Computer Technology |
|
|
|
|
Image segmentation using novel adaptive robust bilateral filter |
ZHANG Jian ting,ZHANG Li min |
Naval Aeronautical and Astronautical University, Yantai 264001, China |
|
|
Abstract At first, an efficient and robust bilateral filtering (BF) model was derived from the view of robust regression. Then, a self-adaptive local smoothing scale algorithm was introduced to effectively remove the intensity variations of local image caused by textures and noise. With the new BF model, intensity and texture gradients could be extracted separately from input images and texture features. At last, gradients were fused with morphological method; watershed transform was used to implement image segmentation. Experiments were carried out with Berkeley segmentation database and some remote sensing images. As indicated, the proposed novel adaptive robust bilateral filter method can obtain visual consistent segmentation results. Meanwhile, compared with traditional algorithms, the method can acquire more accurate boundaries and reduce over-segmentation effect, which means that this method can be an effective tool in the preprocessing stage of image analysis.
|
Published: 22 September 2016
|
|
新型自适应稳健双边滤波图像分割
从稳健回归分析的角度,推导得到高效的稳健双边滤波(BF)模型.针对由纹理、噪声等导致的图像局部梯度变化差异性,提出一种自适应的局部平滑尺度方法.运用该双边滤波模型处理原图像和纹理特征图,提取出灰度梯度和纹理梯度.采用形态学方法进行梯度融合,并使用分水岭变换实现图像分割.在Berkeley图像数据集和遥感图像上的实验结果表明:所提出的新型自适应稳健双边滤波方法可以得到视觉一致的分割效果;与传统方法相比,提高了对象边界识别率并降低了过分割效应;可以作为图像分析预处理过程的有效工具.
|
|
[1] BEGHDADI A, LARABI M C, BOUZERDOUM A, et al. A survey of perceptual image processing methods [J]. Signal Processing: Image Communication, 2013, 28(8): 811-831.
[2] ACHANTA R, SHAJI A, SMITH K, et al. SLIC superpixels compared to stateoftheart superpixel methods [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 34(11): 2274-2282.
[3] WANG P, ZENG G, GAN R, et al. Structuresensitive superpixels via geodesic distance [J]. International Journal of Computer Vision, 2013, 103(1): 1-21.
[4] FELZENSZWALB P, HUTTENLOCHER D. Efficient graphbased image segmentation [J]. International Journal of Computer Vision, 2004, 59(2): 167-181.
[5] WITHARANA C, CIVCO D L, MEYER T H. Evaluation of data fusion and image segmentation in earth observation based rapid mapping workflows [J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2014, 87: 1-18.
[6] TABIB MAHMOUDI F, SAMADZADEGAN F, REINARTZ P. Object oriented image analysis based on multiagent recognition system [J]. Computers and Geosciences, 2013, 54: 219-230.
[7] CHAJI N, GHASSEMIAN H. Texturegradientbased contour detection [J]. EURASIP Journal on Applied Signal Processing, 2006(1): 18.
[8] OCALLAGHAN R J, BULL D R. Combined morphologicalspectral unsupervised image segmentation [J]. IEEE Transactions on Image Processing, 2005, 14(1): 49-62.
[9] CORCORAN P, WINSTANLEY A, MOONEY P. Complementary texture and intensity gradient estimation and fusion for watershed segmentation [J]. Machine Vision and Applications, 2011, 22(6): 1027-1045.
[10] TSIOTSIOS C, PETROU M. On the choice of the parameters for anisotropic diffusion in image processing [J]. Pattern Recognition, 2013, 46(5): 1369-1381.
[11] TOMASI C, MANDUCHI R. Bilateral filtering for gray and color images [C]∥ Sixth International Conference on Computer Vision, 1998. New Delhi: IEEE, 1998: 839-846.
[12] 陈潇红, 王维东. 基于时空联合滤波的高清视频降噪算法 [J]. 浙江大学学报:工学版, 2013, 47(5): 853-859.
CHEN Xiaohong, WANG Weidong. A HDTV video denoising algorithm based on spatialtemporal filtering [J]. Journal of Zhejiang University:Engineering Science, 2013, 47(5): 853-859.
[13] BLACK M J, SAPIRO G, MARIMONT D H, et al. Robust anisotropic diffusion [J]. IEEE Transactions on Image Processing, 1998, 7(3): 421-432.
[14] DONG G, ACTON S T. On the convergence of bilateral filter for edgepreserving image smoothing [J]. IEEE Signal Processing Letters, 2007, 14(9): 617-620.
[15] 刘晓光,高兴宝. 一种基于GNC和增广拉格朗日对偶的非凸非光滑图像恢复方法[J]. 电子学报,2014,42(2): 264271.
LIU Xiaoguang, GAO Xingbao. A method based on the GNC and augmented lagrangian duality for nonconvex nonsmooth image restoration [J]. Acta Electronica Sinica, 2014, 42(2): 264271.
[16] NIKOLOVA M. Markovian reconstruction using a GNC approach [J]. IEEE Transactions on Image Processing, 1999, 8(9): 1204-1220.
[17] NAVA R, ESCALANTERAMREZ B, CRISTBAL G. A comparison study of Gabor and logGabor wavelets for texture segmentation [C]∥ 2011 7th International Symposium on Proceedings of the Image and Signal Processing and Analysis (ISPA). Dubrovnik: IEEE, 2011: 189194.
[18] CLAUSI D A, JERNIGAN M. Designing Gabor filters for optimal texture separability [J]. Pattern Recogniti-on, 2000, 33(11): 1835-1849.
[19] SOILLE P. Morphological image analysis: principles and applications [M]. New York: SpringerVerlag, 2003: 206-216.
[20] YANG Y, WANG Y, QIAN J. Building identification from SAR image based on the modified markercontrolled watershed algorithm [C]∥Geoscience and Remote Sensing Symposium. Milan:IEEE, 2015: 2481-2484. |
|
Viewed |
|
|
|
Full text
|
|
|
|
|
Abstract
|
|
|
|
|
Cited |
|
|
|
|
|
Shared |
|
|
|
|
|
Discussed |
|
|
|
|