An optical method for automated roadside detection and counting of vehicle occupants

Abstract

The monitoring and control of traffic volume is becoming a constant social, economic, and environmental pressure in the UK and elsewhere, because of landmass and current infrastructure strain under a swelling and increasingly mobile population. The viability of high-occupancy vehicle (HOV) lanes for easing traffic congestion, and hence maximising traffic flow, has been proven in countries worldwide. The USA, Australia, and Canada have had HOV installations for some time, controlling the flux of traffic into their most densely populated areas. Experience in these cases has dictated that it is enforcement which is crucial to the successful implementation of such a traffic policy. To date, all enforcement has been manual, i.e. a police officer counting the occupants in a vehicle as it passes by. Studies have concluded that manual enforcement is typically only 65 per cent accurate and, considering the pressures which one individual is put under in these circumstances, this statistic is not surprising. Lighting and environmental conditions, skin tone, and location of the occupants and the alertness of the officer are all variables affecting the accuracy of manually collected data. Hence there is the need for an autonomous detection system to count the occupancy within a vehicle. This could provide the basis of an enforcement or congestion strategy.

Passenger Occupancy Estimation in Vehicles: A Review of Current Methods and Research Challenges

“…In 2008, Tyrer and Lobo published a paper [72] regarding passenger occupation limitations and issues and performed night tests with captured IR images and images in the visible part of the spectrum. The authors display the reflectance spectra of Caucasian, Asian, and African skin as well as the transmission and absorption spectra of typical windscreens.…”

“…Different colours of clothes and skin present another difficulty. Some papers have investigated the influence of such parameters [2,8,62,63,72,146]. Many environmental aspects affect the final accuracy of these systems.…”

Occupant Monitoring System for Traffic Control Based on Visual Categorization

“…Most of the automatic systems for occupancy detection in the last years, are based on cameras placed on the infrastructure [8], [9], [10]. However, recently, the focus has shifted to in-vehicle sensing.…”

Front-view car detection and counting with occlusion in dense traffic flow

“…The captured vehicles may have variances on local shape, color, view and are affected by local occlusion and illumination conditions… This challenges most of the existing object detection methods. Classical visual surveillance techniques have been successfully used on highways such as frame differencing, background estimation [14,15], optical flow [16], edge detection [17]… Traffic analysis on urban traffic domain appear to be more challenging because of high traffic density flow, lower camera angles that lead to a high degree of occlusion [1]. Methods based on background modeling [18,19] may fail to segment slowly moving or stationary vehicles or groups of vehicles could be clustered into the same blob due to occlusion.…”