基于圖像分塊和優(yōu)化累積能量圖的線裁剪算法
doi: 10.11999/JEIT170501
基金項目:
國家自然科學基金(60302018),天津市科技計劃項目(14RCGFGX00846, 15ZCZDNC00130),河北省自然科學基金面上項目(F2015202239)
An Improved Seam Carving Algorithm Based on Image Blocking and Optimized Cumulative Energy Map
Funds:
The National Natural Science Foundation of China (60302018), The Sci-tech Planning Projects Foundation of Tianjin (14RCGFGX00846, 15ZCZDNC00130), The Natural Science Foundation of Hebei Province (F2015202239)
-
摘要: 針對傳統(tǒng)線裁剪方法對圖像過度裁剪造成失真的問題,該文提出一種基于圖像分塊的線裁剪方法。該方法把分塊的思想融入到線裁剪并優(yōu)化累積能量圖,能在一定程度上保護圖像主體區(qū)域,又兼顧背景區(qū)域的裁剪效果。分塊是根據(jù)顯著圖的平均列累加能量向量按照逐列標記的方式把圖像分成保護區(qū)域和非保護區(qū)域,再根據(jù)每個區(qū)域的面積來分配裁剪線的數(shù)目。在裁剪過程中,優(yōu)化了累積能量圖,降低了小面積顯著主體被裁剪掉的可能性。在MSRA數(shù)據(jù)庫上與目前流行的線裁剪及其改進的方法進行對比,并把各種方法得到的縮放結(jié)果圖在互聯(lián)網(wǎng)上進行主觀評價測試,實驗結(jié)果表明該文方法具有更好的主觀縮放效果,對各類圖像的縮放具有普適性。Abstract: For the image distortion of over-carved, this paper proposes a modified Seam Carving (SC) method based on image blocking. Images are segmented into protected and non-protected blocks according to the labelled averaged column summation energy vectors, and then each block is allocated the corresponding carving seams. Moreover, the cumulative energy map is optimized in order to reduce the possibility of the small significant regions to be cut off. This paper fused blocking with the SC method, and optimized the cumulative energy map, which can make a carving balance between the object and background parts. In the MSRA database, the proposed algorithm are compared with the SC method and its improved methods. The experimental results are evaluated on the Internet to test their subjective perceptions, which shows that the proposed method has a better subjective perception, and a general applicability for different images.
-
Key words:
- Image retargeting /
- Seam Carving (SC) /
- Summation energy vector /
- Blocking /
- Cumulative energy map
-
THVENAZ P, BLU T, and UNSER M. Interpolation revisited [medical images application][J]. IEEE Transactions on Medical Imaging, 2000, 19(7): 739-758. doi: 10.1109/42. 875199. SURESHA D and PRAKASH H N. Single picture super resolution of natural images using N-Neighbor Adaptive Bilinear Interpolation and absolute asymmetry based wavelet hard thresholding[C]. 2016 2nd International Conference on Applied and Theoretical Computing and Communication Technology, Bangalore, India, 2016: 387-393. 肖志濤, 馮鐵君, 張芳, 等. 基于角點保護的偏微分方程圖像插值方法[J]. 電子與信息學報, 2015, 37(8): 1892-1899. doi: 10.11999/JEIT141420. XIAO Zhitao, FENG Tiejun, ZHANG Fang, et al. Image interpolation with corner preserving based on partial differential equation[J]. Journal of Electronics Information Technology, 2015, 37(8): 1892-1899. doi: 10.11999/JEIT 141420. CHEN Y L, HUANG T W, CHANG K H, et al. Quantitative analysis of automatic image cropping algorithms: A dataset and comparative study[C]. IEEE Conference on Applications of Computer Vision, Santa Rosa, CA, USA, 2017: 226-234. CHEN Jiansheng, BAI Gaocheng, LIANG Shaoheng, et al. Automatic image cropping: A computational complexity study[C]. IEEE Computer Vision and Pattern Recognition, Las Vegas, NV, USA, 2016: 507-515. AVIDAN S and SHAMIR A. Seam carving for content-aware image resizing[J]. ACM Transactions on Graphics, 2007, 26(3): 101-109. doi: 10.1145/1275808.1276390. 聶棟棟, 馬勤勇, 馬利莊. 基于梯度矢量方向性分析的線裁剪算法[J]. 電子與信息學報, 2012, 34(6): 1506-1510. doi: 10.3724/SP.J.1146.2011.01171. NIE Dongdong, MA Qinyong, and MA Lizhuang. Seam carving algorithm based on gradient vector direction analysis [J]. Journal of Electronics Information Technology, 2012, 34(6): 1506-1510. doi: 10.3724/SP.J.1146.2011.01171. RAZ G, SHMUELI R, and KATZ E. Texture segmentation for seam carving[C]. IEEE Conference on Science of Electrical Engineering, Eilat, Israel, 2017: 1-5. MANSFIELD A, GEHLER P, VAN GOOL L, et al. Scene carving: Scene consistent image retargeting[C]. European Conference on Computer Vision, Heraklion, Crete, Greece, 2010: 143-156. DOMINGUES D, ALAHI A, and VANDERGHEYNST P. Stream carving: An adaptive seam carving algorithm[C]. IEEE International Conference on Image Processing, Hong Kong, China, 2010: 901-904. AGHCHEHKOHAL M G and KUMARA W G C W. Improved seam carving using meta-heuristics algorithms combination[C]. IEEE Signal Processing and Intelligent Systems Conference, Tehran, Iran, 2015: 43-47. LIN Xiao, SHENG Bin, MA Lizhuang, et al. Seamlet carving for shape-aware image resizing[J]. Science China Information Sciences, 2012, 55(5): 1073-1081. doi: 10.1007/s11432-012- 4565-z. ZHOU Bin, WANG Xuanyin, CAO Songxiao, et al. Optimal bi-directional seam carving for compressibility-aware image retargeting[J]. Journal of Visual Communication Image Representation, 2016, 41: 21-30. doi: 10.1016/j.jvcir.2016.09. 002. SHAFIEYAN F, KARIMI N, MIRMAHBOUB B, et al. Image retargeting using depth assisted saliency map[J]. Image Communication, 2016, 50(C): 34-43. doi: 10.1016/j. image.2016.10.006. 趙旦峰, 王博, 楊大偉. 基于隨機置亂的內(nèi)容感知圖像縮放算法[J]. 吉林大學學報(工學版), 2015, 45(4): 1324-1328. doi: 10. 13229/j.cnki.jdxbgxb201504043. ZHAO Danfeng, WANG Bo, and YANG Dawei. Content- aware image based on radom permutation[J]. Journal of Jilin University (Engineering and Technology Edition), 2015, 45(4): 1324-1328. doi: 10.13229/j.cnki. jdxbgxb201504043. ZHU Wangjiang, LIANG Shuang, WEI Yichen, et al. Saliency optimization from robust background detection[C]. IEEE Conference on Computer Vision and Pattern Recognition, Columbus, OH, USA, 2014: 2814-2821. -
計量
- 文章訪問數(shù): 1386
- HTML全文瀏覽量: 168
- PDF下載量: 302
- 被引次數(shù): 0