Conference Paper

Fast segmentation based on a hybrid of clustering and morphological approaches

Fac. of Electron. & Telecommun., Hanoi Univ. of Technol., Hanoi
DOI: 10.1109/CCE.2008.4578952 Conference: Communications and Electronics, 2008. ICCE 2008. Second International Conference on
Source: IEEE Xplore

ABSTRACT This paper proposes a fast segmentation method for still image based on a hybrid of clustering and morphological segmentation approach. The objective of the clustering is to partition an input image into a number of clusters such that the gray levels within each cluster are similar. The clustered image is further processed by using morphological segmentation approach, in which a seeded region growing however plays a role of the decision tool instead of a watershed algorithm for a remarkable improvement of processing time. The performance of the proposed method is evaluated by comparing its region-based coding results with those of the morphological watershed-based segmentation method and the split-and-merge algorithm. The experiments results showed that region-based coding using the proposed algorithm yields PSNR improvement of about 1.5 dB over the morphological watershed-based method. Especially, the total time elapsed to segment an image using the proposed method is reduced about 1/6 and 1/3 compared with those of the watershed-based segmentation and the split-and-merge methods, respectively.

0 Followers
 · 
86 Views
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: A fast and flexible algorithm for computing watersheds in digital gray-scale images is introduced. A review of watersheds and related motion is first presented, and the major methods to determine watersheds are discussed. The algorithm is based on an immersion process analogy, in which the flooding of the water in the picture is efficiently simulated using of queue of pixel. It is described in detail provided in a pseudo C language. The accuracy of this algorithm is proven to be superior to that of the existing implementations, and it is shown that its adaptation to any kind of digital grid and its generalization to n -dimensional images (and even to graphs) are straightforward. The algorithm is reported to be faster than any other watershed algorithm. Applications of this algorithm with regard to picture segmentation are presented for magnetic resonance (MR) imagery and for digital elevation models. An example of 3-D watershed is also provided
    IEEE Transactions on Pattern Analysis and Machine Intelligence 07/1991; 13(6):583-598. DOI:10.1109/34.87344
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The generalized Lloyd algorithm plays an important role in the design of vector quantizers (VQ) and in feature clustering for pattern recognition. In the VQ context, this algorithm provides a procedure to iteratively improve a codebook and results in a local minimum that minimizes the average distortion function. In this research, we propose an efficient method to obtain a good initial codebook that can accelerate the convergence of the generalized Lloyd algorithm and achieve a better local minimum as well.
    IEEE Signal Processing Letters 11/1994; 1(10-1):144 - 146. DOI:10.1109/97.329844
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: This paper presents a region-based coding algorithm for video sequences. The coding approach involves a time-recursive segmentation relying on the pixels homogeneity, a region-based motion estimation, and motion compensated contour and texture coding. This algorithm is mainly devoted to very low bit rate video coding applications. One of the important features of the approach is that no assumption is made about the sequence content. Moreover the algorithm structure leads to a scalable coding process giving various levels of quality and bit rates. The coding as well as the segmentation are controlled to regulate the bit stream. Finally, the interest of morphological tools in the content of region-based coding is extensively reviewed
    Proceedings of the IEEE 07/1995; DOI:10.1109/5.387088

Preview

Download
0 Downloads