In this paper an image retrieval technique based on a novel Short Run Length Descriptor (SRLD) is proposed. SRLD can effectively represent image local and global information. It can be viewed as an integrated representation of both color and texture properties. HSV color space is quantized to 72 bins and SRLD is computed using short run lengths of size two and three for each color in different orientations. Short run lengths at all orientations are combined to get Short Run Length Histogram (SRLH) feature. SRLH can thoroughly describe the spatial correlation between color and texture and have the advantages of both statistical and structural approaches of texture representation. The experimental results clearly demonstrate the effectiveness of the proposed descriptor in image retrieval applications.
CITATION STYLE
Shrivastav, N., & Tyagi, V. (2016). A short run length descriptor for image retrieval. In Advances in Intelligent Systems and Computing (Vol. 434, pp. 1–10). Springer Verlag. https://doi.org/10.1007/978-81-322-2752-6_1
Mendeley helps you to discover research relevant for your work.