The Infona portal uses cookies, i.e. strings of text saved by a browser on the user's device. The portal can access those files and use them to remember the user's data, such as their chosen settings (screen view, interface language, etc.), or their login data. By using the Infona portal the user accepts automatic saving and using this information for portal operation purposes. More information on the subject can be found in the Privacy Policy and Terms of Service. By closing this window the user confirms that they have read the information on cookie usage, and they accept the privacy policy and the way cookies are used by the portal. You can change the cookie settings in your browser.
A 671G0PS/W Semantic Analysis SoC (SASoC) is implemented in 90 nm CMOS technology. Two stream processing systems are integrated with a power-aware frequency scaling technique to simultaneously accelerate video processing and machine-learning algorithms. The input data rate reaches 76.8 Gpixel/s for video processing and 51.2 Gdimension/s for machine-learning algorithms.
Contact burning is one of the main faults of transformer on-load tap changers (OLTC). If contact burning fault diagnosis is not timely, the decreasing contact area, increasing contact resistance and the increasing arc and heat during the witching processes will cause further burning. Researches on contact burning failure were done and a feature analysis method based on wavelet packet energy entropy...
In this paper, we propose a new method to detect liver tumors in CT images automatically. The proposed method is composed of two steps. In the first step, tumor candidates are extracted by EM/MPM algorithm; which is used to cluster liver tissue. To cluster a dataset, EM/MPM algorithm exploits both intensity of voxels and labels of the neighboring voxels. It increases the accuracy of detection, with...
Recently, scene recognition is becoming an additional function in digital camera. Automatic scene understanding is a highest-level operation in computer vision, and it is a very difficult and largely unsolved problem. The conventional methods usually use global features (such as color histogram, texture, edge) for image representation and recognize scene types with some classifiers (such as Bayesian,...
The paradigm of Content-based retrieval (CBR) with query-by-example (QBE) has been prevailing in multimedia applications and systems. While the CRB/QBE approach is easy to use from the end-user's perspective, it is not always possible for a user to prepare/have a suitable "example" in hand; besides, the lack of semantics inherent to this approach may cause strange things to happen (e.g.,...
A new photo retrieval system for mobile devices is proposed in this paper. The system can be used to search for photos with similar spatial layout effectively and efficiently, and it adopts a new algorithm that extracts features of image regions based on hardware K-Means clustering. Since K-Means is computationally intensive for real-time applications in embedded systems, it is necessary to accelerate...
The hue component of face area in coal mine surveillance images is not true of normal school because of nonlinear characteristic of image sample device in dim-lightening surroundings, and its contrast is clearer than that of non-face area. The skin-color of miner's faces surveillance images of multi angles can be segmented fleetly with the segmentation parameter of hue component which is initialized...
Glycosylation is one of the common post-translation modification of protein in eukaryotic cells. Conventional neural network methods have been applied to predict glycosylation sites in protein sequence and the prediction accuracy is dependent on the dimension of feature vector (length of protein sequence). Though the prediction accuracy can be improved by increasing the length of protein sequence,...
Image category recognition is important to access visual information on the level of objects and scene types. This paper combines different feature representations of images and learn a compact subspace of different features for the automatic recognition of object and scene classes. Compact visual-words and low-level-features object class subspaces are automatically learned from a set of training...
In this paper, a new feature selection method with applications to handwritten digit recognition is proposed. This method is based on recursive feature elimination (RFE) in least squares support vector machines (LS-SVM). Digit recognition is achieved by one-against-all LS-SVMs. The RFE method is adapted to multi-class classification in two ways. One is to prune features for each binary LS-SVM classifier...
A fast and efficient approach to color image segmentation and texture feature extraction is developed. In the proposed image segmentation algorithm, a new quantization technique for HSV color space is implemented to generate a color histogram and a gray histogram for K-Means clustering, which operates across different dimensions in HSV color space. Then a texture feature extraction method for content-based...
Image warping and morphing are important visual effect tools in entertainment industry and other research fields. We developed a prototypical automatic facial image manipulation system (AFIM) for face morphing and shape normalization (warping). In our AFIM system, there are two main functions: (1) warping a facial image to a target image (face shape normalization), (2) generation of inter- or intra-personal...
Face recognition is an important technology for managing security, human interface, and so on. Though face recognition techniques have been used in many fields, it still has several problems to be solved, such as robustness to variations of pose, illumination and expression. In this paper, we focus our research on pose-robustness and propose a pose-robust face recognition method based on 3D reconstruction...
Statistical shape model (SSM) is to model the shape variation of an object. The statistical shape models are constructed by analysis of the positions of a set of landmark points based and use the surface information. In this paper, we propose a new PCA based statistical shape modeling technique and its application to medical applications. In the proposed method, boundary points of each slice are used...
In this paper, we proposed an automatic facial caricaturing system based on multi-view Active Shape Model (ASM). The ASM is used for automatic extraction of feature points. But it is difficult to apply the conventional ASM to facial images with pose variations. We developed a multi-view ASM bank, which includes various ASMs corresponding to each pose. The multi-view ASMs can extract feature points...
Image category recognition is important to access visual information on the level of objects and scene types. This paper presents an automatic recognition system of scene and object with PCA-SICEF feature for digital color images. SICEF (scale-invariant color and edge feature) is an extension of the conventional local SIFT (scale-invariant feature transform) feature, which only include edge invariance...
Chaotic time series analysis or forecasting is an important and complex problem in machine learning. As an effective tool, support vector machine (SVM) has been broadly adopted in pattern recognition and machine learning fields. In developing a successful SVM classifier, eliminating noise and extracting feature are very important. This paper proposes the application of kernel PCA to LS-SVM for feature...
Exploration techniques of video knowledge have been proposed for years to help people discover the details about videos. However, existing systems still yield limited information for users. In this paper, we present a video knowledge browsing system, which can establish the framework of a video based on its summarized contents and expand them by using online correlated media. Thus, users can not only...
This paper describes a fast and efficient texture feature extraction method based on image segmentation in HSV color space, which is especially useful in resource-limited embedded systems. In this approach, a query image is first segmented based on its color feature, and the proposed technique is applied on the labeled image to efficiently extract texture feature. Compared with the feature extraction...
Duplicated web pages responded by search engines not only waste valuable storage, but also aggravate burdens of userspsila browse. Web page de-duplication can effectively improve the information retrieval. This paper proposes pretreatment of web pages to improve the effectiveness and efficiency of web page de-duplication based on feature code according to the principle of data clearing. This paper...
Set the date range to filter the displayed results. You can set a starting date, ending date or both. You can enter the dates manually or choose them from the calendar.