An In-depth Analysis of Co-occurrence Matrix-based Features
| Autor: | Seiichi Serikawa, Shahera Hossain |
|---|---|
| Rok vydání: | 2013 |
| Předmět: |
business.industry
ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Pattern recognition Property analysis computer.software_genre Texture (geology) Image (mathematics) Co-occurrence matrix Image texture Data mining Artificial intelligence business computer Image based Feature detection (computer vision) Mathematics |
| Zdroj: | Journal of the Institute of Industrial Applications Engineers. 1:35-42 |
| ISSN: | 2187-8811 |
| DOI: | 10.12792/jiiae.1.35 |
| Popis: | In this paper, we do in-depth analysis of image based on gray-level co-occurrence matrices for various features. Image texture and property analysis has ample applications in various fields. Hence, we employ eleven different features on standard dataset to evaluate them on twenty-one images of varied textures and increasing complexities. We implement these features based on the co-occurrence matrices on each image. We analyze these features and demonstrated their variations and impact on different gray-scale images. This is the first time to evaluate these in such extensive manner and we find significant variations from which one can evaluate the required features for specific image classes. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|