Face detection technology has widely attracted attention due to its enormous application value and market potential, such as face recognition and video surveillance system. Real-time face detection not only is one par...
详细信息
Face detection technology has widely attracted attention due to its enormous application value and market potential, such as face recognition and video surveillance system. Real-time face detection not only is one part of the automatic face recognition system but also is developing an independent research subject. So, there are many approaches to solve face detection. Here the modified adaboost algorithm based on OpenCV is presented, and experiments of real-time face detecting are also given through two methods of timer and dual-thread. The result shows that the method of face detection with dual-thread is simpler, smoother and more precise.
Traditional gender recognition technologies with single feature cannot express an image completely and are limited by their recognition speed and accuracy. In this paper, we explored a way of fulfilling this task by c...
详细信息
ISBN:
(纸本)9781467309653
Traditional gender recognition technologies with single feature cannot express an image completely and are limited by their recognition speed and accuracy. In this paper, we explored a way of fulfilling this task by combing the characteristics of both Haar-like and textural feature and proposed the approach to construct a multiple scale textural feature (MST), meanwhile, in order to achieve high recognition accuracy, we further improved the adaboost algorithm improved by Freidman et al. Data from experiments based on MIT database show that our MST feature working together with the improved adaboost algorithm can obtain a recognition rate of 86%.
This work proposes an intelligent and interactive virtual fence system for security management in a danger- or accident-prone area. The proposed system is designed to be realized in an outdoor and wide-area park. With...
详细信息
ISBN:
(纸本)9781467330848
This work proposes an intelligent and interactive virtual fence system for security management in a danger- or accident-prone area. The proposed system is designed to be realized in an outdoor and wide-area park. With the proposed system, users can visually and arbitrarily define the polygonic alert area for the sake of security. The Gaussian Mixture Model (GMM) and adaboost algorithms are modified and integrated in this work to provide a robust human detection. As soon as the invading people is detected and located in the interior of the alert area, the proposed system will immediately alarm the security guards by MMS of cellular networks or a customized App installed in smart phones. Besides, this work also proposes a sliding-window based alarm mechanism to practically and robustly improve the detection rate and reduce the false alarm rate of detection. The experimental results will demonstrate that the proposed system can be indeed applied in a real field and provide a satisfied performance.
The paper designs a number gesture recognition system based on Kinect. The system uses the depth image captured by Kinect sensor to extract gesture information for recognition. In the process of recognition, the AdaBo...
详细信息
The paper designs a number gesture recognition system based on Kinect. The system uses the depth image captured by Kinect sensor to extract gesture information for recognition. In the process of recognition, the adaboost algorithm in machine learning is adopted firstly, by the help of Visual Gesture Builder to train the number gesture classifier,and export the classifiers to the recognition program to recognize the current gesture. Finally, numbers are outputted according to the recognized gesture, and the basic functions of number gesture recognition system are realized. Through the experimental verification, the recognition system has enough accuracy, and the system is suitable for various factors such as environment, personnel, etc., which has a certain practical value.
A cone obstacle detection method is proposed to detect the cone obstacle. The proposed method uses the Haar features trained by the adaboost algorithm. After training, the cascaded classifiers are produced and used in...
详细信息
ISBN:
(纸本)9781467372220
A cone obstacle detection method is proposed to detect the cone obstacle. The proposed method uses the Haar features trained by the adaboost algorithm. After training, the cascaded classifiers are produced and used in a wheeled mobile robot to detect the cone obstacles. To apply the cone obstacle detection, the wheeled mobile robot uses the proposed cone obstacle detection algorithm to simulate children playing the cone maze. The wheeled mobile robot creates a twisty path to move through and around the cones to go to the end point. Experimental results show that the proposed method is effective to detect the cone obstacles. Furthermore, the wheeled mobile robot can create a twisty path to move through and around the cones to go to the end point.
Face detection methods based on mixed features and adaboost algorithm are presented. A method combining local Walsh features and global PCA features is proposed, and classifiers based on PCA features are ensemble to i...
详细信息
Face detection methods based on mixed features and adaboost algorithm are presented. A method combining local Walsh features and global PCA features is proposed, and classifiers based on PCA features are ensemble to improve the performance. Experiments have been done on MIT+CUM dataset.
The eyes are important facial landmarks, both for image normalization due to their relatively constant interocular distance, and for post processing due to their anchoring model-based *** this paper a hierarchical cla...
详细信息
The eyes are important facial landmarks, both for image normalization due to their relatively constant interocular distance, and for post processing due to their anchoring model-based *** this paper a hierarchical classifier for precise eye location algorithm is *** new algorithm involves three main ***, a classifier based on adaboost algorithm is introduced to locate eye pair *** order to get a reliable location, simple eye location is also introduced for checking the located ***, a precise eye location method based on Minimum Extreme Region (MER) is employed to detect eye ***, a combine processing based on the above two results is used for the last *** algorithm is tested on database CAS-PEAL, JAFFE and *** results prove the merit of this approach.
This paper proposes an automatic recognition method for losing of train bogie center plate bolt (TBCPB),which initially locates the bolt region based on gray projection and spatial filter,and then acquires the accurat...
详细信息
ISBN:
(纸本)9781467321006
This paper proposes an automatic recognition method for losing of train bogie center plate bolt (TBCPB),which initially locates the bolt region based on gray projection and spatial filter,and then acquires the accurate position of bolt region combined with its gradient feature,finally extracts the Haar-like features of bolt region and designs a classifier based on the adaboost *** of experiments demonstrate that the method compresses noise effectively and overcomes the complexity of various *** is also a robust method to images with poor quality,such as blur,poor illumination and excess *** average fault recognition accuracy is 95.9%.
Change detection from remote sensing image is a hot topic in the field of image analysis. This paper describes an object-based change detection method from high resolution image. To obtain meaningful objects, a segmen...
详细信息
Change detection from remote sensing image is a hot topic in the field of image analysis. This paper describes an object-based change detection method from high resolution image. To obtain meaningful objects, a segmentation method based on optimal segmentation scale is discussed, and then by combing the NDVI and band ratio to obtain the difference image, the change area is detected by adaboost algorithm finally. The efficiency of the method has been assessed by using a pair of bitemporal QuickB ird images.
3D reconstruction from the microscopy images of serial sections plays an important role in analyzing structure of biological specimens, such as neuronal circuits in brain tissue. This paper is focusing on the 3D recon...
详细信息
ISBN:
(纸本)9781509023974
3D reconstruction from the microscopy images of serial sections plays an important role in analyzing structure of biological specimens, such as neuronal circuits in brain tissue. This paper is focusing on the 3D reconstruction of synapses which is a branch of micro reconstruction of brain. Having analysed the structure of synapses, we first detect and locate them in serial sections with cascade adaboost algorithm, then optimize the shape of synapses by constructing suitable fitting functions and segment them with morphology processing method. Taking the connection of adjacent sections and synaptic extensibility in space into consideration, we complete the reconstruction of synapses which is based on the segmentation in serial sections. Actually all our works are based on the registration of serial sections.
暂无评论