摘要
图像去雾算法被广泛的应用于自动驾驶、安防监控、国防军事等重要领域。但在PC端实现去雾算法处理具有体积大和计算慢等缺点,难以满足高实时性的应用场景需求。因此该文提出了改进暗通道先验去雾算法并将其移植到ZYNQ小型化硬件平台上,利用该平台PL端强大的并行运算能力加速实现算法对图像的处理,利用PS端控制各模块的初始化与配置PL端各IP核,实现对整个系统的控制。经测试该系统具有出色的图像去雾处理能力,且具有便携性、高实时性、高计算能力、低功耗等特点,适用于对实时性要求较高的图像去雾处理场景需求。
The image dehazing algorithm is widely used in important fields such as autonomous driving,security monitoring,national defense and military.However,implementing dehazing algorithms on the PC side has disadvantages such as large size and slow computation,making it difficult to meet the high real-time application requirements.Therefore,this article proposes an improved dark channel prior dehazing algorithm and ports it to the ZYNQ miniaturized hardware platform.The powerful parallel computing capability of the PL end of the platform is utilized to accelerate the algorithm’s image processing.The PS end is used to control the initialization of each module and configure each IP core of the PL end,achieving control of the entire system.After testing,the system has excellent image dehazing processing capabilities and features portability,high real-time performance,high computing power,low power consumption,etc.It is suitable for image dehazing processing scenarios that require high real-time performance.
作者
王毅鹏
雷千龙
侯硕
侯义
WANG Yipeng;LEI Qianlong;HOU Shuo;HOU Yi(School of Electronics and Information,Xi′an Polytechnic University,Xi′an Shaanxi 710600,China)
出处
《长江信息通信》
2025年第2期49-52,共4页
Changjiang Information & Communications
关键词
ZYNQ
改进暗通道先验法
视频流去雾
硬件加速
软硬件协同
ZYNQ
Improved dark channel prior method
Video stream defogging
Hardware acceler-ation
Software and hardware collaboration
作者简介
王毅鹏(2000-),男,陕西安康人,硕士研究生,主要研究方向:图像处理算法。
