We propose an innovative and efficient approach to improve k-view-template (k-view-T) and k-view-datagram (k-view-D) algorithms for image texture classification. The proposed approach, called the weighted k-view-votin...
详细信息
We propose an innovative and efficient approach to improve k-view-template (k-view-T) and k-view-datagram (k-view-D) algorithms for image texture classification. The proposed approach, called the weighted k-view-voting algorithm (k-view-V), uses a novel voting method for texture classification and an accelerating method based on the efficient summed square image (SSI) scheme as well as fast Fourier transform (FFT) to enable overall faster processing. Decision making, which assigns a pixel to a texture class, occurs by using our weighted voting method among the "promising" members in the neighborhood of a classified pixel. In other words, this neighborhood consists of all the views, and each view has a classified pixel in its territory. Experimental results on benchmark images, which are randomly taken from Brodatz Gallery and natural and medical images, show that this new classification algorithm gives higher classification accuracy than existing k-view algorithms. In particular, it improves the accurate classification of pixels near the texture boundary. In addition, the proposed acceleration method improves the processing speed of k-view-V as it requires much less computation time than other k-view algorithms. Compared with the results of earlier developed k-view algorithms and the gray level co-occurrence matrix (GLCM), the proposed algorithm is more robust, faster, and more accurate. (c) 2012 Society of Photo-Optical Instrumentation Engineers (SPIE). [DOI: 10.1117/***.51.2.027004]
Textural features is very important properties in many types of images. Partitioning an image into homogeneous regions based on textural features is useful in computer vision. Many texture classification algorithms ha...
详细信息
ISBN:
(纸本)9780769547213;9781467319027
Textural features is very important properties in many types of images. Partitioning an image into homogeneous regions based on textural features is useful in computer vision. Many texture classification algorithms have been proposed including Local Binary Patterns, Gray Level Co-Occurrence and k-view based algorithms, to name a few. Among of them, The k-view using Rotation-invariant feature algorithm (k-view-R) and the fast weighted k-view-Voting algorithm (k-view-V) produce higher classification accuracy by compare with those original k-view based algorithms. However, there still have some rooms for improvement. In this paper, by analyzing those k-view based algorithms, an attempt to utilize the advantages of the k-view-R and k-view-V was investigated. The new approach which we called combinatorial k-view based method was presented. To test and evaluate the proposed method, some experiments were carried out on a lot of textural images which taken from a standard database. Preliminary experimental results demonstrated the new method achieved more accurate classification by compare with other k-view based methods.
暂无评论