Abstract:In this paper, we present a superpixel segmentation algorithm called linear spectral clustering (LSC), which is capable of producing superpixels with both high boundary adherence and visual compactness for natural images with low computational costs. In LSC, a normalized cuts-based formulation of image segmentation is adopted using a distance metric that measures both the color similarity and the space proximity between image pixels. However, rather than directly using the traditional eigen-based algorithm, we approximate the similarity metric through a deliberately designed kernel function such that pixel values can be explicitly mapped to a high-dimensional feature space. We then apply the conclusion that by appropriately weighting each point in this feature space, the objective functions of the weighted K-means and the normalized cuts share the same optimum points. Consequently, it is possible to optimize the cost function of the normalized cuts by iteratively applying simple K-means clustering in the proposed feature space. LSC possesses linear computational complexity and high memory efficiency, since it avoids both the decomposition of the affinity matrix and the generation of the large kernel matrix. By utilizing the underlying mathematical equivalence between the two types of seemingly different methods, LSC successfully preserves global image structures through efficient local operations. Experimental results show that LSC performs as well as or even better than the state-of-the-art superpixel segmentation algorithms in terms of several commonly used evaluation metrics in image segmentation. The applicability of LSC is further demonstrated in two related computer vision tasks.

Pixel clustering for color image segmentation

Hierarchical pixel clustering for image segmentation

A Model of Pixel and Superpixel Clustering for Object Detection

A robust image segmentation method using hierarchical color clustering.

Reversible Image Merging for Low-level Machine Vision

Image segmentation based on modified superpixel segmentation and spectral clustering

Pixel Intensity Clustering Algorithm for Multilevel Image Segmentation

Nonparametric clustering for image segmentation

Image Segmentation Algorithm Based on Superpixel Clustering

Image Segmentation by Using Linear Spectral Clustering

An effective color image segmentation approach using superpixel-neutrosophic C-means clustering and gradient-structural similarity

Image Clustering Algorithm Using Superpixel Segmentation and Non-Symmetric Gaussian-Cauchy Mixture Model.

A Semi-Supervised Color Image Segmentation Method

Research for Color Image Segmentation Based on Hybrid Clustering

Efficient Clustering Approach for Adaptive Unsupervised Colour Image Segmentation

Pixel-level clustering network for unsupervised image segmentation

Image segmentation via coherent clustering in Lab* color space

Color Image Segmentation Based on Blocks Clustering and Region Growing

A Practical GrabCut Color Image Segmentation Based on Bayes Classification and Simple Linear Iterative Clustering.

Linear Spectral Clustering Superpixel.

Improved Simple Linear Iterative Clustering Algorithm Using Hsl Color Space