CN105513372A - Capture and recognition method of vehicle provided with license plate anti-photographing device - Google Patents

Abstract

The invention relates to a capture and recognition method of a vehicle provided with a license plate anti-photographing device. The method includes the following steps that when electronic police detect that a vehicle passes, an electric alarm flashlight is firstly controlled to be turned on and to perform first-time photographing on the vehicle for the first time; after the electric alarm flashlight is turned off, the vehicle is photographed for the second time; overexposure regions of pictures taken during the two times are recognized; when the picture taken at the first time has an overexposure region and the picture taken at the second time does not have an overexposure region, it is thought that the vehicle is provided with the license plate anti-photographing device. The method solves the problem that the vehicle cannot be punished after being provided with the license plate anti-photographing device, traffic violation behaviors such as overspeed can be well frightened, and the traffic accident rate is reduced. The electronic police have the functions of snapshot and detection of the installed license plate anti-photographing device, the traffic violation behaviors can be better cracked down, and traffic safety is guaranteed.

Description

That installs car plate anti-bat device vehicle catches recognition methods

Technical field

The present invention relates to a kind of car plate anti-bat device vehicle is installed catch recognition methods.Belong to intelligent transportation field, for carrying out technical perfection to the electronic police equipment such as bayonet socket candid photograph, overspeed snapping of current use, for overcoming the interference that the anti-candid photograph equipment of the illegal optics installed produces.

Background technology

In order to farthest reduce the generation of traffic hazard, the management and control dynamics of traffic department to vehicle drive behavior continues to increase, overspeed snapping, the electric police grasp shoot equipment such as to make a dash across the red light are seen everywhere, and vehicle supervision department's punishment capable to traffic offence also strengthens thereupon.

The candid photograph equipment such as current electronic police are all collected evidence with video mode, and when carrying out traffic offence punishment using number plate of vehicle number as important punishing justification, it is general not as punishing justification that number plate of vehicle takes unclear picture.In order to escape traffic offence punishment, part car owner installs car plate anti-bat device equipment to vehicle, instead capture process, the effect of the anti-bat device of car plate carries out candid photograph in real time to detect, strong reflection is produced in the moment that electronic police equipment is captured, cause car plate in the picture captured over-exposed, None-identified number plate of vehicle number.Install the car owner of car plate anti-bat device because escaping illegal punishment, irregular driving behavior showed increased even malice is illegal, brings larger hidden danger to traffic safety.

Conditional electronic police captures equipment (as overspeed snapping equipment) and is detecting to be correlated with have vehicle through out-of-date, the flashlamp (this flashlamp be called for short herein electricity alert flashlamp) opening electronic police system assembling carries out high brightness light filling in short-term, after light compensating lamp is opened, turn down camera exposure parameters, utilize the brightness of the alert flashlamp of electricity, capture the vehicle of process, the alert brightness of flash lamp of this electricity is more much larger than the environmental light brightness of surrounding, good solution ordinary ambient light is (as the sunshine on daytime, the car bulb etc. at night) interference problem, picture clearly can be photographed under when section in office and weather.The vehicle of the anti-bat device of car plate is installed through electronic police equipment, when the alert flashlamp of electricity is opened, the light sensor of the anti-bat device of car plate can detect that the brightness of surround lighting brightens suddenly, light rapidly the car plate anti-bat device be arranged near car plate and carry flashlamp (this flashlamp is called for short anti-bat flashlamp herein), car plate has good light-reflecting property, anti-bat flashlamp is opened and car plate and the brightness of car plate periphery can be made to promote rapidly, now electronic police video capture to picture near car plate and car plate by over-exposed, become a slice of vast expanse of whiteness, the discernible number-plate number cannot be obtained.

Summary of the invention

The present invention is directed to and install car plate anti-bat device vehicle and capture, the number plate of shooting clear, to collect evidence foundation as law enfrocement official on the one hand, car plate anti-bat device vehicle is installed in identification simultaneously, and recording-related information also uploads warning.

The technical scheme that the present invention is adopted for achieving the above object is: that installs car plate anti-bat device vehicle catches recognition methods, comprises the following steps:

When electronic police detected vehicle by time, first control the alert flashlamp of electricity and open and first time is carried out to vehicle and take pictures; After the alert flashlamp of electricity cuts out, carry out second time to vehicle take pictures;

The identification of over-exposed region is carried out to the picture that twice is taken pictures; When the picture of taking pictures for the first time exists over-exposed region and the picture that second time is taken pictures does not exist over-exposed region, think that the anti-bat device of car plate installed by this vehicle.

Carry out the identification of over-exposed region to comprise the following steps:

Binaryzation is carried out to the gray scale of picture and obtains binary image;

The sliding window of setting is traveled through on binary image, obtains the brightness statistics value in each sliding window corresponding region;

Sliding window region brightness statistics value on binary image be greater than corresponding to setting value K is over-exposed candidate region;

Each over-exposed candidate region is calculated standard deviation, and the standard deviation over-exposed candidate region be less than corresponding to threshold value is over-exposed region.

The described sliding window by setting travels through on binary image, and the brightness statistics value obtained in each sliding window corresponding region is realized by following formula: $A\left[m\right]\left[n\right]=\underset{i=0}{\overset{P-1}{\mathrm{\Σ}}}\underset{j=0}{\overset{Q-1}{\mathrm{\Σ}}}y[m+j][n+i]$ ②

Wherein, A [m] [n] is brightness statistics value; The data that in the gray-scale map that y [m] [n] is picture, m is capable, the n-th row are corresponding; I, j are respectively laterally, longitudinal accumulation step-length; M ∈ [0, W-P), n ∈ [0, H-Q); W is picture traverse, and H is picture altitude; P is sliding window height, and Q is sliding window width.

Described standard deviation

$S=\sqrt{\frac{\underset{i=0}{\overset{P-1}{\mathrm{\Σ}}}\underset{j=0}{\overset{Q-1}{\mathrm{\Σ}}}{\left(Y\right[m+i\left]\right[n+j]-\stackrel{\‾}{Y})}^2}{P\×Q}}$ ③

Wherein parameter the present invention has following beneficial effect and advantage:

1. the invention solves the problem can not carrying out after the anti-bat device of car plate installed by vehicle punishing, the traffic violations such as hypervelocity can well be frightened, reduce traffic hazard incidence.

2. the present invention uses flexibly, can use as stand-alone product, detect the equipment use installing car plate anti-bat device vehicle, also can be embedded in the Related product of electronic police and apply, electronic police is had and captures and detect installation car plate anti-bat device function, better strike traffic violation, ensures traffic safety.

3., when the present invention is embedded in and uses in electronic police Related product, do not need change electronic police products-hardware structure, only need increase algorithm, upgrade software, Project Realization is very convenient.

4., after the present invention's application, because car plate anti-bat device loses original effect, car plate anti-bat device series products also will be made progressively to disappear in the application of the secure context that blocks traffic.

5. the over-exposed detection algorithm of the present invention adopts over-exposed statistics, and coordinates the mode of brightness mean square deviation to judge, detection algorithm operand is little, is beneficial to the real-time detection of system.

Accompanying drawing explanation

Fig. 1 candid photograph process of the present invention sequential chart;

Fig. 2 process flow diagram of the present invention.

Embodiment

Below in conjunction with accompanying drawing, the present invention is described in detail.

As shown in Figure 1, the present invention can by amendment electric police grasp shoot software, detecting that car is through out-of-date, opens flashlamp, video capture first pictures, after candid photograph, close the alert flashlamp of electricity immediately, video capture second pictures, utilizes Recognition Algorithm of License Plate to carry out Car license recognition to the first pictures, if the first pictures can recognize the brand number of vehicle, then show that the anti-bat device of car plate do not installed by this vehicle; If the first pictures does not recognize number plate of vehicle, then detection is further analyzed to the first pictures, if the first pictures does not exist the over-exposed region meeting or exceeding certain size, then illustrates that the anti-bat device of car plate do not installed by this vehicle; If the first pictures exists the over-exposed region that pixels across point is greater than 60 pixels, capture picture to second and carry out Car license recognition, if the second pictures can correctly recognize number plate of vehicle number, illustrate that this vehicle has and car plate anti-bat device suspicion is installed, first pictures is uploaded to back-end server together with the second pictures, upload warning message simultaneously, carry out artificial judgment by the administrative authority be correlated with, make respective handling.

Simultaneously, when candid photograph the second pictures, because there is no flash lighting, can not produce over-exposed, can normally catch the illegal activities of installing car plate anti-bat device vehicle, while whether identification vehicle installs car plate anti-bat device, also make the anti-bat device of vehicle lose efficacy, play the effect that car plate anti-bat device car owner driving behavior is installed in constraint further.

As shown in Figure 2, concrete steps of the present invention are as follows:

1, vehicle detection

Whether have vehicle pass through, after having detected that car passes through, notice video camera standard is originally taken pictures if adopting video, radar or Coil Detector mode to detect.

2, video camera is captured first

After video camera receives candid photograph order, parameter is captured in adjustment camera shutter time and gain etc., opens electric alert flashlamp, captures the first pictures.

3, video camera second time is captured

After capturing first, parameter such as candid photograph such as video camera adjustment time shutter and gain etc. (pre-set according to on-the-spot actual environment and capture parameter) does not start the alert flashlamp of electricity and captures.

Whether 4, the anti-bat device of car plate be installed detect

Pass through Video Detection Algorithm, collection picture is detected, the vehicle installing the anti-bat device of car plate does not collect car plate clearly substantially, namely there is vehicle in the picture of license plate recognition result that the anti-bat device of car plate is not all installed, therefore only the vehicle of the first pictures captured without license plate recognition result is judged.

When the first pictures has license plate recognition result, this picture is according to not installing the process of car plate anti-bat device.To the first pictures without license plate recognition result, judge whether installed the anti-bat device of car plate through vehicle by the over-exposed situation detecting two pictures, detailed process is as follows:

1) binary conversion treatment

Camera acquisition picture is carried out yuv format conversion (format conversion is not needed to the video camera that directly can obtain YUV information), wherein " Y " luminance component (Luminance) that is image, the chromatic component (Chrominance) that " U ", " V " are image, wherein " Y " represents lightness, namely grey decision-making; And " U " and " V " represents is colourity, effect describes colors of image and saturation degree, is used to specify the color of pixel.Because " Y " is luminance component image, over-exposed pixel " Y " value is comparatively large, carries out overall binaryzation, that is: to the component of picture " Y "

$y\left[m\right]\left[n\right]=\left\{\begin{array}{cc}0& Y\left[m\right]\left[n\right]<T\\ 1& Y\left[m\right]\left[n\right]\&GreaterEqual;T\end{array}\right.$ ①

Wherein: Y [m] [n] is for m in image is capable, " Y " component of the pixel of the n-th row, y [m] [n] is the result after Y [m] [n] binaryzation, T is binary-state threshold, select a fixed value close to " Y " maximal value, " Y " maximal value is 255, when T ∈ [235,245] all can meet the demands, select T=240 through a large amount of actual Test Suggestion.

2) over-exposed region coarse search

The car plate size of existing electronic police identification is generally pixels across o'clock in 70 to 160 pixels, and most of car plate comprises 7 characters (as: distant A12345), and namely each character boundary is horizontal 7 ~ 23 pixels.In order to reach to the over-exposed object of car plate, then exposure area will to covering 2 characters, namely exposure area is necessarily greater than 23 × 2=46 pixel, therefore setting is less than the sliding window of 46 × 46 (is convenience of calculation, frame size can be adopted to be 40 × 40), brightness statistics is carried out, that is: in sliding window

$A\left[m\right]\left[n\right]=\underset{i=0}{\overset{P-1}{\mathrm{\&Sigma;}}}\underset{j=0}{\overset{Q-1}{\mathrm{\&Sigma;}}}y[m+j][n+i]$ ②

Wherein, y [m] [n] is the data that the m in formula 1. operation result is capable, the n-th row are corresponding, m ∈ [0, W-P), and n ∈ [0, H-Q), W is picture traverse, and H is picture altitude.P is sliding window height, and Q is sliding window width; I, j are respectively laterally, longitudinal accumulation step-length; In this enforcement, step-length is 1 pixel.

3) over-exposed region is determined

If 2. formula of choosing calculates value (usual desirable K=P × Q × 0.8 being greater than K in A [m] [n], namely in selected areas, " Y " component of more than 80% pixel is greater than T), if qualified A [m] [n] is more, then get three maximal values without overlapping region.Standard deviation is calculated to original luminance component Y [m] [n] of selected A [m] [n] corresponding region, that is:

$S=\sqrt{\frac{\underset{i=0}{\overset{P-1}{\mathrm{\&Sigma;}}}\underset{j=0}{\overset{Q-1}{\mathrm{\&Sigma;}}}{\left(Y\right[m+i\left]\right[n+j]-\stackrel{\&OverBar;}{Y})}^2}{P\&times;Q}}$ ③

Wherein, $\stackrel{\&OverBar;}{Y}=\frac{1}{P\&times;Q}\underset{i=0}{\overset{P-1}{\&Sigma;}}\underset{j=0}{\overset{Q-1}{\&Sigma;}}(Y\&lsqb;m+i\&rsqb;\&lsqb;n+j\&rsqb;).$

Design constant C, as the thresholding decision content of S, to image procossing, can get a fixed integer value of C ∈ (3,10), the concrete exposure parameter adjustment optimum efficiency needed according to video camera.If at least there is P × Q pixel region in the first pictures, make:

$\left\{\begin{array}{c}A\left[m\right]\left[n\right]\&GreaterEqual;K\\ S\&le;C\end{array}\right.$ ④

Set up, then judge that the first pictures exists over-exposed region, otherwise be without over-exposed region.

If there is over-exposed region in the first pictures, then to the second pictures captured according to 1), 2), 3) step processes, and judges whether second picture captured exists over-exposed region.

If capture picture and have over-exposed region for first, there is not over-exposed region in the second pictures, then the vehicle that this picture is caught has installs car plate anti-bat device suspicion, uploads warning message, carries out artificial judgment, otherwise according to not installing the process of car plate anti-bat device.

Claims (4)

1. that installs car plate anti-bat device vehicle catches recognition methods, it is characterized in that comprising the following steps:

When electronic police detected vehicle by time, first control the alert flashlamp of electricity and open and first time is carried out to vehicle and take pictures; After the alert flashlamp of electricity cuts out, carry out second time to vehicle take pictures;

The identification of over-exposed region is carried out to the picture that twice is taken pictures; When the picture of taking pictures for the first time exists over-exposed region and the picture that second time is taken pictures does not exist over-exposed region, think that the anti-bat device of car plate installed by this vehicle.

2. installation car plate according to claim 1 anti-bat device vehicle catch recognition methods, carry out described in it is characterized in that over-exposed region identify comprise the following steps:

Binaryzation is carried out to the gray scale of picture and obtains binary image;

The sliding window of setting is traveled through on binary image, obtains the brightness statistics value in each sliding window corresponding region;

Sliding window region brightness statistics value on binary image be greater than corresponding to setting value K is over-exposed candidate region;

Each over-exposed candidate region is calculated standard deviation, and the standard deviation over-exposed candidate region be less than corresponding to threshold value is over-exposed region.

3. installation car plate according to claim 2 anti-bat device vehicle catch recognition methods, it is characterized in that described by setting sliding window travel through on binary image, the brightness statistics value obtained in each sliding window corresponding region is realized by following formula: $A\&lsqb;m\&rsqb;\&lsqb;n\&rsqb;=\underset{i=0}{\overset{P-1}{\&Sigma;}}\underset{j=0}{\overset{Q-1}{\&Sigma;}}y\&lsqb;m+j\&rsqb;\&lsqb;n+i\&rsqb;$ ②

Wherein, A [m] [n] is brightness statistics value; The data that in the gray-scale map that y [m] [n] is picture, m is capable, the n-th row are corresponding; I, j are respectively laterally, longitudinal accumulation step-length; M ∈ [0, W-P), n ∈ [0, H-Q); W is picture traverse, and H is picture altitude; P is sliding window height, and Q is sliding window width.

4. installation car plate according to claim 2 anti-bat device vehicle catch recognition methods, it is characterized in that described standard deviation

$S=\sqrt{\frac{\underset{i=0}{\overset{P-1}{\&Sigma;}}\underset{j=0}{\overset{Q-1}{\&Sigma;}}{(Y\&lsqb;m+i\&rsqb;\&lsqb;n+j\&rsqb;-\stackrel{\&OverBar;}{Y})}^2}{P\&times;Q}}$ ③

Wherein $\stackrel{\&OverBar;}{Y}=\frac{1}{P\&times;Q}\underset{i=0}{\overset{P-1}{\&Sigma;}}\underset{j=0}{\overset{Q-1}{\&Sigma;}}(Y\&lsqb;m+i\&rsqb;\&lsqb;n+j\&rsqb;).$