High Resolution Video Stitching Algorithm Based On FPGA

Video stitching technology has become an important part of many fields including computer vision, safety and surveillance and battlefield information perceiving. The video stitching quality is one of the key determinants to research on object detection and recognition, analysis and decision-making. Due to the factors such as large volume of data, complexity and real-time requirements, research on high-resolution video stitching has been faced with many challenges, which have restricted the development and application of video stitching technology. FPGA(Field Programmable Gate Array) is a suitable platform for video stitching. This paper has proposed a high-resolution video stitching method based on FPGA. The research includes the following aspects:1. An advanced RANSAC algorithm based on key point dispersion and pre-selection is proposed, which can disperse the key points and filter the errors based on the information obtained from the key points.2. A look-up table method for video registration is proposed to improve the real-time performance of video registration, which can reduce the computing complexity.3. A novel FPGA-oriented seam-searching algorithm based on advanced dynamic programming is designed to improve the real-time performance and accuracy of seam searching. The parameters of this algorithm are also analyzed.4. In order to deal with the moving objects in overlapping areas, a FPGA-oriented local update algorithm of seam based on ROD( Region of Difference) is proposed to make detour of moving objects and improve the quality of video fusion.5. A FPGA framework of high resolution video stitching is designed and the implementation of different modules is discussed.The effectiveness of the proposed methods has been verified by the comparison experiments with conventional algorithms.