图像检索系统外文翻译--在flickr网站上的基于文本和基于内容的图像检索演示(编辑修改稿)

图像检索系统外文翻译--在flickr网站上的基于文本和基于内容的图像检索演示(编辑修改稿)内容摘要：

Due to the independence between these approaches, is likely that their bination could improve the performance of a search system by benefiting of both approaches. In the present work, we present an image retrieval system based on a bined search of text and content. II. SYSTEM OVERVIEW The present image search system has been implemented using Java , C++, and PostgreSQL. The set of image object were taken from Flickr web site1 using the SAPIR collection [3]. In the offline phase, the images are downloaded from Flickr using the URL provided by the SAPIR collection. The contentbased descriptors extracted from the images were: Color Histogram 3 3 3 using RGB color space (a 27d vector), Gabor Wavelet (a 48d vector), Efficient Color Descriptor (ECD) 8 1 using RGB color space (a 32d vector), ECD 8 1 using HSV color space (a 32d vector), and Edge Local 4 4 (a 80d vector). The Color Histogram and Gabor Wavelet descriptors were implemented in C++ with the OpenCV library, and the other ones were implemented in Java. The textbased descriptors (title, description, and tags) were extracted from the SAPIR collection. The feature vectors were calculated using the vectorial model and the tfidf weighing [1]. Six feature vectors were created for each image, three for the text using the Porter stemming algorithm [4] and three without stemming. A wordlist, stoplist, and the reverse file for the text features (with and without stemming) were also calculated and stored in the PostgreSQL database. In the online phase, the user enters the query image, a search text, and a weighed distance function for each available feature. The distance functions can be metric (like Euclidean distance) or nonmetric (like DPF and cosine distance). Then, the system performs a kNN search using a weighed bination of distances, normalized by the maximum distance of a feature to the origin. All the textbased and contentbased features (up to 11 vectors for an image) are stored in a PostgreSQL database for efficient retrieval of a small subset, and in a binary file designed for efficient linear scan. Currently, the system contains more than 115,000 images in the collection and the binary file size is about 130 MB. The system contains two different Graphical User I。