نمونه متن انگلیسی مقاله
In this paper, three image features are proposed for image retrieval. In addition, a feature selection technique is also brought forward to select optimal features to not only maximize the detection rate but also simplify the computation of image retrieval. The first and second image features are based on color and texture features, respectively called color co-occurrence matrix (CCM) and difference between pixels of scan pattern (DBPSP) in this paper. The third image feature is based on color distribution, called color histogram for K-mean (CHKM). CCM is the conventional pattern co-occurrence matrix that calculates the probability of the occurrence of same pixel color between each pixel and its adjacent ones in each image, and this probability is considered as the attribute of the image. According to the sequence of motifs of scan patterns, DBPSP calculates the difference between pixels and converts it into the probability of occurrence on the entire image. Each pixel color in an image is then replaced by one color in the common color palette that is most similar to color so as to classify all pixels in image into k-cluster, called the CHKM feature. Difference in image properties and contents indicates that different features are contained. Some images have stronger color and texture features, while others are more sensitive to color and spatial features. Thus, this study integrates CCM, DBPSP, and CHKM to facilitate image retrieval. To enhance image detection rate and simplify computation of image retrieval, sequential forward selection is adopted for feature selection. Besides, based on the image retrieval system (CTCHIRS), a series of analyses and comparisons are performed in our experiment. Three image databases with different properties are used to carry out feature selection. Optimal features are selected from original features to enhance the detection rate.
Many contemporary scholars have been very much devoted to the design of image databases [1–6], as similarity retrieval is important for applications such as medical imaging, office automation, digital library, computer aided design, and multimedia publications. Traditional image retrieval systems are based on the features of the original data [1,3], such as file name, note title, keyword, and indexing icon. When applied to large-scale image databases, these features become troublesome and time-consuming, and even unable to adequately describe image contents. Thus, many feature-based image retrieval systems have been proposed in the academic arena [6–18]. Using a single attribute to describe image features is not enough. Despite the extensive applications of textures [6,8,16], colors [12–15], spatial relations , and shapes  in image retrieval, the results have limited effects on discrimination. When describing image features, the relations between colors and textures are critical. Hence, in this study, colors and textures are employed as attributes in similarity retrieval to develop an innovative and effective image retrieval system (CTCHIRS). Huang and Dai  proposed a texture based image retrieval system which combines the wavelet decomposition  and gradient vector . The system associates a coarse feature descriptor and a fine feature descriptor with each image. Both descriptors are derived from the wavelet coefficients of the original image. The coarse feature descriptor is used at the first stage to quickly screen out non-promising images; the fine feature descriptor is subsequently employed to find the truly matched images. The image retrieval system introduced in Jhanwar et al.  is based on motif co-occurrence matrix (MCM), which converts the difference between pixels into a basic graphic and computes the probability of its occurrence in the adjacent area as an image feature. To obtain color difference between adjacent pixels, we propose a better technique integrated with color co-occurrence matrix (CCM) and difference between the pixels of a scan pattern (DBPSP) to improve texture description. Color histogram  is one of the common techniques used in image retrieval systems. However, when directly used to describe color features, more features have to be recorded. Thus, we propose color histogram for K-mean (CHKM) to clearly describe color features with a smaller number of features. This method is expected to effectively shorten image retrieval time and enhance retrieval performance. CCM, DBPSP, and CHKM are able to effectively describe various properties of an image. To enhance retrieval performance, CCM, DBPSP, and CHKM are integrated to develop an image retrieval system based on texture distribution and color features (CTCHIRS system). The integration of multiple features may certainly reduce retrieval performance. For the highest detection rate and a simplified image retrieval process, we apply a feature selection technique (SFS) in the image retrieval system to shorten the computation time.