but personally I haven't used in production (just for testing) these 2 libraries since the documentation is not very good/detailed, and since this kind of problem is quite "cutting edge" so the documentation and an explanation paper is much more important than any source code.
If none of the above approaches work for you, I would suggest to use a MATLAB based iris recognition examples (because there are many MATLAB based implementations, papers, theses) and than convert them to Java or use the directly from Java (since MATLAB has support for this).
I tried the project iris tool. It has the foundation blocks for IRIS Matching SYstem. It has the following features: * Basic QT UI - Needs qt-make and qt-sdk on top of gc++ and make to build * Iris Detection from Image file. - The Image should be 320x280 windows bmp greyscale 8 bit indexed image. - You can take an iris snap from you camera phone. It took a 12M regular jpeg image and converted it to required format using gimp. * Iris Image database: Text file db to store Iris Images. - Means you could extend the tool to create a Iris Search database - I could take another iris image and try to match with existing image
发布评论
评论（4）
我没有使用过“只是完成”您所描述的全部工作的 Java 库，但使用 JavaVis 并不困难（http://sourceforge.net/projects/javavis/) 如以下论文中描述的示例代码所示：http://imai-software.com/openlab/data/IMAI-JOUNALVI.pdf#page=43（参见从第 43 页开始）-那篇论文非常好，而且长度也很长:)。
当然，有一些“有效”的解决方案，例如：
但我个人没有在生产中使用（仅用于测试）这两个库，因为文档不是很好/详细，并且由于此类问题非常“前沿”，因此文档和解释文件比任何源代码都重要得多。
如果上述方法都不适合您，我建议使用基于 MATLAB 的虹膜识别示例（因为有许多基于 MATLAB 的实现、论文、论文），然后将它们转换为 Java 或直接使用 Java 中的（因为 MATLAB 有对此表示支持）。
I haven't used Java libraries that "just do" that entire work you describe, but it's not that difficult to use JavaVis (http://sourceforge.net/projects/javavis/) as in the sample code described in the following paper: http://imai-software.com/openlab/data/IMAI-JOUNALVI.pdf#page=43 (see from page 43) - that paper is quite good and for it's length :).
Of course, there are a few "working" solutions, e.g.:
but personally I haven't used in production (just for testing) these 2 libraries since the documentation is not very good/detailed, and since this kind of problem is quite "cutting edge" so the documentation and an explanation paper is much more important than any source code.
If none of the above approaches work for you, I would suggest to use a MATLAB based iris recognition examples (because there are many MATLAB based implementations, papers, theses) and than convert them to Java or use the directly from Java (since MATLAB has support for this).
我会研究OpenCV，它是一个用于计算机视觉的 C++ 库。 Java 包装器可用：
我相信您会想研究圆的霍夫变换。 此页面提到了它们
I would look into OpenCV which is a C++ library for computer vision. Java wrappers are available:
I believe you will want to look into Hough transforms for circles. This page mentions them
IEEE 材料可用，但需要订阅。
http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber= 4579428
IEEE material available , but it needs subscriptions.
http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber=4579428
我尝试了项目 iris 工具。它具有 IRIS 匹配系统的基础模块。
它具有以下特点：
* 基本的 QT 用户界面
- 需要在 gc++ 和 make 之上构建 qt-make 和 qt-sdk
* 从图像文件中检测虹膜。
- 图像应为 320x280 windows bmp 灰度 8 位索引图像。
- 您可以使用拍照手机拍摄虹膜快照。它获取了 12M 的常规 jpeg 图像，并使用 gimp 将其转换为所需的格式。
* 虹膜图像数据库：用于存储虹膜图像的文本文件数据库。
- 意味着您可以扩展该工具来创建虹膜搜索数据库
- 我可以拍摄另一个虹膜图像并尝试与现有图像匹配
图像采集不可用。
I tried the project iris tool. It has the foundation blocks for IRIS Matching SYstem.
It has the following features:
* Basic QT UI
- Needs qt-make and qt-sdk on top of gc++ and make to build
* Iris Detection from Image file.
- The Image should be 320x280 windows bmp greyscale 8 bit indexed image.
- You can take an iris snap from you camera phone. It took a 12M regular jpeg image and converted it to required format using gimp.
* Iris Image database: Text file db to store Iris Images.
- Means you could extend the tool to create a Iris Search database
- I could take another iris image and try to match with existing image
Image acquisition is not available.