A new real-time algorithm is proposed in this paperfor detecting moving object in color image sequencestaken from stationary cameras.This algorithm combines a temporal difference with an adaptive background subtractio...A new real-time algorithm is proposed in this paperfor detecting moving object in color image sequencestaken from stationary cameras.This algorithm combines a temporal difference with an adaptive background subtraction where the combination is novel.Ⅷ1en changes OCCUr.the background is automatically adapted to suit the new conditions.Forthe background model,a new model is proposed with each frame decomposed into regions and the model is based not only upon single pixel but also on the characteristic of a region.The hybrid presentationincludes a model for single pixel information and a model for the pixel’s neighboring area information.This new model of background can both improve the accuracy of segmentation due to that spatialinformation is taken into account and salientl5r speed up the processing procedure because porlion of neighboring pixel call be selected into modeling.The algorithm was successfully used in a video surveillance systern and the experiment result showsit call obtain a clearer foreground than the singleframe difference or background subtraction method.展开更多
Exactly capturing three dimensional (3D) motion i nf ormation of an object is an essential and important task in computer vision, and is also one of the most difficult problems. In this paper, a binocular vision s yst...Exactly capturing three dimensional (3D) motion i nf ormation of an object is an essential and important task in computer vision, and is also one of the most difficult problems. In this paper, a binocular vision s ystem and a method for determining 3D motion parameters of an object from binocu lar sequence images are introduced. The main steps include camera calibration, t he matching of motion and stereo images, 3D feature point correspondences and re solving the motion parameters. Finally, the experimental results of acquiring th e motion parameters of the objects with uniform velocity and acceleration in the straight line based on the real binocular sequence images by the mentioned meth od are presented.展开更多
Image sequences processing and video encoding are extremely time consuming problems. The time complexity of them depends on image contents. This paper presents an estimation of a block motion method for video coding w...Image sequences processing and video encoding are extremely time consuming problems. The time complexity of them depends on image contents. This paper presents an estimation of a block motion method for video coding with edge alignment. This method uses blocks of size 4 × 4 and its basic idea is to find motion vector using the edge position in each video coding block. The method finds the motion vectors more accurately and faster than any known classical method that calculates all the possibilities. Our presented algorithm is compared with known classical algorithms using the evaluation function of the peak signal-to-noise ratio. For comparison of the methods we are using parameters such as time, CPU usage, and size of compressed data. The comparison is made on benchmark data in color format YUV. Results of our proposed method are comparable and in some cases better than results of standard classical algorithms.展开更多
In this paper, evacuation experiments are carried out to study pedestrian movement behaviors in building bottleneck. An image processing method based on mean-shift algorithm is used to extract pedestrian movement traj...In this paper, evacuation experiments are carried out to study pedestrian movement behaviors in building bottleneck. An image processing method based on mean-shift algorithm is used to extract pedestrian movement trajectory. Based on the extracted trajectory, we analyze the microscopic movement characteristics of pedestrians such as lane formation, change of velocity and distance between two sequential pedestrians. A pedestrian lane is a group of pedestrians moving in a column. The lane formation is verified by the pedestrian trajectory and distribution of pedestrian’s lateral positions (x direction in the paper): lane number changes from one to two, three or even more with the increasing bottleneck width when pedestrians pass through the bottleneck. By analyzing the pedestrian movement behaviors in the same pedestrian lane, we find three typical movement modes in the bottleneck: time-lag acceleration, synchronous acceleration, and avoiding deceleration. Through analyzing the time intervals when successive pedestrians pass through the bottleneck, we find that most pedestrians adjust their velocities according to the distance to the forward pedestrians. Results also indicate that due to different cultures, pedestrians flux in China and Germany may have some differences besides their similarities.展开更多
基金National Natural Science Foundation Grant No.60072029
文摘A new real-time algorithm is proposed in this paperfor detecting moving object in color image sequencestaken from stationary cameras.This algorithm combines a temporal difference with an adaptive background subtraction where the combination is novel.Ⅷ1en changes OCCUr.the background is automatically adapted to suit the new conditions.Forthe background model,a new model is proposed with each frame decomposed into regions and the model is based not only upon single pixel but also on the characteristic of a region.The hybrid presentationincludes a model for single pixel information and a model for the pixel’s neighboring area information.This new model of background can both improve the accuracy of segmentation due to that spatialinformation is taken into account and salientl5r speed up the processing procedure because porlion of neighboring pixel call be selected into modeling.The algorithm was successfully used in a video surveillance systern and the experiment result showsit call obtain a clearer foreground than the singleframe difference or background subtraction method.
文摘Exactly capturing three dimensional (3D) motion i nf ormation of an object is an essential and important task in computer vision, and is also one of the most difficult problems. In this paper, a binocular vision s ystem and a method for determining 3D motion parameters of an object from binocu lar sequence images are introduced. The main steps include camera calibration, t he matching of motion and stereo images, 3D feature point correspondences and re solving the motion parameters. Finally, the experimental results of acquiring th e motion parameters of the objects with uniform velocity and acceleration in the straight line based on the real binocular sequence images by the mentioned meth od are presented.
文摘Image sequences processing and video encoding are extremely time consuming problems. The time complexity of them depends on image contents. This paper presents an estimation of a block motion method for video coding with edge alignment. This method uses blocks of size 4 × 4 and its basic idea is to find motion vector using the edge position in each video coding block. The method finds the motion vectors more accurately and faster than any known classical method that calculates all the possibilities. Our presented algorithm is compared with known classical algorithms using the evaluation function of the peak signal-to-noise ratio. For comparison of the methods we are using parameters such as time, CPU usage, and size of compressed data. The comparison is made on benchmark data in color format YUV. Results of our proposed method are comparable and in some cases better than results of standard classical algorithms.
基金supported by the National Natural Science Foundation of China (Grant No. 91024025)the NCET Project (Grant No. 08-0518)
文摘In this paper, evacuation experiments are carried out to study pedestrian movement behaviors in building bottleneck. An image processing method based on mean-shift algorithm is used to extract pedestrian movement trajectory. Based on the extracted trajectory, we analyze the microscopic movement characteristics of pedestrians such as lane formation, change of velocity and distance between two sequential pedestrians. A pedestrian lane is a group of pedestrians moving in a column. The lane formation is verified by the pedestrian trajectory and distribution of pedestrian’s lateral positions (x direction in the paper): lane number changes from one to two, three or even more with the increasing bottleneck width when pedestrians pass through the bottleneck. By analyzing the pedestrian movement behaviors in the same pedestrian lane, we find three typical movement modes in the bottleneck: time-lag acceleration, synchronous acceleration, and avoiding deceleration. Through analyzing the time intervals when successive pedestrians pass through the bottleneck, we find that most pedestrians adjust their velocities according to the distance to the forward pedestrians. Results also indicate that due to different cultures, pedestrians flux in China and Germany may have some differences besides their similarities.
