基于加权融合的水下图像增强算法及FPGA设计  

Design of Underwater Image Enhancement Algorithm and FPGA Based on Weighted Fusion

作　　者：黄晓鑫 朱铮涛 李敬韬 HUANG Xiaoxin;ZHU Zhengtao;LI Jingtao(Institute of Mechanical and Electronic Engineering,Guangdong University of Technology,Guangdong 510006,China.)

机构地区：[1]广东工业大学机电工程学院,广东510006

出　　处：《电子技术(上海)》2025年第1期106-110,共5页Electronic Technology

摘　　要：阐述针对水下机器人在采集图像时存在的实时性差、图像低对比度和色偏问题,设计一种基于加权融合的水下图像增强算法,并完成了现场可编程门阵列(FPGA)映射。通过在HSV空间设计权重,并对图像分别进行直方图均衡化和灰度世界法处理,对获得的融合子图进行加权融合,最终将算法实现FPGA映射。实验结果表明,改进的加权融合算法优于多种现有图像增强算法,对比融合子算法的UIQM指标分别提升了29.5%和46.6%。FPGA采用流水线设计,处理一帧640×480图像的总耗时为12.3mS,理论帧率为81.3fps,功耗为1.08W,能够满足水下图像实时增强的需求。This paper describes the of an underwater image enhancement algorithm based on weighted fusion in view of the problems of poor real-time performance,low contrast and color bias,which completed the field programmable gate array(Field Programmable Gate Array,FPGA)mapping.By designing the weights in HsV space and balancing the histogram and the gray world method,the obtained fusion subgraphs are weighted and fused,and finally all the algorithms realize FPGA mapping.The experimental results show that the improved weighted fusion algorithm is better than many existing image enhancement algorithms,and the UiQM index of the comparative fusion subalgorithm improves by 29.5%and 46.6%respectively.FPGA adopts pipeline design,and the total time consumption of processing one frame 640×480 image algorithm is 12.3 ms,the theoretical frame rate is 81.3 fps,and the power consumption is 1.08w,to meet the requirements of real-time enhancement of water image.

关 键 词：计算机工程 加权融合 图像增强 FPGA UIQM 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TP183[自动化与计算机技术—计算机科学与技术]