| 基于双线性模板和分块策略的虹膜色素斑检测 |
| |
| 引用本文： | 朱立军,苑玮琦.基于双线性模板和分块策略的虹膜色素斑检测[J].仪器仪表学报,2015,36(12):2714-2721. |
| |
| 作者姓名： | 朱立军 苑玮琦 |
| |
| 作者单位： | 1.沈阳工业大学信息科学与工程学院； 2.沈阳化工大学计算机科学与技术学院 |
| |
| 基金项目： | 国家自然科学基金(61271365)项目资助 |
| |
| 摘 要： | 本文色素斑是一种重要的虹膜特征纹理。然而,由于该纹理颜色深浅不一,形状各异,因此如何从具有复杂背景的可见光虹膜图像中检测出该纹理仍然是目前亟待解决的一个难题。提出一种基于双线性模板和分块策略的可见光虹膜色素斑检测方法,首先使用两个不同长度的线性模板来提取虹膜上所有不同类型的特征纹理;然后,采用分块策略得到一个能描述该纹理灰度分布特征的特征向量;最后,使用BP神经网络实现对色素斑纹理的检测。该算法对图库中色素斑纹理检测的正确率达到81.25%,对非色素斑纹理的正确识别率达到83.97%,总的正确检测率达到83.23%。实验结果表明,该方法能够从背景复杂的可见光虹膜图像中检测出色素斑纹理,为虹膜的特征提取提供了一个新的思路。
|
| 关 键 词： | 虹膜色素斑 双线性模板 分块策略 区域生长 BP神经网络 |
Iris pigmented spot detection based on bilinear templates and
block partitioning strategy |
| |
| Zhu Lijun,Yuan Weiqi.Iris pigmented spot detection based on bilinear templates and
block partitioning strategy[J].Chinese Journal of Scientific Instrument,2015,36(12):2714-2721. |
| |
| Authors: | Zhu Lijun Yuan Weiqi |
| |
| Abstract: | Pigmented spot is an important feature of iris texture. However, due to the fact that the shades and shapes of the pigmented spots are different, how to detect the texture from the visible iris image with complex background is still a challenging task to be solved at present. In order to solve this problem, a visible iris pigmented spot detection method based on bilinear templates and block partitioning strategy is proposed in this paper. First, two linear templates with different lengths are used to extract all the different types of feature textures on the iris; then, the block partitioning strategy is applied to obtain a feature vector that can describe the grayscale distribution characteristics of the feature textures; and finally, the textures of the pigmented spots are detected using BP neural network. To test the performance of this method, the method was used to detect the textures of the pigmented spots in our image gallery; and the results indicate that the correct detection rate for the textures of pigmented spots is 81.25%, while the correct detection rate for the textures of non-pigmented spots is 83.97%, and the total correct detection rate is 83.23%. The experiment results show that the proposed method can detect the textures of the pigmented spots from the visible iris images with complex background, which provides a new idea for iris feature extraction. |
| |
| Keywords: | iris pigmented spot bilinear template block partitioning strategy region growing BP neural network |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《仪器仪表学报》浏览原始摘要信息 |
|
点击此处可从《仪器仪表学报》下载全文 |
|
