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.
We discuss FPGA implementations of object (such as face) detectors in video streams using the accurate Haar-feature based algorithm. Rather than creating one implementation for one FPGA, we develop a method to generate a series of implementations that have different size and performance to target different FPGA devices. The automatic generation was enabled by custom design space exploration on a particular...
Background subtraction methods are widely exploited for moving objects detection in surveillance video sequences acquired by static camera. Gaussian Mixture Model (GMM), whose convergence speed is rather slow, can be used to model the background of complex scene. This paper adds virtual Gaussian component into GMM and optimizes the updating process of parameters in GMM, in order to increasing the...
This paper proposes an efficient scheme for detecting different object classes in an imaging sensor network. The object detection system detects all the instances of objects (for which the classifier was trained) in the given image, regardless of their scales and locations. Therefore, the image can be thus seen as a set of sub-windows that are to be evaluated by the detector. The detector selects...
In this paper we explore the application of Genetic Programming (GP) to the problem of domain-independent image feature extraction and classification. We propose a new GP-based image classification system that extracts image features autonomously, and compare its performance against a baseline GP-based classifier system that uses human-extracted features. We found that the proposed system has a similar...
Considering the simplicity and fast training speed of Haar-like features, the high detecting precision of HOG features, a combined method is proposed on the basis of the two features. Several rectangular features which can describe local human characteristics based on original features are added. The combined method can retain the precision of HOG features and increase the speed of detection at the...
According to the features of the changing color of fire flames, the periodically increasing of area and the irregular changing in the shape of fire when the early fire breaks, we take fire flame as a kind of special moving object and make a research of it. The color characteristics in fire video images as well as the morphological changes are discussed in this paper, a comprehensive analysis of all...
Background subtraction is commonly used to detect foreground objects in video surveillance. Traditional background subtraction methods are usually based on the assumption that the background is stationary. However, they are not applicable to dynamic background, whose background images change over time. In this paper, we propose an adaptive Local-Patch Gaussian Mixture Model (LPGMM) as the dynamic...
In the present work, automatic corner detection in soccer games based on image features (e.g., object-based features) has been studied. For this purpose, a framework has been proposed that consists of five steps. This paper mainly focuses on the first three steps and specially ball detection step. Ball position on the field plays an important role in determining which event has occurred in the game...
High-resolution satellite imagery is being increasingly employed for large-scale topographic mapping, and especially for updating spatial databases. GeoEye-1 is one of the latest in a series of commercial high-resolution earth observation satellites. The paper presents a method for building detection from GeoEye-1 imagery. The initial building mask was created with supervised classification based...
An algorithm was proposed to remove the moving shadow in video surveillance. This method gets the moving objects from motion detection in first, in which removal shadow on invariable features of colors and local texture, then gets rid of the change suddenly pixels in edge of shadow. Experiments show that having a strong ability to adapt to different scenes, the algorithm retains the integrity of the...
This paper propose a method for automatic vehicle detection in QuickBird images of small highways, with relatively low traffic density, frequent occurrence of tree shadows, and changes in illumination conditions. The vehicle detection is based on an elliptical Laplacian of Gaussian scale space methodology, where the vehicle locations are detected at local extrema in the image response to convolution...
Car license plate detection is a basic task in ITS. A MSER license plate detection method is proposed in the paper constrained by some priori knowledges of license plate. First the input image is pre-processed with grayscaling and gray stretching, etc, and then, the candidate MSER(Maximally Stable Extremal Region) license plate regions are chosen according to the pixel sum, the scale and length-width...
In this paper, we proposed an approach of multi-person movement tracking in office environment without any identity conflicts. Simple image processing with frame differentiation method is applied to identify multiple human motion. An Expert System is applied to predict next camera occurrence of the tracking human. The main objective of this work is to detect and track multi-human motion using single...
Mathematical morphology supplies powerful tools for low level image analysis, with applications in many areas. In this paper, the development of a novel reconfigurable hardware using a genetic algorithm and a pipeline architecture is proposed for the task of shape recognition in binary images. For the recognition process, a large sized convex structuring element representing the object shape to be...
In order to realize infrared small object track detection, the paper analyses the principle of the target tracking detection, the technology combining Mean Shift track algorithm and image processing is proposed. According to the principle, every tache of detection system is analyzed, and then the optic detection model is set up. The detection model of dynamic object is set up. The paper analyses the...
Background Modeling is the normal method on the moving objects detection, it plays a key role in the moving objects detection and tracking, the Gaussian mixture model is one of the most successful methods on the detection. But it converges slowly in the complex scene. This paper proposes a new method named “additive increase” and the “additive decrease” to adjust the weight of the matched distributions...
In this paper, a novel fast logo detection approach in document images is presented. Logos with separated parts usually can affect the logo detection process. To overcome this problem, some specifications of logos are considered. Our proposed method divided in three main sections. In the first section, a horizontal dilation operator is used to merge separated parts of logo in horizontal direction...
This paper presents a novel and robust approach to detect and follow a human with a mobile robotic platform. In order to follow a human, both the initial detection of human and the subsequent tracking need to be implemented. As the robot is initially static, initial human detection is done using a background subtraction technique. To remove the outliers objects, filters are formulated based on the...
Object tracking is a paramount task in video surveillance systems. Although many efforts have been accomplished on object tracking during the last years more work is still needed in order to generate more robust systems. A new fuzzy method for object tracking is presented in this paper. The proposed method is composed of two Sugeno type systems with weighted average memory output functions. One of...
Landmark detection has proven to be a very challenging task in biometrics. In this paper, we address the task of facial component-landmark detection. By “component” we refer to a rectangular subregion of the face, containing an anatomical component (e.g., “eye”). We present a fully-automated system for facial component-landmark detection based on multi-resolution isotropic analysis and adaptive bag-of-words...
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.