CN201215688Y - Full view field fixed investigation device for transportation flow - Google Patents

Abstract

The utility model relates to a whole-view video fixed traffic flow investigating device, wherein analog signals produced by a video camera (1) get into an A/D conversion circuit (5) through an input circuit (4) and converse analog signals into digital signals to be connected with an I/O input/output interface (8), and the I/O input/output interface (8) is connected with a CPU (9), the CPU (9) is connected with a memory (10), the I/O input/output interface (8) further reaches an executive system (11) through an output circuit (6), and the executive system (11) gets into a data middle point (14) with a transmission protocol through a transmission medium (13). The whole-view video fixed traffic flow investigating device overcomes the shortcomings of bad detection precision of the existing test equipment and the like. The utility model has the advantages of excellent stability, high detection precision and the like.

Description

Full ken video fixed traffic flow investigation equipment

Technical field

The utility model relates to a kind of traffic flow investigation apparatus, and more particularly it is a kind of full ken video fixed traffic flow investigation equipment.It utilizes the Video Detection principle, by the computer general-purpose technology, realizes the information of vehicle flowrate collection of highway, highway, bridge, tunnel and special road section, reflects traffic flow data authentic and validly.

Background technology

Utilizing video technique to detect traffic dynamic information (flow, speed, occupation rate, following distance etc.) just begins one's study as far back as the seventies developed country, so far studied for three more than ten years, but the method for Video Detection traffic dynamic information never changes, can not follow the tracks of in the full ken and detect, but open up a detection window at a certain section of road, only this window half-tone information is carried out Treatment Analysis, be similar to ground induction coil.The shortcoming of this method is as follows: A, the characteristic information that obtains the vehicle that flows are few, and identification vehicle difficulty is big; B, to obtain the vehicular movement track short, and it is big that the speed of a motor vehicle detects error; C, to not according to the easy omission of road driving vehicle or heavily the inspection; D, vehicle inverted image can cause heavily inspection; E, video camera installation site require the center, otherwise can't guarantee accuracy of detection.F, to follow the tracks of the computing ken less, single complete equipment sensing range little (being generally in 4 tracks).If it is more to detect the section track, many cover checkout equipments (containing video camera and video processor) must be installed, with high costs.

Along with VLSI (very large scale integrated circuit) great-leap-forward development in recent years, the bottleneck of foreground detecting device arithmetic capability is progressively eliminated, and in the full FOV (Field of View) moving image being followed the tracks of detection becomes possibility.

Summary of the invention

The purpose of this utility model is to overcome above-mentioned existing disadvantages of background technology part, and a kind of full ken video fixed traffic flow investigation equipment is provided.

The ultimate principle of full ken video fixed traffic flow investigation equipment: detect road surface image black white contrast and change.Vehicle recognition principle: according to a plurality of external appearance characteristics such as length and width height and windscreen, preceding faces.Speed of a motor vehicle recognition principle: measure the Time Calculation speed of a motor vehicle by whole camera field of view.Mounting means: video camera is installed above the road surface.Advantage: probe does not contact with vehicle, and long service life is not destroyed the road surface.Shortcoming: need special-purpose column.Accuracy of detection: the flow detection precision is higher daytime, night accuracy of identification descend to some extent, the vehicle classification precision is very high daytime, night accuracy of identification descend to some extent, speed of a motor vehicle accuracy of identification is higher, the distance that tests the speed reaches more than 100 meters.

The purpose of this utility model reaches by following measure: full ken video fixed traffic flow investigation equipment, it is characterized in that it comprises that two stands are located at the video camera of road top, every video camera is taken the vision signal of a wagon flow direction, the simulating signal that described video camera produces enters the A/D change-over circuit by input circuit and becomes digital signal to connect the I/O input/output interface analog signal conversion, or video camera 1 enters connection I/O input/output interface with the digital signal that produces by input circuit, the I/O input/output interface connects CPU9, the CPU9 connected storage, the I/O input/output interface also connects output circuit and arrives executive system, and described executive system enters the data mid point by transmission medium and host-host protocol.

In technique scheme, described transmission medium is microwave wireless, ADSL, optical fiber or GPRS.

Technology such as the full ken video of the utility model fixed traffic flow investigation equipment integrated use advanced person's communication information, network, control automatically, traffic engineering, improve the ruuning situation of traffic and transportation system, improve conevying efficiency and security, reduce traffic hazard, reduce environmental pollution, thereby set up the overall transportation system of intelligentized, safe, convenient, comfortable a, environmental protection, improved work efficiency and quality.

Description of drawings

Fig. 1 is the structural representation that video camera is installed at road in the utility model.

Fig. 2 is the circuit block diagram of input of the present utility model and output.

Fig. 3 is a system signal transmission synoptic diagram of the present utility model.

Embodiment

Describe performance of the present utility model in detail below in conjunction with accompanying drawing, but they do not constitute to qualification of the present utility model, simultaneously by illustrating that advantage of the present utility model will become clear more and understands easily.

Consult accompanying drawing as can be known: the full ken video of the utility model fixed traffic flow investigation equipment, it comprises that two stands are located at the video camera 1 of road top, every video camera is taken the vision signal of a wagon flow direction, the simulating signal that described video camera 1 produces enters A/D change-over circuit 5 by input circuit 4 and becomes digital signal to connect I/O input/output interface 8 analog signal conversion, or video camera 1 enters connection I/O input/output interface 8 with the digital signal that produces by input circuit 7, I/O input/output interface 8 connects CPU9, CPU9 connected storage 10, I/O input/output interface 8 also arrives executive system 11 by output circuit 6, and described executive system 11 enters data mid point 14 by transmission medium 13 and host-host protocol.Described transmission medium 13 is microwave wireless, ADSL, optical fiber or GPRS.

Every cover system mainly contains: 2 stands are located at the video camera of road top, and every video camera is taken the vision signal of a wagon flow direction; 1 traffic information collection device includes parts such as embedded computer board, CPU, internal memory, CF card, video frequency collection card, encryption, communication; 1 outer cabinet, inside are equipped with traffic information collection device, power supply, communication, lightning protection, equipment such as antitheft; The backstage traffic information management system.

Equipment set is installed on the greenbelt or the roadside of site road central authorities by vertical rod, the vision signal that collector for processing is obtained from video camera, obtain traffic data information and compressed video stream informations such as vehicle, flow, the speed of a motor vehicle, by wired or wireless communication information is sent to data processing centre (DPC) and further handles.

Vehicle recognition principle of the present utility model: 1. big at domestic vehicle class many (the especially nonstandard transformations of the way), feature difference, sample obtains factor such as cost height and causes the big problem of vehicle identification difficulty, the utility model video traffic information system selection neural network algorithm, by extract vehicle length and two geometric properties of width, four not displacement feature and three textural characteristics of deriving by gray level co-occurrence matrixes the eigenvector formed of totally nine characteristic components vehicle is described.2. the speed of a motor vehicle detects principle: this programme adopts dummy line speed of a motor vehicle detection algorithm (as shown in Figure 1): preestablish two dummy line (can set automatically or by hand) in the video camera ken, and import distance L between two dummy line, the vehicle that a certain detection territory is occurred and passes through, as long as obtain passing through the time of vehicle, just can calculate the speed of vehicle by following computing formula:

$V\left(k\right)=\frac{L}{T^{\mathrm{End}}\left(K\right)-T^{\mathrm{Start}}\left(K\right)}---\left(1\right)$

K is the sequence number that vehicle occurs

The mean flow rate of road surface wagon flow is:

$V_{\mathrm{avg}}=\frac{1}{N_2-{\mathrm{<figure-callout\; id="1"\; label="N"\; filenames="Y200820068135E00091.png"\; state="\{\{state\}\}">N</figure-callout>}}_1}=\underset{k=N_1}{\overset{N_2}{\mathrm{\&Sigma;}}}V\left(k\right)---\left(2\right)$

N ₂, N ₁Be respectively the zero-time of metering time.

The full ken target dynamic of the utility model tracking detection method ultimate principle is: enter the analyzable ken from destination object, just constantly follow the tracks of and detect its shape, orientation, motion state and sequential relationship etc., thereby obtain corresponding information.This method, relate to the still jejune forefront technology of computer vision, the difficulty that realizes is very big, but owing to can record the motion state of vehicle exactly, and can obtain a plurality of samples of same vehicle, therefore, can obtain quite high accuracy of detection and obtain particularly vehicle movement parameter of more traffic parameter.Particularly can detect traffic parameter comparatively exactly under no illumination condition evening, this is that other similar Video Detection product is not available in the world.System is under most of light (comprise and do not have illumination night) condition, and the accuracy rate of flow detection is greater than 90%, and can detect Ministry of Communications's regulation 9 kinds of vehicles.This method also has a key advantage to be because system need not to catch the instant that vehicle is breasted the tape, therefore can on the non-real time operating system platform, carry out product development, thereby can adopt more standardization and more ripe operating system and hardware thereof, this greatly improves the reliability of system.

The utility model is as follows with the advantage that the conventional video checkout equipment is compared:

	Conventional video detects	The utility model type Video Detection
			Place capacity	The single device sensing range is less.	Single device can detect 16 tracks.
Video camera is installed	Must install above center line of road, the installation restriction is many.	Can install in the road both sides, construction does not influence traffic.
			Vehicle identification	Extract geometric properties vectors such as simple length, width, the vehicle recognition accuracy is higher, but it is few to discern the vehicle kind.	Extract two geometric properties of vehicle, four not displacement feature, three textural characteristics the eigenvector formed of totally nine characteristic components vehicle is described, it is many to discern the vehicle kind, the accuracy rate height.
The speed of a motor vehicle detects	Mobile track of vehicle short (being generally average vehicle length), range rate error is big.	Mobile track of vehicle long (can reach 100 meters) tests the speed accurately.
			Anti-interference	Anti-background interference ability is strong, but it is low that vehicle is blocked the ELIMINATION OF ITS INTERFERENCE rate.	Integral body by the full feature vector is remembered vehicle, so the total system antijamming capability is strong.
System debug	The artificial vehicle characteristic parameter of adjusting, consuming time more, debugging is complicated, change vehicle complex operation.	The artificial setting, system intelligence study vehicle, feature is analyzed and extracted to software automatically, and debugging is simple, supports the User Defined vehicle.
			Hardware requirement	Operand is little, and hardware requirement is low	The computing complexity, the hardware requirement height

The full ken video of the utility model fixed traffic flow investigation equipment performance index:

Claims (2)

1, full ken video fixed traffic flow investigation equipment, it is characterized in that it comprises that two stands are located at the video camera (1) of road top, every video camera is taken the vision signal of a wagon flow direction, the simulating signal that described video camera (1) produces enters A/D change-over circuit (5) by input circuit (4) and becomes digital signal to connect I/O input/output interface (8) analog signal conversion, or video camera (1) enters connection I/O input/output interface (8) with the digital signal that produces by input circuit (7), I/O input/output interface (8) connects CPU (9), CPU (9) connected storage (10), I/O input/output interface (8) also arrives executive system (11) by output circuit (6), and described executive system (11) enters data mid point (14) by transmission medium (13) and host-host protocol.

2, full ken video fixed traffic flow investigation equipment according to claim 1 is characterized in that described transmission medium (13) is microwave wireless, ADSL, optical fiber or GPRS.

Images