病理切片远程会诊中图像拼接技术的研究

研究图像拼接技术优化问题,是当前图像处理领域的研究热点。由于技术原因,在拼接时会造成误配。目前有利用全局优化的拼接方法进行修正的,理论上比较合理,但在速度和效果方面,尤其是精度要求比较高的场合还不理想。为解决上述问题,提出一种新方法,根据图片拼接时位置之间的相关性进行修正。提出拼接参数的相关公式,分析并提出修正简化规律。最后对十六张病理切片拼接图中误配部分进行修正。实验证明,算法可以将误差控制在一个像素以内,取得良好的效果。证明方法适用于对拼接精度要求高的应用领域,有较高的实用价值。

Study on Pathology Image Mosaic in Teleconsultation

Image Mosaicing technology has become a popular research topic in the field of image processing.But due to many reasons,the mosaicing errors of mismatch are unavoidable.The errors are usually resolved with the approach of global optimization,but the speed and efficiency are unsatisfactory.A new method was proposed for correcting the mosaiced image.According to the relevance of the positions of the images in the image stitching process,the correction formula of stitching parameters was proposed.Based on the method,the simplified correction rules were put forward.The method was tested with sixteen pictures,four by four.The results of correction present that the mismatch errors are restrained at 1 pixel.It is a high-precision image mosaicing method,with high practical value.