Vision-based automatic vehicle counting system using motion estimation with Taylor series approximation

Abstract

Automatic tracking vehicle in urban traffic video surveillance is a challenging problem in computer vision. Although many issues have been solved, some are still unsolved, such as video surveillance problem of complex traffic intersection in congested condition. In this paper, we develop a vehicle counting system using a motion estimation with Taylor series approximation with embedded virtual entering and exiting boxes. The result shows that the system provides the counting success rate as high as 100% and the lowest counting rate is 14.29%. The mistakes are from the wrong direction prediction because of the very complex traffic condition.

Vehicle Classification using IPCP and EsKNN Algorithm for Surveillance Camera

“…Even though unsupervised clustering method is used for extracting path of trajectories, this method unable to distinguish pedestrian and vehicles. The other method of vehicles detection using motion estimation based detector was introduced by [5]. They develop vehicles detection and counting based upon Taylor series approximation with embedded virtual entering and virtual bboxes.…”

“…They develop vehicles detection and counting based upon Taylor series approximation with embedded virtual entering and virtual bboxes. Proposed method [5] uses optical flow and block matching based on vehicle detection and tracking. This object's features were extracted based on the difference of motion pixels.…”

Vehicle Counting Based on Vehicle Detection and Tracking from Aerial Videos

“…Srijongkon [ 12 ] proposed a vehicle counting system based on ARM/FPGA processor, which uses adaptive background subtraction and shadow elimination to detect moving vehicle and then counts the vehicles in video screen. Prommool [ 13 ] introduced a vehicle counting framework using motion estimation (block matching and optical flow combination). In [ 13 ], the area box is set at the intersection to determine whether the vehicle passes through.…”

“…Prommool [ 13 ] introduced a vehicle counting framework using motion estimation (block matching and optical flow combination). In [ 13 ], the area box is set at the intersection to determine whether the vehicle passes through. Swamy [ 14 ] presented a vehicle detection and counting system based on color space model, which uses color distortion and brightness distortion of image to detect vehicle and then counts vehicle using a pre-defined line.…”

“…Although researchers improve the traditional traffic monitoring methods and apply vehicle detection and tracking on vehicle counting, the traditional surveillance camera still cannot be applied for monitoring large areas, which is very restrictive for vehicle counting. Besides, the research in [ 12 , 13 , 14 , 15 ] does not use a multi-object management system to confirm the uniqueness of vehicles, which is unreliable for long sequence vehicle counts and effects the efficiency of vehicle counting. In contrast to traditional traffic monitoring sensors, the UAV can be flexibly deployed to the regions that need to be monitored.…”