计算机科学
现场可编程门阵列
动态范围
计算机视觉
智能摄像头
人工智能
图像传感器
高动态范围
时滞与积分
宽动态范围
图像处理
高动态范围成像
CMOS芯片
帧速率
计算机硬件
作者
Yi-Zheng Lang,Yun-Sheng Qian,Hong-Gang Wang,Xiang-Yu Kong,Sheng Wu
标识
DOI:10.1007/s11082-022-03679-8
摘要
To capture more details and enhance image’s contrast in a low-light-level (LLL) environment and improve the performance of the LLL night vision system, a method of extending the dynamic range for the LLL night vision system based on guided filter and adaptive detail enhancement is proposed. Moreover, a complete high dynamic range (HDR) intensified complementary metal oxide semiconductor (ICMOS) camera based on this method is implemented. Unlike conventional technologies to create HDR images, the proposed camera based on an HDR CMOS can better process images’ edge information while avoiding artifacts’ appearance and solve the storage problem in a multi-images fusion so that real-time image is realized processing. Based on this, the ICMOS camera can capture details in a high contrast environment with an extended dynamic range of up to 80 dB. As the core processing unit of this camera, a Xilinx Spartan 6 series FPGA has completed the work of CMOS driving, image detail enhancement, image storage control, video compression output. Finally, this camera can realize an output video with a resolution of 1280 × 1024 at the rate of 60fps, and experimental results show that this camera has a good performance under a 10−4 lx environment.
科研通智能强力驱动
Strongly Powered by AbleSci AI