Abstract: Due to no effective quality evaluation methods for bird-eye panoramic image stitching,a novel metric was proposed,which is based on the sampling distribution of the final image to the original image．According to this metric,a novel framework for image quality assessment integrated with sampling distribution was elaborated in the following section．Firstly,we briefly described a general algorithm of bird-eye panoramic image．Then,the metric based on sampling distribution was proposed．At last,other metrics,such as image variance,entropy and average gradient were compared.The comprehensive experiments demonstrate our method is effective and robust.

Key words: Panorama parking system,Image-stitching,Image quality evaluation,Pixel mapping