This paper describes a 5times5 window-based, power-of-two approximation algorithm for 2D Gaussian smoothing filter and its generic hardware reference design. With a small change in standard deviation sigma, the Gaussi...
详细信息
This paper describes a 5times5 window-based, power-of-two approximation algorithm for 2D Gaussian smoothing filter and its generic hardware reference design. With a small change in standard deviation sigma, the Gaussian smoothing filter is able to provide different levels of smoothing and noise reduction capability, which is highly desirable in the early stage of an image processing flow. By using power-of-two terms, the digital-approximated 2D Gaussian filter can be implemented by simple hardware shifters. A hardware reference design is also proposed to achieve such functionality. Experiments in using the proposed filter with an existing edge detection algorithm show the flexibility and effectiveness of the proposed smoothing mask. The synthesized hardware of such combination shows a 200MHz high operating frequency under UMC 0.18mum technology, which proves the proposed filter is able to achieve the real-time high frame-rate and high-resolution video processing requirements
In recent years, people pay more and more attention to security. Our system presented in this paper is a real-time application to intelligent environment security. It can recognize simple human behaviors and send out ...
详细信息
In recent years, people pay more and more attention to security. Our system presented in this paper is a real-time application to intelligent environment security. It can recognize simple human behaviors and send out alert message intelligently based on human behavior analysis results. This paper mainly describes the behavior analysis methods used in the system, such as moving object detection, human region classification and eigenspace algorithm to recognize human behaviors. Since people usually change their appearances with different dressing, we process images with skeleton algorithm to reduce the impact of appearances. The skeleton structure image sets with all the postures is used to build general eigenspace. Once the general eigenspace is formed, we can recognize behaviors by projecting an unknown human posture into the eigenspace. In our application, six human behaviors (walking, standing, sitting, squatting, leaning and lying) are used. Experimental results show that our method is efficient to recognize these postures
An algorithm using wavelets and Kalman filtering method has been developed for QRS detection. The algorithm takes following steps: i) We use a RLS adaptive algorithm to establish a model of the electrocardiogram. Thus...
详细信息
An algorithm using wavelets and Kalman filtering method has been developed for QRS detection. The algorithm takes following steps: i) We use a RLS adaptive algorithm to establish a model of the electrocardiogram. Thus the algorithm can change its own model according to the change of the electrocardiogram, ii) Based on the model established in stage i), we combined the wavelet and Kalman algorithm, decomposed the signal to several scales, filtering the signal in each scale. And then, iii) A decision rule is used to determine the QRS complex. The method was tested on ECG data form the MIT database, and experimental results show that the method is effective, and the influence of the noise can be eliminated
In this paper, we describe a single antenna time-reversal based detection scheme that can be used to enhance the performance of a radar system in a multipath rich scattering environment. Using time-reversal, the respo...
详细信息
In this paper, we describe a single antenna time-reversal based detection scheme that can be used to enhance the performance of a radar system in a multipath rich scattering environment. Using time-reversal, the response from a cluttered medium is first nulled. When the target enters into the medium, the electromagnetic energy focuses around the target so that a stronger echo is obtained. The experimental results show that depending on the amount of clutter in the medium, the performance of a radar detection scheme can be improved up to 4.8 dB using time-reversal methods with a single antenna
In application fields of self-direction of vehicle, unmanned drive, trace robot and trace finder, tracking and identification of trace signal have key effects on tracking reliability and accuracy. Trace characterized ...
详细信息
In application fields of self-direction of vehicle, unmanned drive, trace robot and trace finder, tracking and identification of trace signal have key effects on tracking reliability and accuracy. Trace characterized with magnetic induction intensity is actually electromagnetic field signal produced around conductor when current flows inside conductor, and magnetic intensity has direct relationship with distance from conductor. Trace detecting device on vehicle can detect magnetic induction intensity and its change trend, and then degree and speed of deviation from trace can be calculated by linear regression algorithm, thus, trace tracking and identification can be achieved. Trace characterized with magnetic information can adapt to many complex working environments, and intensity control of trace signal and fault diagnosis can be easily realized
Determining the correspondence of image patches is one of the most important problems in Computer Vision. When the intensity space is variant due to several factors such as the camera gain or gamma correction, one nee...
详细信息
Determining the correspondence of image patches is one of the most important problems in Computer Vision. When the intensity space is variant due to several factors such as the camera gain or gamma correction, one needs methods that are robust to such transformations. While the most common assumption is that of a linear transformation, a more general assumption is that the change is monotonic. Therefore, methods have been developed previously that work on the rankings between different pixels as opposed to the intensities themselves. In this paper, we develop a new matching method that improves upon existing methods by using a combination of intensity and rank information. The method considers the difference in the intensities of the changed pixels in order to achieve greater robustness to Gaussian noise. Furthermore, only uncorrelated order changes are considered, which makes the method robust to changes in a single or a few pixels. These properties make the algorithm quite robust to different types of noise and other artifacts such as camera shake or image compression. Experiments illustrate the potential of the approach in several different applications such as changedetection and feature matching.
Motion segmentation is important in many computer vision application, which aims to detect motion regions such as moving vehicles and people in natural scenes. Detecting moving blobs provides a focus of attention for ...
详细信息
Motion segmentation is important in many computer vision application, which aims to detect motion regions such as moving vehicles and people in natural scenes. Detecting moving blobs provides a focus of attention for later processes such as tracking and activity analysis. However, changes from weather, illumination, shadow and repetitive motion from clutter make motion segmentation difficult to process quickly and reliably. In this paper we proposed a method using minimum graph cuts based on hybrid algorithm approach for moving object segmentation. Experiments are carried out to examine the efficiency of the proposed approach
Sequential detection allows the analysis of an incoming data flow and the detection of changes in the distribution of these measurements. In this paper, we develop the sequential detection algorithm for a target under...
详细信息
Sequential detection allows the analysis of an incoming data flow and the detection of changes in the distribution of these measurements. In this paper, we develop the sequential detection algorithm for a target under compound-Gaussian clutter. Both the target and clutter parameters are assumed unknown. We first derive estimates for these parameters, then discuss the sequential detection algorithm for two cases: target parameter is known and unknown. We consider detections for both the target appearance and disappearance. We examine the relationship between several performance measurements for the sequential detector, including the false-alarm rate and the average detection delay. In the numerical example part, we first illustrate the performance of our algorithms. Then we present an example of the optimal polarimetry design in the sequential detection.
暂无评论