilizes steady instead of illumination. Lastly, the picture of the tunnel is improved by the picture model. Moreover, we propose a narrowing region approach for improving the overall computing performance, due to the real-time requirements of the algorithm. Taking account of the highway tunnel features, which are a blurred scene and difficult to identify from the context, we use a multi-function integration approach to detect the enhanced image. We described a novel filter in this article that is commonly used in computer vision & graphics. Guided algorithm filter is MATLAB simulated. Results of the experimental and comparative assessment indicate that the suggested technique can quickly and efficiently enhance the picture of the tunnel and highly enhance the impact of pedestrian detec
Pedestrian detection is a necessary means of support in modern traffic management. The error and miss detection rate of traditional pedestrian detection are high due to uneven illumination, dim environment in the tunn...
详细信息
Pedestrian detection is a necessary means of support in modern traffic management. The error and miss detection rate of traditional pedestrian detection are high due to uneven illumination, dim environment in the tunnel, and the blurred monitored image, which makes it difficult for the subsequent identification. Therefore, in this paper, a fast image enhancement algorithm based on imagingmodelconstraint is proposed and narrowed to the pedestrian ROI in the pavement near the street under the scene of highway tunnel. First, the method uses the combination of global atmospheric light and partitioned atmospheric light to estimate the local atmospheric light. Second, transmission is estimated based on the formula derived from the imagingmodelconstraints. Third, the method uses constant instead of illumination to balance tunnel image illumination. Last, the tunnel image is enhanced according to the imagingmodel. Furthermore, because of the algorithm's real-time requirement, we propose a narrowing region method to thoroughly improve the overall computing efficiency. Considering about the characteristics of high way tunnel, which is a blurred scene and has difficulty recognizing the foreground from the background, we adopt a method of multi-feature integration to detect the enhanced image. Experimental and comparative analysis results show that the proposed method can rapidly and effectively enhance the tunnel image, and improve the effect of pedestrian detection in high way.
暂无评论