SIFT image stitching based on geometric image registration solution

被引：0
|
作者：
Zou C. [1 ,2 ]
Hou X. [1 ,2 ]
Ma J. [1 ,2 ]
机构：
[1] Department of Computer Science and Technology, Wuhan University of Technology, Wuhan
[2] Hubei Key Laboratory of Technology of Transportation of Things, Wuhan University of Technology, Wuhan
来源：
| 2016年 / Huazhong University of Science and Technology卷 / 44期
关键词：
Image processing; Image stitching; Phase correlation; Scale-invariant feature transform; Transformation matrix; Voronoi diagram;
D O I：
10.13245/j.hust.160407
中图分类号：
学科分类号：
摘要：
There is a problem that large amount of calculation in the calculation of characteristic vector and feature point matching with the scale-invariant feature transform (SIFT) algorithm. Based on the SIFT algorithm and the above problem, corresponding improvement was achieved. Firstly, overlap area of images were localized by the phase correlation method roughly, and then the feature points of interest of overlap were extracted, with which the Voronoi diagram was structured. The image was divided into 4 rows and 4 columns. In every small area according to the Voronoi, 4 match points were found respectively. Transformation matrix was calculated by the corresponding image. Combining with eight transform matrix, the image transformation relations were calculated. Finally, the fade in fade out image fusion algorithm was used to stitch images. Looking for quantitative point within a specific area can reduce the number of the points need to pair, thereby increasing the efficiency of image stitching. © 2016, Huazhong University of Science and Technology. All right reserved.
引用
收藏
页码：32 / 36
页数：4
相关论文
共 7 条
  • [1] Lowe D.G., Distinctive image features from scale-invariant keypoints, International Journal of Computer Vision, 60, 2, pp. 91-110, (2004)
  • [2] Laraqui A., Saaidi A., Jarrar A., Et al., Image stitching based on the geometric solution, Proc of 2014 Third IEEE International Colloquium on Information Science and Technology, pp. 340-344, (2014)
  • [3] Ke Y., Sukthankar R., PCA-SIFT: a more distinctive representation for local image descriptors, Proc of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, pp. 506-513, (2004)
  • [4] Cheng C., Ma L., Yu Y., Et al., A novel image mosaic method based on improved SIFT algorithm and contourlet transform, Proc of 2013 32nd Chinese Control Conference, pp. 3657-3661, (2013)
  • [5] Wang Z., Yan F., Zheng Y., An adaptive uniform distribution surf for image stitching, Proc of 2013 6th International Congress on Image and Signal Processing, 2, pp. 888-892, (2013)
  • [6] Reddy B.S., Chatterji B.N., An FFT-based technique for translation, rotation, and scale-invariant image registration, IEEE Transactions on Image Processing, 5, 8, pp. 1266-1271, (1996)
  • [7] Bohler C., Cheilaris P., Klein R., Et al., On the complexity of higher order abstract Voronoi diagrams, Lecture Notes in Computer Science, 48, pp. 539-551, (2013)
1