An Object Detection Method Based on the Joint Feature of the H-S Color Descriptor and the SIFT Feature
Autor: | Yong Kang Peng, Yi Lai Zhang, Shi Dong Zhao, Xi En Cheng, Yi Cheng Li |
---|---|
Rok vydání: | 2018 |
Předmět: |
Fast speed
Computer science business.industry ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Cognitive neuroscience of visual object recognition Scale-invariant feature transform Pattern recognition 02 engineering and technology 010501 environmental sciences Invariant (physics) 01 natural sciences Object detection Computer Science::Computer Vision and Pattern Recognition Histogram Computer Science::Multimedia 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Chromatic scale Artificial intelligence business 0105 earth and related environmental sciences |
Zdroj: | 2018 International Conference on Audio, Language and Image Processing (ICALIP). |
DOI: | 10.1109/icalip.2018.8455641 |
Popis: | Object detection base on pattern feature is an important approach of object recognition. In this paper, we proposed a joint feature consisted of the chromatic feature and the structural feature. First the sift feature of the pattern is adapted to express the structural feature, it has the same invariance with the sift vector. Second we utilize the correlations of the points in sift vector to resolve the match region in patterns. Then the hue-saturation histogram of match region is calculated, and applied to be the color descriptor (denoted as H-S color descriptor) of the match region in our detection method. It is robust and invariant of rotation, scale and transformation. Experiments indicates that our approach has fast speed and high accuracy in our private ceramic datasets. |
Databáze: | OpenAIRE |
Externí odkaz: |