Design of Real-time Haze Image Restoration System Based on FPGA Technology

摘要

With the development of science and technology, the outdoor image acquisition system has been widely used in various sectors of the national economy, such as transport, agriculture and aerospace. But by the impact of inclement weather, the image quality will be a serious decline. Therefore, the defog processing of digitized images is one of the research focus in recent years. The traditional approach is to upload the image data to the host computer and then process the image on software platform. Since it is difficult to achieve real-time processing by this method, portability and mobility of the platform is too weak, we present a FPGA-based image defogging system. The system includes algorithm and hardware platform. The algorithm is based on fog degradation model and has been optimized and streamlined for FPGA platform. The hardware part using D5M provided by Terasic to acquire the image, using DE2-70 provided by Altera to process the image and using LTM provided by Terasic to display the result. Experimental results show that this system has achieved real-time processing and has also met the design requirements of de-fog effects.