An adaptive threshold segmentation algorithm based on HS joint statistics using HSI color model is introduced in order to improve the performance of image segmentation and the robustness of object recognition to illum...
详细信息
An adaptive threshold segmentation algorithm based on HS joint statistics using HSI color model is introduced in order to improve the performance of image segmentation and the robustness of object recognition to illumination change. Statistics of H and S component value of all the pixels in the region of interest (ROI) is obtained after image segmentation in HSI color space. The changing rule of statistics is verified while an empirical scale factor is applied to reflect the difference between the influence of H value and S value on segmentation threshold. The threshold is thus adjusted online according to the proposed approach. Experimental results demonstrate the decrease trend of H and S component value with intensified light. Furthermore, it validates the robustness of the approach under the circumstance of changing illumination. Thus the accuracy of the vision-based robot grasping is significantly improved.
A class of complex hybrid time-delay models for network controlsystems have been constructed by taking the characteristics of network into account. Based on a LMI(linear matrix inequality) method, a kind of Lyapunov-...
详细信息
A class of complex hybrid time-delay models for network controlsystems have been constructed by taking the characteristics of network into account. Based on a LMI(linear matrix inequality) method, a kind of Lyapunov-Krasovskii functional is employed, and the controller design is discussed for this type of systems. As input control is in discrete-time and the input delay constant is known, while state is in continuous-time and the state delay constant is not known exactly, so the system is constructed based on two models in both continuous-time domain and discrete-time domain. At the same time, a new type of adaptive control strategy for the unknown delay parameter is proposed in this paper, which realizes that the unknown delay parameter can always be reflected in the memory state-feedback controller. Delay-dependent sufficient conditions for the existence of the feedback controller in terms of LMIS are obtained.
The classical mean shift tracking algorithm is apt to make errors or lose the target if the target is occluded for a very long time. Thus an improved mean shift tracking algorithm is proposed. This algorithm divides t...
详细信息
The classical mean shift tracking algorithm is apt to make errors or lose the target if the target is occluded for a very long time. Thus an improved mean shift tracking algorithm is proposed. This algorithm divides the target into multiple fragments and integrates spatial information by using different weights of each image fragment. The similarity coefficient between target template and candidate template consists of the Bhattacharyya coefficients of the corresponding multiple fragments. Experimental results show that the proposed method is efficient when the target is occluded for a long time. A new method named edge-histogram is used. This method is based on original scale updating mechanism and make a further judgment that whether the target is smaller or not by calculating the Bhattacharyya coefficient between the target's edge-histograms of the current frame and the previous one. Experimental results show that the proposed algorithm can deal with the scale problem very well.
A conceptual map model of environments is proposed for task planning of service robots in semantic knowledge space. Such model is characterized with a layered structure containing ontology, spatial and user knowledge....
详细信息
A conceptual map model of environments is proposed for task planning of service robots in semantic knowledge space. Such model is characterized with a layered structure containing ontology, spatial and user knowledge. It is capable of inferring environmental knowledge and mapping semantic plans onto the instantial perception space. A scene recognition based method for conceptual map building and updating is proposed to support the environment learning capabilities of robots. Using such model, a layered topological and metric navigation strategy is proposed, in which map switching technique ensures that Monte Carlo localization and metric path planning are performed within each small-scale grid map. Experimental results in large-scale office environments validate the effectiveness and efficiency of robot localization and navigation, and a semantic navigation manner is achieved.
In order to achieve the goal of data statistic and task management under the embedded systems, together with the consideration of processing speed and the internal storage capacity, the embedded database is introduced...
详细信息
In order to achieve the goal of data statistic and task management under the embedded systems, together with the consideration of processing speed and the internal storage capacity, the embedded database is introduced. The features of embedded database are small in size, multi-functional, portable, highly efficient and stable. Taking the data management in the home energy monitor system as an example, a statistical method of time sharing is proposed. The embedded database Berkeley DB is transplanted, and the resource occupancy and response speed before and after using Berkeley DB are compared. Testing data show that although taking up more storage resources, the use of Berkeley DB provides more efficient data management capabilities. The application of Berkeley DB improves the system response, and provides a better human-computer interaction. Under the allowance of the hardware resources of device, this practice of using space for performance is feasible.
When it comes to the issue of path planning for mobile service robots in structured indoor environments, traditional A* algorithm fails to display a good performance of real-time path tracking. Therefore, a new method...
详细信息
When it comes to the issue of path planning for mobile service robots in structured indoor environments, traditional A* algorithm fails to display a good performance of real-time path tracking. Therefore, a new method is proposed to design a near-optimal smooth path. Mobile robots can thus track the path smoothly and reach the goal fast. At first, the improved A* algorithm is applied to plan a proper trajectory for the mobile robot. Then a key-point optimization process is applied to make the global path simplified to be connections among a series of key points. The most important improvement lies in the introduction of the polar polynomials curve to smooth the path according to the constraints of position, curvature and slope. However, when the arc turns become 90° or more, the optimized piecewise-polynomial-function curve is generated to replace the original one. Experimental results prove that the path generated meets the dynamic characteristics of the mobile robots as well as the geometric features. The requirements of quick real-time computation of path tracking process can be well satisfied.
In order to solve the problems of inaccurate positioning in current vehicle monitoring systems, an intelligent vehicle monitoring system based on mobile video and WebGIS (Web geographic information system) is proposed...
详细信息
In order to solve the problems of inaccurate positioning in current vehicle monitoring systems, an intelligent vehicle monitoring system based on mobile video and WebGIS (Web geographic information system) is proposed and designed. The system is composed of front-end equipment, system platform and client. The instant image information captured by vehicle mobile video and the GPS(global positioning system) positioning location information received by the terminal are transmitted from front-end to system platform via the GPRS(general packet radio service) mobile communications platform. Also, a highly compressed digital video and audio codec technology standard named H.264/AVC is employed to compress the real-time mobile video information and the Google Maps API is called to display the current position of vehicles on the electronic map. Furthermore, the spatial analysis capabilities of GIS plus the GPS information not only make tracking feasibly, but also provide aid in decision making. It is shown that the system can not only reproduce the video data timely and accurately, but also accurately position the mobile vehicles so as to provide a better service in monitoring, control, management, analysis, decision-making and command functions, which expands the application fields and development prospects of WebGIS and mobile video monitoring technology.
The memory state feedback control problem for a class of discrete-time systems with input delay and unknown state delay is addressed based on LMIs and Lyapunov-Krasovskii functional method. Under the action of our des...
详细信息
The memory state feedback control problem for a class of discrete-time systems with input delay and unknown state delay is addressed based on LMIs and Lyapunov-Krasovskii functional method. Under the action of our designed adaptive control law, the unknown time-delay parameter is included in memory state feedback controller. Using LMI technique, delay-dependent sufficient conditions for the existence of the feedback controller are obtained. Finally, the effectiveness of the proposed design method is demonstrated by a numerical example.
In order to improve the face recognition performance using Gabor wavelet, a multi-channel dimension reduction scheme (MDRS) employing PCA algorithm is proposed. In addition, the selection of scales and orientations in...
详细信息
Object tracking algorithm using modified Particle filter in low frame rate(LFR) video is proposed in this paper,which the object moving significantly and randomly between consecutive frames in the low frame rate ***,P...
详细信息
Object tracking algorithm using modified Particle filter in low frame rate(LFR) video is proposed in this paper,which the object moving significantly and randomly between consecutive frames in the low frame rate ***,Particle filtering use motion transitions to model the movement of the ***,in object tracking with low frame rate sequences,it is very difficult to model significant random jumps of *** key notion of our solution is that using the object detection and extraction to locate the tracked object,while not using the dynamical *** propagate the sample set around the detected regions,which the samples are assumed to be uniformly distributed in the neighborhoods of the detected *** is similar to the general particle filter to propagate *** we compute the likelihood between the target model and the candidate regions,which are based on color histogram *** extensive experiments show that the proposed algorithm performs robustly in a large variety of tracking scenarios.
暂无评论