基于F直方图的形状全方向顺序特征描述
|
摘要
为提高二维对象的形状描述识别精度,本文提出基于F直方图的形状全方向顺序特征描述方法。该方法在全方向上通过方向线对形状进行分割,计算各方向线上对象分割段之间的顺序特征描述量,构建形状的全方向顺序特征描述量,计算形状间的相似度。通过地物形状数据集、MPEG-7 CE-1 Part B形状数据集和瑞典叶子数据集对本方法进行了检索识别性能测试,并和其他形状相似性描述方法进行了对比。实验结果表明,本方法具有更高的检索准确率。形状全方向顺序特征描述方法将形状的边界和区域信息融合,可以准确描述形状的全局特征与局部特征,具有平移、旋转、尺度不变性,具有较强的形状描述识别能力。
To improve the shape description and identification accuracy of 2 D objects,an all-directional sequence feature description based on F-histogram is proposed. In this method,the shape is segmented omnidirectionally by the direction lines,the sequential feature description quantity among object segments in the direction line is computed,the omnidirectional sequence feature description quantity of the shape is established,and the similarity among shapes is calculated. The ground object shape data set,MPEG-7 CE-1 Part B shape data set,and Swedish leaf database are used to test the retrieval and identification performance of this method. The method is then compared with other description methods on shape similarity. Experimental results show that this method exhibited high retrieval accuracy. The all-directional sequence feature description method integrates the boundary and regional information of the shape,and it can accurately describe global features and partial features of the shape. This method is invariant to translation,rotation,scaling,and it has strong ability to describe and recognize shapes.
To improve the shape description and identification accuracy of 2 D objects,an all-directional sequence feature description based on F-histogram is proposed. In this method,the shape is segmented omnidirectionally by the direction lines,the sequential feature description quantity among object segments in the direction line is computed,the omnidirectional sequence feature description quantity of the shape is established,and the similarity among shapes is calculated. The ground object shape data set,MPEG-7 CE-1 Part B shape data set,and Swedish leaf database are used to test the retrieval and identification performance of this method. The method is then compared with other description methods on shape similarity. Experimental results show that this method exhibited high retrieval accuracy. The all-directional sequence feature description method integrates the boundary and regional information of the shape,and it can accurately describe global features and partial features of the shape. This method is invariant to translation,rotation,scaling,and it has strong ability to describe and recognize shapes.
引文
[1]周瑜,刘俊涛,白翔.形状匹配方法研究与展望[J].自动化学报,2012,38(6):889-910.ZHOU Yu,LIU Juntao,BAI Xiang.Research and perspective on shape matching[J].Acta automatica sinica,2012,38(6):889-910.
[2]ZHAO C J,SHI W K,DENG Y.A new Hausdorff distance for image matching[J].Pattern recognition letters,2005,26(5):581-586.
[3]BAI X,WANG B,YAO C,et al.Co-transduction for shape retrieval[J].IEEE transactions on image processing,2012,21(5):2747-2757.
[4]LING H B,JACOBS D W.Shape classification using the inner-distance[J].IEEE transactions on pattern analysis and machine intelligence,2007,29(2):286-299.
[5]LI Z,KUANG Z,LIU Y,et al.Multiscale shape context and re-ranking for deformable shape retrieval[J].Computers&graphics,2016,54:8-17.
[6]KOPF S,HAENSELMANN T,EFFELSBERG W.Enhancing curvature scale space features for robust shape classification[C]//2005 IEEE International Conference on Multimedia and Expo.New York,2005:478-481.
[7]ALAJLAN N,EL RUBE I,KAMEL M S,et al.Shape retrieval using triangle-area representation and dynamic space warping[J].Pattern recognition,2007,40(7):1911-1920.
[8]FAN H,ZIPF A,FU Q,et al.Quality assessment for building footprints data on Open Street Map[J].International journal of geographical information science,2014,28(4):700-719.
[9]田泽宇,门朝光,汤亚楠.基于形状及空间关系的场景相似性检索[J].电子学报,2016,44(8):1892-1898.TIAN Zeyu,MEN Chaoguang,TANG Yanan.Spatial-scene similarity retrieval based on shape and spatial relation[J].Acta electronica sinica,2016,44(8):1892-1898.
[10]YANG J,WANG H,YUAN J,et al.Invariant multi-scale descriptor for shape representation,matching and retrieval[J].Computer vision and image understanding,2016,145:43-58.
[11]KAOTHANTHONG N,CHUN J,TOKUYAMA T.Distance interior ratio:a new shape signature for 2D shape retrieval[J].Pattern recognition letters,2016,78:14-21.
[12]王斌.一种基于多尺度拱高形状描述的图像检索方法[J].电子学报,2013,41(9):1821-1825.WANG Bin.Image retrieval using multi-scale arch height shape description[J].Acta electronica sinica,2013,41(9):1821-1825.
[13]WANG B,GAO Y.Hierarchical string cuts:a translation,rotation,scale,and mirror invariant descriptor for fast shape retrieval[J].IEEE transactions on image processing,2014,23(9):4101-4111.
[14]MEI Y,ANDROUTSOS D.Robust affine invariant regionbased shape descriptors:the ica zernike moment shape descriptor and the whitening zernike moment shape descriptor[J].IEEE signal processing letters,2009,16(10):877-880.
[15]ERDEM A,TARI S.A similarity-based approach for shape classification using Aslan skeletons[J].Pattern recognition letters,2010,31(13):2024-2032.
[16]GURU D S,NAGENDRASWAMY H S.Symbolic representation of two-dimensional shapes[J].Pattern recognition letters,2007,28(7):144-155.
[17]FREITAS A M,TORRES R S,MIRANDA P A V.TSS&TSB:Tensor scale descriptors within circular sectors for fast shape retrieval[J].Pattern recognition letters,2016,83:303-311.
[18]王斌.一种不变的基于傅立叶变换的区域形状描述子[J].电子学报,2012,40(1):84-88.WANG Bin.An invariant region-shape descriptor based on fourier transform[J].Chinese journal of electronics,2012,40(1):84-88.
[19]NI J,MATSAKIS P.An equivalent definition of the histogram of forces:theoretical and algorithmic implications[J].Pattern recognition,2010,43(4):1607-1617.
[20]ZUNIC J,ROSIN P L,KOPANJA L.On the orientability of shapes[J].IEEE transactions on image processing,2006,15(11):3478-3487.
[2]ZHAO C J,SHI W K,DENG Y.A new Hausdorff distance for image matching[J].Pattern recognition letters,2005,26(5):581-586.
[3]BAI X,WANG B,YAO C,et al.Co-transduction for shape retrieval[J].IEEE transactions on image processing,2012,21(5):2747-2757.
[4]LING H B,JACOBS D W.Shape classification using the inner-distance[J].IEEE transactions on pattern analysis and machine intelligence,2007,29(2):286-299.
[5]LI Z,KUANG Z,LIU Y,et al.Multiscale shape context and re-ranking for deformable shape retrieval[J].Computers&graphics,2016,54:8-17.
[6]KOPF S,HAENSELMANN T,EFFELSBERG W.Enhancing curvature scale space features for robust shape classification[C]//2005 IEEE International Conference on Multimedia and Expo.New York,2005:478-481.
[7]ALAJLAN N,EL RUBE I,KAMEL M S,et al.Shape retrieval using triangle-area representation and dynamic space warping[J].Pattern recognition,2007,40(7):1911-1920.
[8]FAN H,ZIPF A,FU Q,et al.Quality assessment for building footprints data on Open Street Map[J].International journal of geographical information science,2014,28(4):700-719.
[9]田泽宇,门朝光,汤亚楠.基于形状及空间关系的场景相似性检索[J].电子学报,2016,44(8):1892-1898.TIAN Zeyu,MEN Chaoguang,TANG Yanan.Spatial-scene similarity retrieval based on shape and spatial relation[J].Acta electronica sinica,2016,44(8):1892-1898.
[10]YANG J,WANG H,YUAN J,et al.Invariant multi-scale descriptor for shape representation,matching and retrieval[J].Computer vision and image understanding,2016,145:43-58.
[11]KAOTHANTHONG N,CHUN J,TOKUYAMA T.Distance interior ratio:a new shape signature for 2D shape retrieval[J].Pattern recognition letters,2016,78:14-21.
[12]王斌.一种基于多尺度拱高形状描述的图像检索方法[J].电子学报,2013,41(9):1821-1825.WANG Bin.Image retrieval using multi-scale arch height shape description[J].Acta electronica sinica,2013,41(9):1821-1825.
[13]WANG B,GAO Y.Hierarchical string cuts:a translation,rotation,scale,and mirror invariant descriptor for fast shape retrieval[J].IEEE transactions on image processing,2014,23(9):4101-4111.
[14]MEI Y,ANDROUTSOS D.Robust affine invariant regionbased shape descriptors:the ica zernike moment shape descriptor and the whitening zernike moment shape descriptor[J].IEEE signal processing letters,2009,16(10):877-880.
[15]ERDEM A,TARI S.A similarity-based approach for shape classification using Aslan skeletons[J].Pattern recognition letters,2010,31(13):2024-2032.
[16]GURU D S,NAGENDRASWAMY H S.Symbolic representation of two-dimensional shapes[J].Pattern recognition letters,2007,28(7):144-155.
[17]FREITAS A M,TORRES R S,MIRANDA P A V.TSS&TSB:Tensor scale descriptors within circular sectors for fast shape retrieval[J].Pattern recognition letters,2016,83:303-311.
[18]王斌.一种不变的基于傅立叶变换的区域形状描述子[J].电子学报,2012,40(1):84-88.WANG Bin.An invariant region-shape descriptor based on fourier transform[J].Chinese journal of electronics,2012,40(1):84-88.
[19]NI J,MATSAKIS P.An equivalent definition of the histogram of forces:theoretical and algorithmic implications[J].Pattern recognition,2010,43(4):1607-1617.
[20]ZUNIC J,ROSIN P L,KOPANJA L.On the orientability of shapes[J].IEEE transactions on image processing,2006,15(11):3478-3487.