基于圆心定位的全景图像拼接与展开

针对特殊的全景采集镜头，经拼接及展开运算，得到人眼容易辨识的图像。系统镜头由两片像素阵列为4096×3072的CCD组成，拼接所形成的传感区域包含像素阵列为4096×4096，文中提出一种基于圆心定位的拼接算法，此算法通过采集到的两幅全景图像的圆心，以圆心为基准进行剪切定位，从而实现图像的拼接。并应用改进的分块查表法，将目标图像变为环带块，对拼接后的图像进行展开运算。仿真实验表明，本拼接算法可完成特殊镜头所采集的图像拼接工作、速度快、可靠性能强、符合实际应用要求。

Panorama Stitching and Spread Based on Central Location

This article aims mainly at the special panoramic acquisition lens.By splicing and starting operation,the recognizable image for humans is obtained.The system lens is composed of two pieces of pixel array CCD of 4 096×3 072.The mosaic formed by the sensing region contains a 4096 x 4096 pixel array.This paper presents a stitching algorithm method based on the central location.By searching the two collected panoramic image circles,with the center of the circle as a benchmark for the shearing and location,this algorithm achieves an accurate image stitching.By improvement and application of the block look-up table method,the target image is transformed into a ring block.Operation is started on the spliced images.Simulation results show that the algorithm can stitch images collected by a special lens with fast speed and reliable performance.It meets the requirements of practical applications.