纹织CAD中图像的自动拼接算法

针对织物扫描获得的各个分块图像需要进行拼接的问题,结合扫描图像的数据特点,首先采用一种改进的SIFT(Scale Invariant Feature Transform)算法进行特征点的获取与匹配,并且用RANSAC(Random Sampling Consensus)算法精选匹配点对以求得转换矩阵,然后将待拼接的织物扫描图像通过转换矩阵映射到参考空间上,最后用线性过渡法来实现图像间的无缝拼接.实验结果表明:该方法不仅有效快速,而且对于织物扫描图像之间存在的平移、旋转、亮度差异以及一定程度上的形变都保持了很好的稳定性,能很好地适应纹织CAD中图像的自动拼接.

Algorithm of Automatic Image Stitching in the Jacquard CAD

Firstly,an improved SIFT(Scale Invariant Feature Transform)algorithm was used to locate the feature points in the scanned images targeting at the problem of automatically stitching of textile images which were scanned in divided pieces.The RANSAC(Random Sampling Consensus)algorithm was employed for getting the best matching points to compute the transforming matrix.Then the image was conversed to the reference coordinate system.Finally,the overlapped area linear transition method was incorporated for seamless stitching.The results showed that the algorithm was fast and stabile with respect to various sizes,luminance,and rotations,which could be suitable to the stitching of images in the jacquard CAD.