基于控制点配准算法的无人机遥感影像自动无缝拼接技术

传统无人机遥感影像自动无缝拼接技术无法精准匹配影像信息，导致无人机遥感影像拼接结果出现大量缝隙，拼接效果差。因此提出了基于控制点配准算法的无人机遥感影像自动无缝拼接技术。遵循无人机影像成像原理，获取无人机遥感影像，并将数据以图像格式文件形式存储。设置阈值，剔除最邻近域和次邻域比值大于阈值的控制点，对影像坐标平移和缩放数据标准化处理，彻底消除坐标变换对图像配准影响。构建相似变换矩阵，获取新的控制点集，使用直接线性变换算法预估变换矩阵，得到线性解。经过粗、细配准，确定不同图像重叠区域。搜索最佳拼接线，使用加权平均融合法消除拼接缝，由此设计拼接流程。由实验结果可知，该技术能够精准匹配影像信息，检测到影像最大分辨率为1000*800，具有良好拼接效果。

Automatic seamless stitching technology of UAV remote sensing images based on control point registration algorithm

Traditional UAV remote sensing image automatic seamless stitching technology cannot accurately match the image information, resulting in a large number of gaps in the splicing result of the UAV remote sensing image, and the splicing effect is poor. Therefore, an automatic seamless stitching technology of UAV remote sensing images based on control point registration algorithm is proposed. Observe the drone image imaging principle, obtain the drone remote sensing image, and store the data in the form of image format file. Set the threshold, remove the control points whose ratio of the nearest neighbor and the next neighbor is greater than the threshold, standardize the image coordinate translation and zoom data, and completely eliminate the effect of coordinate transformation on image registration. Construct a similar transformation matrix, obtain a new set of control points, use a direct linear transformation algorithm to predict the transformation matrix, and obtain a linear solution. After coarse and fine registration, the overlapping areas of different images are determined. Search for the best stitching line and use the weighted average fusion method to eliminate the stitching seam, thus designing the stitching process. It can be seen from the experimental results that the technology can accurately match the image information, and the maximum resolution of the image is detected to be 1000*800, which has a good stitching effect.