Optics: measuring and testing – Dimension – Length
Reexamination Certificate
1999-03-08
2002-06-11
Kim, Robert H. (Department: 2877)
Optics: measuring and testing
Dimension
Length
C356S635000, C356S636000, C356S614000, C356S615000, C250S559220, C250S559260
Reexamination Certificate
active
06404506
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention pertains generally to laser-based systems for detecting objects moving across a planar surface, and more particularly to a non-intrusive laser-based detection system for determining vehicle travel time which is positioned above the plane of detection and configured such that it can unambiguously find the object boundaries in all lighting conditions independent of the time-of-flight of the laser.
2. Description of the Background Art
Travel time is the most important aspect of the Intelligent Transportation System (ITS). For example, travel time is a good indicator of other direct constraints on ITS efficiency, such as cost, risk, and attentive workload. The importance of travel time has been verified in ATIS user surveys which indicate that what travelers most want from a transportation system is (almost always) reduced travel time and higher reliability [15]. Every traveler must implicitly or explicitly make an assessment of these various travel time options before embarking on every trip; therefore, this information is definitely of high value. Because trip travel time is the parameter the public most wants to minimize, this is the parameter that is most important for transportation service providers to measure and minimize.
Speed is commonly used as an indicator of the travel time across a link. In current practice, speed is measured at one or more points along a link and extrapolated across the rest of the link [14]. This extrapolation method is used regardless of the mechanism of detection. Examples of detection methods include loops-which determine speed from two elements twenty feet apart [8]; radar, which can directly determine speed from the carrier frequency shift (Doppler effect); and video image processing, which tracks vehicles across the pixel elements within the field of view [9][10]. The extrapolation from a point to a line, however, is not necessarily valid. At the onset of flow breakdown, the speed variations along the length of a link can be quite large. Also, the onset of flow breakdown occurs when routing decisions are most time-critical and accurate information has the highest value, so inaccurate extrapolations could have detrimental effects to the traveler.
An alternate method to determine the traverse travel time (e.g. the true link speed) is to use Vehicles As Probes (VAP). A VAP system determines travel time directly by identifying vehicles at the start of the link and re-identifying them at the end of the link, with the time difference being the true travel time. The problem with VAP systems, however, is that they require large numbers of both vehicle tags and tag readers to be effective, and the cost justification of such a system may be unwarranted in the light of other options. The key aspect to measuring the actual travel time is simply to identify some distinguishing characteristic of a vehicle at the beginning of a link and then to re-identify that same characteristic on the same vehicle at the end of the link. This is the basic idea of VAP; however the characteristic does not have to be entirely unique (as in a vehicle tag), and it does not necessitate the infrastructure set-up costs of VAP [13].
As an alternative to VAPs, if a characteristic can be found to separate the fleet into, for example,
100
classifications, the “maximum probability fit” can be determined for the same sequence of classifications at the downstream detector as was identified at the upstream detector [2]. This is what is currently being done in Germany with the low-resolution imaging provided by new high-speed loops [12], and has been demonstrated in California. If a higher-resolution detector is used so that it is possible to get a few thousand classes, then it is theoretically possible to perform 100% upstream-downstream Origin and Destination (O/D) identification using time gating and other relatively straight-forward signal processing techniques (even if a significant percentage of the vehicles switch lanes). The mechanism of detection must allow highly resolved delineation between commonly available “commuter” vehicles, because commuter vehicles represent the majority of the vehicle stream during the period that traverse travel time information is most needed (e.g. the peak hours).
It is recognized that any mechanism to measure travel time, by definition, is only determining the “past state” of the transportation system. Collecting data on what happened in the past, however, has no utility unless it is used to infer what may happen in the future. Therefore, all decisions by definition are based on an inference of future consequences. For example, when a traveler learns that speed on a route is 50 MPH, the traveler generally infers that the speed will remain 50 MPH when he or she traverses it. This may or may not be a reasonable inference. Travelers want to know the “state” of the system in the future when they traverse it. In the simplest case, this is just a straight extrapolation of current “state.” More sophisticated travelers may develop their own internal conceptual model of the typical build up and progression of congestion along routes with which they are familiar. A major benefit of ITS will be to provide travelers with a much more valid and comprehensive “look ahead” model of the short term) future state of the transportation system.
In current practice, vehicle features are most commonly measured using inductive loops or video image processing. Because loops are buried beneath the pavement, installation requires heavy equipment, and traffic must be re-routed [19]. It is for this reason that loops are expensive to install and repair. More importantly, loop detectors cannot be relied upon to produce accurate speed (and therefore length) measurements because the inductive properties of the loop and loop detectors vary [19]. Video can be used to directly measure the length of vehicles, but the use of real time video image processing is problematic due to its computationally intensive nature. However, because video is a passive system (gathering ambient light), video images are dependent on the lighting conditions. Vehicle length measurements taken from video, even on the game vehicle, may not produce consistent results depending on time of day and weather conditions. For truly site and time independent vehicle length measurements, video would require an external source of illumination.
One system that addresses some of these problems is the Automatic Vehicle Dimension Measurement System (AVDMS) developed by the University of Victoria [4]. The AVDMS uses laser time of flight data to classify vehicles based on length, width, or height, and is based on the Schwartz Electro-Optics Autosense III sensor [7][20][21][22]. The Schwartz systems are entirely dependent on time-of-flight laser measurements with moving parts, similar to conventional LIDAR. In addition, the fundamental mechanism of detection is that the Schwartz detector determines the range (or distance) from the detector to the objects being detected. Furthermore, the laser of the Schwartz detector reflects off the vehicle to determine the size, shape, and “presence” of the vehicle.
In view of the foregoing, there is a need for a system that is easier to install and maintain than loop detectors, and which is mounted above the road so that, once installed, it can be maintained without disrupting the flow of traffic. There is a further need for a system that operates on a simple “on/off” basis, requiring much less computation for vehicle detection, and consequently much less computational hardware. While systems are known where the detectors are mounted above the road, such as in a Schwartz system, there is a further need for a system which does not operate on time-of-flight because time-of-flight systems are complex and susceptible to errors introduced by, for example, ambient temperature d
Cheng Harry H.
Hu Xudong
Katwyk Kirk Van
Larson Jonathan E.
Palen Joe
Kim Robert H.
Nguyen Sang H.
O'Banion John P.
The Regents of the University of California
LandOfFree
Non-intrusive laser-based system for detecting objects... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Non-intrusive laser-based system for detecting objects..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Non-intrusive laser-based system for detecting objects... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2962751