论文部分内容阅读
针对车载鱼眼图像拼接过程中的“拼接缝”的问题,对拼接误差的来源及其理论模型进行了分析。通过对全景影像的拼接过程进行梳理,对拼接误差的来源进行了论证,并对误差的理论模型进行了详细推导。最后,结合实际车载鱼眼数据,通过在不同目标距离情况下进行全景图拼接实验,证实了实际拼接成像效果与根据理论模型推导所得的拼接缝的大小基本吻合,说明了该文所提出的理论模型的可靠性,从而为后续改善全景图像拼接效果提供了理论支持。
In order to solve the problem of “splicing seam” in the process of car fisheye image splicing, the origin of the splicing error and its theoretical model are analyzed. By combing the splicing process of panoramic images, the source of splicing error is demonstrated, and the theoretical model of error is deduced in detail. Finally, based on the actual data of fisheye in the vehicle, the panoramic mosaic experiment is conducted under different target distances, which confirms that the actual stitching effect is in good agreement with the size of the stitching seam deduced from the theoretical model. The reliability of the theoretical model, which provides theoretical support for the subsequent improvement of the panoramic image mosaic effect.