A robust multimedia surveillance system for people counting

Abstract

Closed circuit television cameras (CCTV) are widely used in monitoring. This paper presents an intelligent CCTV crowd counting system based on two algorithms that estimate the density of each pixel in each frame and use it as a basis for counting people. One algorithm uses scale-invariant feature transform (SIFT) features and clustering to represent pixels of frames (SIFT algorithm) and the other uses features from accelerated segment test (FAST) corner points with SIFT features (SIFT-FAST algorithm). Each algorithm is designed using a novel combination of pixel-wise, motion-region, grid map, background segmentation using Gaussian mixture model (GMM) and edge detection. A fusion technique is proposed and used to validate the accuracy by combining the result of the algorithms at frame level. The proposed system is more practical than the state of the art regression methods because it is trained with a small number of frames so it is relatively easy to deploy. In addition, it reduces the training error, setup time, cost and open the door to develop more accurate people detection methods. The University of California (UCSD) and Mall datasets have been used to test the proposed algorithms. The mean deviation error, mean squared error and the mean absolute error of the proposed system are less than 0.1, 16.5 and 3.1, respectively, for the Mall dataset and less than 0.07, 5.5 and 1.9, respectively, for UCSD dataset.

Energy Efficient Camera Solution for Video Surveillance

“…The depth images were captured using single sensor mounted in zenithal position in both indoor and outdoor environments. SIFT-FAST algorithm for counting people in a crowded environment was proposed by [20]. They used traditional CCTV video sequences for detection purposes.…”

Development of a Sensor to Measure Physician Consultation Times

“…Sensors that count and track people are commercially available and deliver a wide range of applications in airports, museums, libraries, or any other type of (public) buildings. Possible applications are crowd management, transport systems, heating as well as air conditioning, and optimisation of evacuation planning [3]. Significant user groups are retailers and shopping centres.…”

Image Processing Based Ambient Context-Aware People Detection and Counting

“…In the ROI methods, people in a specific region are counted [24] whereas, in the LOI methods, people who cross a real or virtual line are counted [16]. These counters can also be classified into four categories: features regression-based, features trajectories based, people detection based and pixel-wise based algorithms [27]. Features regression-based algorithms start by extracting useful features and then use them to count people [21].…”

“…Each method has some advantages and disadvantages but people counting systems based on conventional camera are one of the best choices because CCTV cameras are widely used. For example, there are 5.9 million CCTV cameras in the UK [27]. People counting system is one of the most challenging systems in computer vision to implement [4], [21], [28]- [30].…”