New graph-based features for shape recognition

Shape recognition is a main challenging problem in computer vision. Different approaches and tools are used to solve this problem. Most existing approaches to object recognition are based on pixels. Pixel-based methods are dependent on the geometry and nature of the pixels, so the destruction of pixels reduces their performance. In this paper, we construct a graph that captures the topological and geometrical properties of the object. Then, using the coordinate and relation of its vertices, we extract features that are robust with respect to noise, rotation, scale variation and articulation. To evaluate our method, we provide different comparisons with state-of-the-art results on various known benchmarks, including Kimia's, Tari56, Tari1000, Tetrapod and articulated datasets. We provide the analysis of our method against different variations. The results confirm the advantage of this method in different datasets, especially in the presence of noise.

Automated summary

This paper introduces a graph-based method for shape recognition, which can extract features robust to noise, rotation, scale variation, and articulation, showing advantages in different datasets.