基于内容的图像检索,即CBIR(Content-based image retrieval),是计算机视觉领域中关注大规模数字图像内容检索的研究分支。典型的CBIR系统,允许用户输入一张图片,以查找具有相同或相似内容的其他图片。而传统的图像检索是基于文本的,即通过图片的名称、文字信息和索引关系来实现查询功能。
本文是Wiki上统计的当前主要的基于内容的图像检索系统。
商业图像搜索引擎:
Elastic Vision | Smart image searcher with content-based clustering in a visual network. | No | No | Private Company | Closed | |
Google Image Search | Google's CBIR system, note: does not work on all images | Yes | Yes | Public Company | Closed | |
Baidu Image Search | Baidu's CBIR system | Yes | Yes | 1000M | Public Company | Closed |
ID My Pill | Automatic prescription pill identification (CBIR) | Yes | No | Private Company | Open (via API) | |
Imense Image Search Portal | CBIR search engine, by Imense. | No | Yes | 3M | Private Company | Closed |
Imprezzeo Image Search | CBIR search engine, by Imprezzeo. | No | Yes | Private Company | Closed | |
Visual Image Search | CBIR search engine, by pixolution | Yes | No | 10M | Private Company | Closed |
Incogna Image Search | CBIR search engine, by Incogna Inc. | No | Yes | 100M | Private Company | Closed |
Like.com | Shopping & fashion based CBIR engine | No | Yes | 1M | Private Company | Closed |
Chic Engine | Visual fashion search engine (CBIR) | Yes | No | Private Company | Closed | |
MiPai similarity search engine | Online similarity search engine | Yes | Yes | 100M | Individual | Closed |
Piximilar | Demo engine, developed by Idee Inc. | No | No | 3M | Private Company1 | Closed |
Empora | Product comparison & shopping using CBIR for product images. Previously known as Pixsta | No | Yes | 0.5M | Private Company | Closed |
Shopachu | Shopping & fashion CBIR engine, by Incogna Inc. | No | Yes | 1M | Private Company | Closed |
TinEye | CBIR site for finding variations of web images, by Idee Inc. | Yes | No | 1800M | Private Company | Closed |
Tiltomo | CBIR system using Flickr photos | No | Yes | Private Company | Closed | |
eBay Image Search | Image Search for eBay Fashion | No | Yes | 20M | Public Company | Closed |
Visual Recognition Factory | Visual Recognition Factory | Yes | Yes | Private Company | Closed | |
IMMENSELAB | CBIR search engine byKBKGROUP. | Yes | No | 10M | Private Company | Closed |
Macroglossa Visual Search | CBIR visual search engine | Yes | No | Private Company | Closed |
研究项目(有一部分是开源的):
名称 |
外部图像查询 |
元数据查询 |
索引大小 (估计值, Millions of Images) |
组织 |
协议 (Open/Closed) |
Yes |
Yes |
15M |
University |
Closed |
|
Yes |
Yes |
University |
Closed |
||
Yes |
No |
0.225M |
University |
Open |
|
Yes |
No |
University |
GPL |
||
Yes |
No |
Desktop-based |
University |
||
No |
No |
University |
Open |
||
Yes |
No |
Desktop-based |
GNU |
||
Yes |
Yes |
free-beta limited to 4k images |
Private Company |
Closed |
|
Yes |
No |
Desktop-based |
Individual |
Closed |
|
Yes |
No |
Individual |
|||
Yes |
Yes |
University |
Closed |
||
Yes |
No |
Desktop-based |
Research Institute |
Closed |
|
Yes |
Yes |
University |
GPL |
||
Yes |
Yes |
106M |
Research Institute |
Closed |
|
No |
No |
106M |
Research Institute |
Open |
|
Yes |
No |
Desktop-based |
University |
Closed |
|
Yes |
Yes |
130M |
University |
Closed |
|
No |
No |
12M |
University |
Closed |
|
Yes |
No |
Desktop-based |
Private company |
Closed |
|
Yes |
No |
Private Company / University |
Closed |
||
No |
No |
University |
Closed |
||
No |
No |
University |
Closed |
||
Yes |
No |
0.002M |
University |
Closed |
|
No |
Yes |
Institute |
Closed |
||
Yes |
Yes |
2.221M |
University |
Closed |
|
Yes |
No |
University |
Closed |
||
Yes |
No |
University |
Closed |
||
SHIATSU |
Yes |
Yes |
University |
Closed |
用过其中的LIRE和Caliph & Emir,还是比较好用的
相关推荐
使用MFC开发的基于内容的图像检索系统,实现以图搜图。程序主要基于图片的颜色、纹理、形状特征,支持多特征权重结合进行检索。可设置本地带检索图片、检索目录、检索方法等参数,按相似度依次显示检索图片并给出...
基于内容的图像检索系统是一种先进的图像搜索技术,它不再依赖于传统的基于文本的关键词搜索,而是通过分析和比较图像本身的特征来查找相似的图像。这个实验报告主要涉及颜色特征的提取、其他特征的选取、相似度度量...
新基于内容的图像检索系统是计算机视觉领域的一个重要研究方向,它主要依赖于对图像内容的理解,而非传统的基于关键字或元数据的检索方式。在这个系统中,图像被转化为可计算的特征向量,这些特征可以反映图像的颜色...
《基于内容图像检索技术》从理论方法研究与实现技术角度,总结归纳了基于内容图像检索(CBIR)技术的研究与进展,并融入了作者多年来... 第11章 图像检索系统性能评价 第12章 基于内容图像检索的原型系统及应用实例
在实际应用中,基于内容的图像检索系统可能会包含以下步骤: 1. **特征提取**:从原始图像中提取HSV和CVV特征,这可能涉及到颜色量化、纹理分析和形状描述。 2. **特征表示**:将提取的特征转化为适合比较的形式,...
【基于VC++的图像检索系统】是一个利用Microsoft Visual C++ 6.0(简称VC6)编译器开发的高效应用程序,主要目标是实现快速、有效的图像搜索功能。在这个系统中,开发者可能利用了VC++的强大性能和丰富的库支持来...
**基于OpenCV的图像检索系统**是计算机视觉领域的一个重要应用,主要目的是通过特定的算法在大量图像数据库中快速找到与目标图像相似的图像。在这个系统中,我们使用了OpenCV(开源计算机视觉库)这一强大的工具,它...
这些文件很可能是用来测试或训练图像检索系统的样本图像。通过运行MATLAB程序,我们可以分析这些图像的颜色和纹理特征,然后进行图像之间的匹配和检索,从而验证和优化算法的性能。 总的来说,"基于内容图像检索"是...
【基于颜色的图像检索系统】 图像检索是计算机视觉领域中的一个重要课题,其目的是从大量图像数据库中找到与用户查询图像最相似的图像。基于颜色的图像检索系统是一种常见的方法,它主要依赖于图像的颜色特征来识别...
在传统的基于内容的图像检索系统中,主要依赖于图像的低级可视化特征(如颜色、纹理、形状等)来进行检索。但由于存在所谓的“语义鸿沟”问题,即机器从低级的可视化特征得到的相似性与人从高级的语义特征得到的相似...
Java 实现的基于内容的图像检索工具包是一个强大的软件框架,专门用于在大量图像数据库中寻找与查询...理解并实践这个工具包的原理和实现,不仅可以提升图像处理技能,也能为开发自己的图像检索系统提供宝贵的参考。
基于内容的图像检索系统(CBIR,Content-Based Image Retrieval)是一种先进的图像搜索技术,它允许用户通过输入图像或图像的特定特征(如颜色、纹理、形状等)来查询相似的图像。在这个“颜色+纹理”系统中,我们...