Abstract
In order to combine multi-views into one ultra-wide single-view for present surveillance desire, a video fusion scheme based on FPGA is proposed in this paper. The geometric transformation and image fusion algorithms used in our scheme are simulated by Matlab for verification and optimization. Then, a combination of two 640*480 video images captured by cameras is implemented in a Altera FPGA platform DE2-115. Moreover, the weighted fusion method is used to realize video fusion for better performance; and larger size of 1280*768@60Hz video images' combination are also realized in DE2-115. Experiments show that the proposed approach can realize the mosaic of two video images with a delay of 16 clock cycles, i.e. a total of 0.16us. The error induced by the fusion process is less than 1 pixel. The implementation in DE2-115 for two RGB 1280*768@60 pixel size image combination can be 1743 times faster than software implementation with a good splicing performance.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
