In this paper, we propose an active contour with threshold value to detect objects and at the same time get rid of unimportant parts rather than extract all information. The basic ideal of our model is to introduce a ...
详细信息
The aim of this research to develop a real-time, robust, vision based hand-gesture control system capable of operating in different living room environments including low light conditions, and using inexpensive hardwa...
详细信息
ISBN:
(纸本)9781601321541
The aim of this research to develop a real-time, robust, vision based hand-gesture control system capable of operating in different living room environments including low light conditions, and using inexpensive hardware components. The system is designed for remote control of a TV set-top box using a preset gesture vocabulary. A camera is located above a TV screen, across from the user. Gestures may be performed from anywhere within the sensors field of view, without the requirement that the operator obtain visual or auditory feedback. The system is capable of using either a near-IR or a visible color webcam. Hand tracking is based on a modified CAMShift algorithm with face detection and blocking. Gesture classification is based on a limited number of gesture motion trajectory features using a tree-based classifier ensemble, and a modified Longest Common Subsequence (LCS) method. Preliminary testing and evaluation of the systems real time operation is reported.
A novel double-level parallelized firing pulse coupled neural networks (DLPFPCNN) model is presented in this paper, which is used for the segmentation of remotesensingimage with water area as low contrast, low signa...
详细信息
The traditional Mallat's algorithm uses an orthogonal wavelet basis, but the transform is not shift-invariant, and the Mallat's algorithm has a major disadvantage for it contains interpolation and decimation, ...
详细信息
In this paper, the knowledge modeling, architecture design and detailed implementation of an ontology-based knowledge base for target recognition in remotesensingimages is presented. Knowledge base is a critical com...
详细信息
In this paper, the knowledge modeling, architecture design and detailed implementation of an ontology-based knowledge base for target recognition in remotesensingimages is presented. Knowledge base is a critical component of a large ground target recognition system which is a hybrid system that combines knowledge base with remotesensingimageprocessing module. The knowledge base provides necessary knowledge for remotesensingimageprocessing module, and remotesensingimageprocessing module can perform multiple tasks with the support of the knowledge base. The existence of the knowledge base makes the whole recognition system more flexible and more generic than those systems without knowledge base. The effectiveness of the presented knowledge base shows the good prospect of the application of knowledge base in remotesensingimageprocessing domain.
it is becoming study hotspot of academia that using visible 3D expression of face curved surface to recognize face. These issues involved in founding 3D-face recognition system were analyzed such as basic process, sys...
详细信息
An adaptive level set model with feature selection for remotesensingimage segmentation is proposed. The traditional C-V Model based on level set pays much attention to the color features, but with less emphasis on t...
详细信息
An adaptive level set model with feature selection for remotesensingimage segmentation is proposed. The traditional C-V Model based on level set pays much attention to the color features, but with less emphasis on texture features. In the processing of remotesensingimage, sometimes texture feature is more important for the purpose of image segmentation. To solve the problem, this paper firstly takes the components of different color spaces and the texture features as the initial feature set. Then feature selection is performed through local similarity analysis. Meanwhile, the weights of different features are adjusted accordingly. The selected features are utilized in the C-V model as inputs to segment the remotesensingimage. Experimental results on various remotesensingimagery show that the newly proposed approach not only outperforms the traditional model efficiently, but also reduces the time cost greatly.
An adaptive level set model with feature selection for remotesensingimage segmentation is proposed. The traditional C-V Model based on level set pays much attention to the color features, but with less emphasis on t...
详细信息
The objective of this work was to evaluate the potential use of the Multiple Endmember Spectral Mixture Analysis (MESMA) when applied to EO-1 Hyperion hyperspectral data to discriminate land covers in the southern sta...
详细信息
The objective of this work was to evaluate the potential use of the Multiple Endmember Spectral Mixture Analysis (MESMA) when applied to EO-1 Hyperion hyperspectral data to discriminate land covers in the southern state of Rio Grande do Sul, Brazil. The methodology involved: (a) pre-processing and atmospheric correction of Hyperion data;(b) sequential use of the Minimum Noise Fraction (MNF), Pixel Purity Index (PPI) and n-Dimensional Visualizer techniques in the 454-2334 nm range for the initial selection of a general group of endmember candidates (first spectral library) and of another group of pixels to be used for model validation;(c) use of the Visualization and imageprocessing for Environmental Research Tools (VIPER Tools) to perform the final selection of endmembers based on the first spectral library and to obtain MESMA models;and (d) evaluation of resultant fraction images and root mean square error (RMSE) values to determine the optimal number of components of the MESMA model. Results showed that a four-endmember MESMA model (soil = dunes and dry fields;green vegetation = pinus, eucalyptus and grasslands;water = without sediments, with sediments, and with chlorophyll;and shade) adequately described the diversity of the scene components, including that of materials within the same class (e. g., pinus and eucalyptus) and produced the largest fractions and the lowest RMSE values on a per-pixel basis. Results demonstrated the potential use of the MESMA with EO-1 Hyperion hyperspectral data to discriminate land covers in the coastal plains of Rio Grande do Sul, even considering the low signal-to-noise ratio of the instrument, especially in the shortwave infrared range.
Object recognition is challenging problem in computer vision due to appearance variation and presence of visual clutter and occlusions. Recently manifolds are thought to be fundamental for visual perception, and manif...
详细信息
暂无评论