Study On Real Time Offsetting Digital Image Rotation Based On FPGA

The observation images are often rotated since the detector's rotation is relative to the hose machine on target identification and so on. Thus it is necessary to do the real time image rotation in order to eliminate the influence to calibration and observation of image rotation. It studies on offsetting image rotation from electronic method after analysis and compares with offsetting image rotation technique at present. And it also gives a complete structure of the system and the corresponding algorithm design.Based on deep researching of the principle of electronic image rotation, a optimize fast method of image rotation is developed, then simulates using Visual C++ 6.0 and guarantees the image quality after interpolation processing. Based on this, build a digital image real time rotation system in this paper which takes ACEX EP1K100 as the core, and realize rotation method using the VHDL hardware description language. Image transmission using the high speed camera and Camera Link connection has raised system run rate.Finally, we carry on the hybrid simulation experiment using QuartusII and Matlab to the entire method design. Experimental result proves that this system can fulfill offsetting rotation of gray image which collect successfully, the revolving image clear and stable, the maximum pixel error is less than one pixel. Has a time delay but less than 20ms for video signal, reaches the system parameters requirement.