-
摘要: 针对目前图像拼接中计算量较大、实时性较差的问题，本文提出了一种图像局部特征自适应的快速尺度不变特征变换（SIFT）拼接方法。首先，对待拼接图像分块，确定图像局部块的特征类型；接着自适应采用不同的简化方法提取各局部块的特征点。然后，通过特征匹配求出变换矩阵，并结合RANSAC算法去除伪匹配对。最后，通过图像融合得到最终的拼接图像。文中使用提出的方法对3组待拼接图像进行实验。从实验结果可以看出：与标准拼接方法相比，本文改进方法的计算速度提升了30%~45%。因此，这种方法能够在保证图像拼接质量的前提下，有效提高图像拼接的效率，克服图像拼接中计算复杂度高的问题，在实际图像拼接中具有一定的应用价值。Abstract: Aiming at the massive calculation burden and poor real-time performance of the existing image stitching methods, a fast image stitching method based on fast Scale Invariant Feature Transform(SIFT) algorithm with adaptive local image feature is proposed in this paper. Firstly, the images are divided into blocks. And the feature types of thses local image blocks are determined. The feature points of the local image blocks are extracted using different simplified method adaptively. Secondly, we use feature matching to get the transform matrix and the RANSAC algorithm is applied to remove the wrong matching point pairs. Finally, the stitched image can be obtained by image blending. In this paper, three groups of to-be-stitched images are used to test the performance of the proposed method. Experimental results show that compared with the standard stitching algorithm, the calculation speed by the proposed method is increased by about 30%-45%. In conclusion, the proposed method improves the stitching efficiency and efficiently overcomes the shortcomings of heavy computation in the process of image stitching while it consistently guarantees the quality of stitched image. It has a certain application value in the actual image stitching.
-
表 1 高斯尺度空间特征
Table 1. Feature of Gaussian scale space
特征描述 差分特征 Dx,Dy,Dxx,Dyy,Dxy Hessian矩阵 λ1,λ2,Det(H), 
下载: 导出CSV
表 2 两种方法特征点提取阶段比较
Table 2. Comparison results of SIFT feature point extraction using two methods
下载: 导出CSV
-
[1] 李宏升,张健.基于小波粗糙集算法的图像拼接研究[J].液晶与显示,2014,29(2):298-303.LI H SH,ZHANG J. Image mosaic research based on wavelet and rough set algorithm[J]. Chinese J. Liquid Crystals and Displays,2014,29(2):298-303.(in Chinese) [2] 史光辉,杨威.用于图像拼接的电视摄像光学系统[J].中国光学,2014,7(4):638-643.SHI G H,YANG W. Optical system used to compose images in television photograph[J]. Chinese Optics,2014,7(4):638-643.(in Chinese) [3] 张云峰.基于DSP的实时图像拼接技术[J].液晶与显示,2013,28(6):963-967.ZHANG Y F. Real-Time image mosaic technology based on DSP[J]. Chinese J. Liquid Crystals and Displays,2013,28(6):963-967.(in Chinese) [4] 何宾,陶丹,彭勃.高实时性F-SIFT图像拼接算法[J].红外激光工程,2013,42(S2):440-444.HE B,TAO D,PENG B. High real F-SIFT image stitching algorithm[J]. Infrared and Laser Engineering,2013,42(S2):440-444.(in Chinese) [5] 李军,吴洁明.一种改进图像拼接算法的仿真研究[J].计算机仿真,2012,29(2):273-313.LI J,WU J M. Simulation study of an improved image stitching algorithm[J]. Computer Simulation,2012,29(2):273-313.(in Chinese) [6] 王新华,王晓坤.十亿像素瞬态成像系统实时图像拼接[J].中国光学,2015,8(5):785-792.WANG X H,WANG X K. Real time image mosaic of the transient gigapixel imaging system[J]. Chinese Optics,2015,8(5):785-792.(in Chinese) [7] 王新华,黄玮,欧阳继红.多探测器拼接成像系统实时图像配准[J].中国光学,2015,8(2):211-219.WANG X H,HUANG W,OUYANG J H. Real-time image registration of the multi-detectors mosaic imaging system[J]. Chinese Optics,2015,8(2):211-219.( in Chinese) [8] CHEN J W,FENG H J,PAN K C,et al.. An optimization method for registration and mosaicking of remote sensing images[J]. Optik - International J. Light and Electron Optics,2014,125(2):697-703. [9] SZELISKI R. Video mosaics for virtual environments[J]. Computer Graphics and Applications,1996,16(2):22-30. [10] PELEG S,ROUSSO B,RAV-ACHA A,et al.. Mosaicing on adaptive manifolds[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2000,22(10):1144-1154. [11] BROWN M,LOWE D G. Automatic panoramic image stitching using invariant features[J]. International J. Computer Vision,2007,74(1):59-73. [12] MAHESH,SUBRAMANYAM M V. Automatic image mosaic system using steerable Harris corner detector[J]. Machine Vision and Image Processing,2012:87-91. [13] LOWE D G. Distinctive image features from scale-invariant keypoints[J]. International J. Computer Vision,2004,60(2):91-110. [14] 许佳佳.结合Harris与SIFT算子的图像快速配准算法[J].中国光学,2015,8(4):574-581.XU J J. Fast image registration method based on Harris and SIFT algorithm[J]. Chinese Optics,2015,8(4):574-581.(in Chinese) [15] 刘立,彭复员,赵坤,等.采用简化SIFT算法实现快速图像匹配[J].红外与激光工程,2008,37(1):181-184.LIU L,PENG F Y,ZHAO K,et al.. Simplified SIFT algorithm for fast image matching[J]. Infrared and Laser Engineering,2008,37(1):181-184.( in Chinese) [16] 王睿,朱正丹.融合全局颜色信息的尺度不变特征变换[J].光学 精密工程,2015,23(1):295-301.WANG R,ZHU ZH D. SIFT matching with color invariant characteristics and global context[J]. Opt. Precision Eng.,2015,23(1):295-301.( in Chinese) [17] LUO J,GWUN O. A comparison of SIFT, PCA-SIFT and SURF[J]. International J. Image Processing(IJIP),2009,3(4):143-152. [18] ANNIS F,KARTHIK R,VAIDEHI V. Image stitching with combined moment invariants and SIFT features[J]. Procedia Computer Science,2013,19:420-427. [19] 王灿进,孙涛,陈娟.局部不变特征匹配的并行加速技术研究[J].液晶与显示,2014,29(2):266-273.WANG C J,SUN T,CHEN J. Speeding up local invariant feature matching using parallel technology[J]. Chinese J. Liquid Crystals and Displays,2014,29(2):266-273.(in Chinese) [20] 韩冬松,何昕,魏仲慧,等.采用区域特征匹配的三维弹痕自动配准[J].液晶与显示,2014,29(5):761-767.HAN D S,HE X,WEI ZH H,et al.. Automatic registration of 3-D bullet marks by matching regional feature[J]. Chinese J. Liquid Crystals and Displays,2014,29(5):761-767.(in Chinese) [21] 聂海涛,龙科慧,马军,等.采用改进尺度不变特征变换在多变背景下实现快速目标识别[J].光学 精密工程,2015,23(8):2349-2356.NIE H T,LONG K H,MA J,et al.. Fast object recognition under multiple varying background using improved SIFT method[J]. Opt. Precision Eng.,2015,23(8):2349-2356.( in Chinese) [22] HARRIS C,STEPHENS M. A combined corner and edge detector[C]. Proceedings 4th Alvey Vision Conference,1988,147-151. -
下载:
图(5) / 表(3)
计量
- 文章访问数: 2626
- HTML全文浏览量: 868
- PDF下载量: 1044
- 被引次数: 0
