Fuzzy-based algorithm for color recognition of license plates
Fuzzy-based algorithm for color recognition of license plates
复制标题
DOI：
10.1016/j.patrec.2008.01.026
复制
发表时间：
2008-05
期刊：
影响因子：
--
通讯作者：
Feng Wang;Lichun Man;Bangping Wang;Yijun Xiao;Wei Pan;Xiaochun Lu
中科院分区：
文献类型：
--
作者：
Feng Wang;Lichun Man;Bangping Wang;Yijun Xiao;Wei Pan;Xiaochun Lu
车牌颜色识别在车牌识别系统中起着重要的作用。但由于车牌的外观受到各种因素的影响，如照明、相机特性等，因此这可能是一项具有挑战性的任务。而且不同地方的车牌颜色特征可能会有很大的不同。为了解决这些问题，本文提出了一种基于模糊逻辑的算法。采用HSV （hue, saturation and value）色彩空间进行色彩特征提取。首先将HSV空间的三个分量根据不同的隶属函数映射到模糊集；然后，用三个加权隶属度的融合来描述颜色识别的模糊分类函数。为了适应所提出的算法，我们还提出了一种获取相关参数的学习算法。在基于dsp的嵌入式LPR平台上，在三组测试图像中与其他分类器进行比较。实验结果表明，该算法具有较高的分类精度和较好的自适应性。
Color recognition of license plates plays an important role in a license plate recognition (LPR) system. But it can be a challenging task as the appearances of license plates are affected by various factors such as illumination, camera characteristics, etc. And the color features of license plates in different places may be quite different. To address these concerns, this paper presents an algorithm based on fuzzy logic. The HSV (hue, saturation and value) color space is employed to perform color feature extraction. Three components of the HSV space are firstly mapped to fuzzy sets according to different membership functions. The fuzzy classification function for color recognition is, then, described by the fusion of three weighted membership degrees. For adaptation of the proposed algorithm, we also present a learning algorithm to obtain the correlative parameters. On a DSP-based embedded LPR platform, comparisons were drawn with other classifiers within three sets of test images. Experimental results show that the proposed algorithm achieves higher classification accuracy and better adaptability.