[1]刘华军,张瑞珏,刘建锋,等.基于FPGA的高分辨率视频图像实时增强去雾系统[J].郑州大学学报(工学版),2020,41(02):22-27.
点击复制

基于FPGA的高分辨率视频图像实时增强去雾系统()
分享到:

《郑州大学学报(工学版)》[ISSN:1671-6833/CN:41-1339/T]

卷:
41
期数:
2020年02期
页码:
22-27
栏目:
出版日期:
2020-05-31

文章信息/Info

Title:
High Resolution Video Image Real-time Enhancement System Based on FPGA
作者:
刘华军张瑞珏刘建锋王盛夏巧桥
文献标志码:
A
摘要:
已有去雾系统大多针对单幅图像进行处理,稳定性和实时性不足,且传统图像去雾算法不仅复杂度较高还存在亮光区域直方图过曝的缺陷,难以应用于实时视频图像修复中。针对该问题,本文设计并构建了一种基于FPGA的1080P全高清视频图像实时增强去雾系统,首先在RGB颜色空间采用分块分通道进行直方图统计,然后通过预设阈值对多通道合并后的直方图进行削减,限制对比度过度放大,最后进行双线性插值计算。对改进的限制对比度自适应直方图均衡算法(Contrast Limited Adaptive Histogram Equalization,CLAHE)进行仿真验证并对系统进行有雾视频、图像以及户外场景进行去雾测试。实验结果表明,本系统能实时处理每秒30帧1920x1080分辨率的图像,与主流方法相比,去雾效果更好更稳定,系统功耗低于5瓦,可应用于低功耗领域。
Abstract:
Most defogging systems mainly focused on single image , lacking in stability and real-time perform-ance. Besides, the traditional image haze removal algorithm had high complexity and the histogram of brightarea was overexposed, it was difficult to be applied to real-time video image restoration. Facing this problems ,a 1080P full-hd video image real-time enhanced defogging system based on FPGA was designed and co nstructed. At first , the histogram statistics was conducted in RGB color space by using blocks and dividing the chan-nels. And then, the combined histogram of multi-channel was clips at a predetermined value, thereby limitingthe contrast amplification. Finally, the bilinear interpolation is performed between the new histogram. We sim-ulated the improved CLAHE algorithm, and tested the system for foggy video, image and outdoor scenes. Ex-perimental results showed that the system can real-time process 30 frames per second 1 920x 1080 resolutionimages, compared with the mainstream methods, the defogging effect was better and more stable. The systempower was less than 5 watts, so it could be applied to low power consumption.
更新日期/Last Update: 2020-05-30