Abstract: It has been widely used in computer vision, military investigation, medical care and some other fields. The widely used seamline removal methods include overlappingimage method, wavelet transform method and hardcorrection method. An improved method based on the correctionratio of average gray value of the images on the two sides of the seamline was proposed. This method calculated the rate of the corrected value and the former value. Then the ratio gradually changed from the seamline to the boundary so that the seamline was removed. The experiments have shown that the method proposed above is effective to eliminate seamlines, especially when the difference of the gray level is big.

Key words: seam-line removal, hard-correction method, gray average value, correction-ratio of average value, histogram matching

摘要： 目前，普遍采用的拼接缝消除方法主要有基于重叠影像的消除方法、小波变换法和强制改正法。提出一种新的基于拼接缝两侧图像灰度均值改正比的拼接缝消除方法，计算缝两侧图像均值的平均值与原图像均值之比，进而将该比值由缝向两侧逐渐过渡，最终实现拼接缝的消除。实验表明，该算法可行、有效，拼接缝消除后具有较好的视觉效果，尤其在灰度差值较大时，该算法比强制改正法具有明显的优势。

关键词: 拼接缝消除, 强制改正法, 灰度均值, 均值改正比, 直方图匹配