Colour segmentation and highlight detection on a 1-D feature space |
| |
| Affiliation: | 1. School of Mathematical Science, Anhui University, Hefei 230601, PR China;2. Key Laboratory of Computer Network and Information Integration (Southeast University), Ministry of Education, Nanjing 211189, PR China;3. Department of Communication Engineering, North University of China, Taiyuan, Shan’xi 030051, PR China;4. Department of Modern Physics, University of Science and Technology of China, Hefei 230026, PR China;1. Department of Library, Information and Archives, Shanghai University, Shanghai 200444, China;2. School of Information Management, Wuhan University, Wuhan 430072, China;3. School of Informatics and Computing, Indiana University, IN 47408, USA;4. University Library, Tongji University, Shanghai 200092, China;1. Institute of Fundamental and Frontier Science, University of Electronic Science and Technology of China, Chengdu 610054, China;2. Yingcai Honors of school, University of Electronic Science and Technology of China, Chengdu 610054, China;3. School of Education, Shannxi Normal University, Xi’an 710062, China;4. Science, Mathematics and Technology Cluster, Singapore University of Technology and Design (SUTD), S487372, Singapore;5. SUTD-Massachusetts Institute of Technology International Design Centre, Singapore;1. School of Transportation and Logistics, Southwest Jiaotong University, Chengdu 611756, China;2. National United Engineering Laboratory of Integrated and Intelligent Transportation, Southwest Jiaotong University, Chengdu 610031, China;3. National Engineering Laboratory of Integrated Transportation Big Data Application Technology, Southwest Jiaotong University, Chengdu 610031, China;4. Institute of System Science and Engineering, Southwest Jiaotong University, Chengdu 611756, China;1. Polytechnic School, Catholic University of Murcia, Campus Los Jeronimos, s/n, E-30107 Murcia, Spain;2. Department of Computer Science and Artificial Intelligence, University of Alicante, Campus de San Vicente, Ap. Correos 99, E-03080, Alicante, Spain |
| |
| Abstract: | Specularly reflecting surfaces often corrupt a real scene and degrade system performance. In this paper, we present an algorithm for colour segmentation as well as highlight detection. This algorithm models the human colour vision perception with the physical properties of sensors, illumination and surface reflectances. For image to be dependent only on the body reflectance, the reflection due to illumination, shading and highlights should be discounted. The dichromatic reflection model which describes the colour of the reflected light as a linear combination of the colour of the light due to surface reflection (highlights) and body reflection (object colour) is used. The input image data is first mapped from device coordinates onto the CIE 1931 xy chromaticity diagram where colour clustering occurs. Since the feature space is 3-D, colour clustering is a computationally expensive process. A more efficient method is the dimensionality reduction of the feature space. Using information from the xy chromaticity diagram, the estimated colour clusters are then projected onto a line for 1-D thresholding. The error of projection is minimized by the Fisher's linear discriminant method. A program recognizing nine colours on a 1024 × 1024 image is presented. Colour segmentation of 99% accuracy within 25 s is also achieved. |
| |
| Keywords: | |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
