摘要
为实现嫦娥三号月球车全景相机图像的无缝镶嵌,针对全景相机序列图像光照不均、月表影像的特殊性等问题,提出了一种基于加速稳健特征(SURF)算法的嫦娥三号全景相机图像全景镶嵌方法。采用SURF算法提取特征点,用SURF描述子的欧氏距离作为判定度量进行特征匹配。求取相邻两幅图像间的变换矩阵,并利用Levenberg-Marquardt非线性优化算法进行优化。采用基于线性插值的渐入渐出方法对图像重叠区域进行融合,实现月表图像的无缝镶嵌。结果显示,该方法针对嫦娥三号全景相机图像实现了快速准确的镶嵌,生成的镶嵌图满足月球车探测目标选取及路径规划的需求。
Considering the uneven illumination of Chang＇E-3 pancam images sequences and the particularity of the lunar surface image, a new panoramic mosaics method based on speed up robust features （SURF） algorithm is put forward to achieve the panoramic mosaics of Chang＇ E-3 pancam images. The feature points of images are extracted using SURF algorithm, and euclidean distance of SURF descriptor is used as the measurement to obtain the corresponding image points. The transformation matrix are calculated and optimized between adjacent images by Levenberg-Marquardt algorithm. Fusion images can be achieved by using the fade in fade out method based on linear interpolation, realizing seamless mosaic of the lunar surface image. The results show that this method can achieve mosaic of Chang＇E-3 pancam images in a rapid and accurate way, and the mosaic satisfy the demands of rover for selecting detect target and path planning.
出处
《光学学报》
EI
CAS
CSCD
北大核心
2014年第9期174-182,共9页
Acta Optica Sinica
基金
国家重大专项(探月工程)
国家自然科学基金面上项目(41371414
11273037)
国家自然科学基金青年科学基金(41304021)
