CN104899880A - Method for automatically detecting open and closed states of vehicle door of public traffic vehicle - Google Patents

Abstract

The present invention discloses a method for automatically detecting open and closed states of a vehicle door of a public traffic vehicle. According to the method, image data of the vehicle door is acquired by a video acquisition device; an image frame of the vehicle before departure is stored as a background image; a detection area is calibrated; an acquired real-time image frame and the background image are subjected to comparison analysis in the calibrated area; when both gray level difference and comparison difference of the image frame and the background image are large, the vehicle door is considered to be in the open state; and otherwise the vehicle door is judged to be in the closed state. In the specific operation process, a camera is mounted at the position of a vehicle roof above a driver, and a shooting angle and a focal length of the camera are regulated, such that the vehicle door is enabled to be positioned at the center of a shot image and occupy the integral shot image to the greatest extent.

Description

A kind of public transit vehicle opening/closing door of vehicle state automatic testing method

Technical field:

The present invention relates to a kind of expanded application of bus camera monitoring system, particularly relate to a kind of public transit vehicle opening/closing door of vehicle state automatic testing method based on image processing techniques.

Background technology:

Information due to video capture device collection has the information such as abundant color, structure, texture and time, is usually widely used in traffic and transport field, as supervisory system in bus.Monitoring camera installed by current most bus, and has corresponding storage facilities and transmission interface.In bus, supervisory system is by installing multiple camera in bus, in vehicle operation, records the inside and outside abnormal conditions of car at any time, for the operation of public transport, management and optimization provide investigational data, the driving of specification public transit vehicle and operation.But at present also not for public transit vehicle opening/closing door of vehicle state automatic testing method.

Summary of the invention:

The object of the present invention is to provide a kind of public transit vehicle opening/closing door of vehicle state automatic testing method based on image processing techniques.

For achieving the above object, the present invention adopts following technical scheme to be achieved:

A kind of public transit vehicle opening/closing door of vehicle state automatic testing method, comprises the steps:

Step 0: gather bus and to dispatch a car the video background picture frame F of front door _t;

Step 1: video background image frame F before the bus of collection is dispatched a car _tgray processing, obtains background grey-level image frame f _t;

Step 2: obtain Background Bg by averaging method in background grey-level image frame;

Step 3: binaryzation is carried out to the Background Bg that step 2 obtains, obtains background binary figure MG;

Step 4: carry out closing operation of mathematical morphology to the background binary figure MG that step 3 obtains, obtains gradient connection layout OG;

Step 5: morphology opening operation is carried out to the gradient connection layout OG that step 4 obtains, obtains gradient map CG;

Step 6: carry out negate and star process of zero-suppressing to the gradient map CG that step 5 obtains, add up the pixel count in each connection district, extracts the maximum connected region of pixel count as car door Prototype drawing MB;

Step 7: the Background Bg that car door Prototype drawing MB step 6 obtained and step 2 obtain carries out dot product, obtain background detection figure ej, computing formula is as follows:

ej＝MB·×Bg (6)

Wherein, × represent point multiplication operation symbol;

Step 8: the average gray value eb calculating background detection figure ej;

Step 9: the car door picture frame E after Real-time Collection vehicle sends _k;

Step 10: to the car door picture frame E gathered in step 9 _kaccording to the method for step 1, gray processing process is carried out to it, obtain the car door gray-scale map of kth frame, be designated as Eh _k;

Step 11: the car door gray-scale map Eh that car door Prototype drawing MB step 6 obtained and step 10 obtain _kcarry out dot product, obtain kth frame car door detection figure e _k, computing formula is as follows:

e _k＝MB·×Eh _k(8)

Step 12: the kth frame car door detection figure e obtained in calculation procedure 11 _kaverage gray value eh _k;

Step 13: calculate car door and detect figure e _kwith the gray difference degree ed of background detection figure ej _k, its calculating formula is as follows:

ed _k＝|eh _k-eb| (10)

Step 14: calculate car door and detect figure e _kwith contrast difference's degree r of background detection figure ej _k;

Step 15: opening/closing door of vehicle state judges; Opening/closing door of vehicle state through type (12) judges;

Wherein, fl _kfor the mark of opening/closing door of vehicle state, fl _k=0 represents that car door is closed condition, fl _k=1 represents that car door is opening; T ₁for gray difference degree threshold value, T ₁span be: 3 ~ 10; T ₂for contrast difference's degree threshold value, its span is-0.5 ~ 0.

The present invention further improves and is, also comprises the steps:

Step 16: the opening/closing door of vehicle state storing kth frame car door picture frame;

Step 17: whether public transit vehicle reaches home; No, proceed to step 18; Then proceed to step 19;

Step 18: gather next frame, proceed to step 9.

The present invention further improves and is, in step 0: the video background picture frame of 5s before collection bus sets out, wherein, sample frequency is 0.5s/ frame, gathers 10 two field pictures altogether, is designated as F _t, wherein, t is frame number, t=1,2 ..., 10; The image size collected is M capable and N row, and set the coordinate of certain pixel as (x, y), x and y represents the row and column of pixel (x, y), satisfied: x and y is integer, and 1≤x≤M, 1≤y≤N, then t frame background image frame F _tthe RGB color value symbolically of middle pixel (x, y) is (R _t(x, y), G _t(x, y), B _t(x, y)).

The present invention further improves and is, in step 1: step 0 is processed the t frame background image frame F obtained _tcarry out gray processing process, obtain background grey-level image frame and be designated as f _t, its calculating formula is as follows:

f _t(x,y)＝0.3×R _t(x,y)+0.59×G _t(x,y)+0.11×B _t(x,y) (1)

Wherein, R _t(x, y) represents background image frame F _tthe R component value of middle pixel (x, y); G _t(x, y) represents background image frame F _tthe G component value of middle pixel (x, y); B _t(x, y) represents background image frame F _tthe B component value of middle pixel (x, y); f _t(x, y) represents background grey-level image frame f _tthe gray-scale value of middle pixel (x, y).

The present invention further improves and is, in step 2: the computing formula of Background Bg is shown below:

$\mathrm{Bg}(x,y)=\frac{1}{10}\underset{t=1}{\overset{10}{\mathrm{\Σ}}}{f}_t(x,y)---\left(2\right)$

Wherein, Bg (x, y) represents the gray-scale value of pixel (x, y) in Background Bg;

In step 3: carry out binaryzation to Background Bg and obtain background binary figure MG, its calculating formula is as follows:

Wherein, T is binary-state threshold, and its value is: 110 ~ 140; MG (x, y) represents the value of pixel (x, y) in background binary figure MG, MG (x, y)=0 represents that pixel (x, y) is possible car door region, MG (x, y)=1 represents that pixel (x, y) is non-car door region.

The present invention further improves and is, in step 4: closed operation background binary figure MG being carried out to mathematical morphology, and obtain gradient connection layout OG, its calculating formula is as follows:

Wherein, Se is structural element, gets the square structure element of 3 × 3 ~ 10 × 10; MG ο Se represents that background binary figure MG is carried out closed operation operation by construction operator Se; ο represents closing operation of mathematical morphology; ⊕ represents dilation operation; represent erosion operation.

The present invention further improves and is, in step 5: carry out mathematical morphology open operator to gradient connection layout OG, obtain gradient map CG, and to eliminate the burr in gradient connection layout OG, its calculating formula is as follows:

$\mathrm{CG}=\mathrm{OG}\·\mathrm{Se}=(\mathrm{OG}\⊗\mathrm{Se})\⊕\mathrm{Se}---\left(5\right)$

Wherein, represent morphology opening operation, OGSe represents that gradient connection layout OG is carried out opening operation operation by construction operator Se.

The present invention further improves and is, in step 8: the calculating formula of average gray value eb is as follows:

$\mathrm{eb}=\frac{1}{\mathrm{MN}}\underset{x=1}{\overset{M}{\mathrm{\Σ}}}\underset{y=1}{\overset{N}{\mathrm{\Σ}}}\mathrm{ej}(x,y)---\left(7\right)$

Wherein, ej (x, y) represents the gray-scale value of pixel (x, y) in background detection figure ej.

The present invention further improves and is, in step 12: average gray value eh _kcalculating formula as follows:

${\mathrm{eh}}_k=\frac{1}{\mathrm{MN}}\underset{x=1}{\overset{M}{\mathrm{\Σ}}}\underset{y=1}{\overset{N}{\mathrm{\Σ}}}{e}_k(x,y)---\left(9\right)$

Wherein, e _k(x, y) represents kth frame car door detection figure e _kthe gray-scale value of middle pixel (x, y).

The present invention further improves and is, in step 14: contrast difference's degree r _kcalculating formula as follows:

$r_k=\frac{\underset{x=1}{\overset{M}{\mathrm{\Σ}}}\underset{y=1}{\overset{N}{\mathrm{\Σ}}}\mathrm{\Σ}(e_k(x,y)-{\mathrm{eh}}_k)(\mathrm{ej}(x,y)-\mathrm{eb})}{\sqrt{\left(\underset{x=1}{\overset{M}{\mathrm{\Σ}}}\underset{y=1}{\overset{N}{\mathrm{\Σ}}}{(e_k(x,y)-{\mathrm{eh}}_k)}^2\right)\left(\underset{x=1}{\overset{M}{\mathrm{\Σ}}}\underset{y=1}{\overset{N}{\mathrm{\Σ}}}{(\mathrm{ej}(x,y)-\mathrm{eb})}^2\right)}}---\left(11\right)$

Wherein, r _kfor contrast difference's degree, its value is larger then illustrates that car door detects figure e _klarger with the diversity factor of background detection figure ej, on the contrary also anti-.

Relative to prior art, the present invention has following advantage:

The car door view data that the present invention utilizes video capture device to gather, the picture frame of vehicle before dispatching a car is saved as Background, and demarcate surveyed area, in demarcation region, the realtime graphic frame of collection and Background are analyzed, when the gray difference degree of picture frame and Background and contrast difference's degree all larger time, think that car door is in opening; Otherwise judge that car door is closed condition.Camera is installed on roof location above driver, regulates the shooting angle of camera and focal length: make car door be in the central authorities of shooting picture, and take whole shooting picture as far as possible.

Further, the open and-shut mode that the present invention detects public transit vehicle car door automatically has important practical significance: one is trigger data acquisition.In vehicle traveling with when stopping, the region paid close attention in bus is different, and the data that corresponding sensor gathers also are not quite similar, and the open and-shut mode by detection car door carrys out the opening and closing of trigger data acquisition sensor; Two is the abnormal conditions monitoring car door.By the open and-shut mode of bus door and driver, the operation to car door contrasts, and whether normally can monitor car door duty, the abnormality of Timeliness coverage opening/closing door of vehicle, ensure the safety of people, car, thing.Three is realizations of following unmanned public transit vehicle, first needs the open and-shut mode detecting car door in real time, could travel subsequently to vehicle, the manipulation such as stop.Given this, the present invention proposes a kind of detection method of public transit vehicle opening/closing door of vehicle state.The advantages such as the method has simple, economic, and real result is reliable.

Accompanying drawing illustrates:

Fig. 1 is Background Bg.

Fig. 2 is background binary figure MG.

Fig. 3 is gradient connection layout OG.

Fig. 4 is gradient map CG.

Fig. 5 is car door Prototype drawing MB.

Fig. 6 is background detection figure ej.

Fig. 7 is car door picture frame E _k.

Fig. 8 is car door gray-scale map Eh _k.

Fig. 9 is car door detection figure e _k.

Embodiment:

Below in conjunction with drawings and Examples, the present invention is described in further detail.

A kind of public transit vehicle opening/closing door of vehicle of the present invention state automatic testing method, comprises the steps:

Step S0: bus dispatch a car before the collection of video background image frame.The video background picture frame of 5s before collection bus sets out, sample frequency is 0.5s/ frame, gathers 10 two field pictures altogether, is designated as F _t, wherein, t is frame number, t=1,2 ..., 10.The image size collected is 180 × 250, i.e. M=180, N=250, assuming that the coordinate of certain pixel is (x, y), x and y represents the row and column of pixel (x, y), meets: x and y is integer, and 1≤x≤180,1≤y≤250.T frame background image frame F _tthe RGB color value symbolically of middle pixel (x, y) is (R _t(x, y), G _t(x, y), B _t(x, y)).

Proceed to step S1.

Step S1: video background picture frame gray processing.By the background image frame F that step S0 process obtains _tcarry out gray processing process, obtain background grey-level image frame and be designated as f _t, its calculating formula is as follows:

f _t(x,y)＝0.3×R _t(x,y)+0.59×G _t(x,y)+0.11×B _t(x,y)，t＝1,2,……10 (1)

Wherein, R _t(x, y) represents background image frame F _tthe R component value of middle pixel (x, y); G _t(x, y) represents background image frame F _tthe G component value of middle pixel (x, y); B _t(x, y) represents background image frame F _tthe B component value of middle pixel (x, y); f _t(x, y) represents background grey-level image frame f _tthe gray-scale value of middle pixel (x, y).

Proceed to step S2.

Step S2: obtain Background by averaging method.Background Bg represents (as shown in Figure 1), and its computing formula is shown below:

$\mathrm{Bg}(x,y)=\frac{1}{10}\underset{t=1}{\overset{10}{\mathrm{\Σ}}}{f}_t(x,y)---\left(2\right)$

Wherein, Bg (x, y) represents the gray-scale value of pixel (x, y) in Background Bg.

Proceed to step S3.

Step S3: binaryzation is carried out to the Background Bg that step S2 obtains.Carry out binary conversion treatment to Background Bg, obtain background binary figure MG (as shown in Figure 2), its calculating formula is as follows:

Wherein, T is binary-state threshold, and in the present embodiment, T value is 128; MG (x, y) represents the value of pixel (x, y) in background binary figure MG, MG (x, y)=0 represents that pixel (x, y) is possible car door region, MG (x, y)=1 represents that pixel (x, y) is non-car door region.

Proceed to step S4.

Step S4: closing operation of mathematical morphology is carried out to the background binary figure MG that step S3 obtains.Background binary figure MG is carried out to the closed operation of mathematical morphology, obtain gradient connection layout OG (as shown in Figure 3), its calculating formula is as follows:

Wherein, Se is structural element, gets the square structure element of 8 × 8 in the present embodiment; MG ο Se represents that background binary figure MG is carried out closed operation operation by construction operator Se; ο represents closing operation of mathematical morphology; ⊕ represents dilation operation; represent erosion operation.

Proceed to step S5.

Step S5: morphology opening operation is carried out to the gradient connection layout OG that step S4 obtains.Carry out mathematical morphology open operator to gradient connection layout OG, obtain gradient map CG (as shown in Figure 4), to eliminate the burr in gradient connection layout OG, its calculating formula is as follows:

$\mathrm{CG}=\mathrm{OG}\·\mathrm{Se}=(\mathrm{OG}\⊗\mathrm{Se})\⊕\mathrm{Se}---\left(5\right)$

Wherein, represent morphology opening operation, OGSe represents that gradient connection layout OG is carried out opening operation operation by construction operator Se.

Proceed to step S6.

Step S6: carry out negate and star process of zero-suppressing to the gradient map CG that step S5 obtains, add up the pixel count in each connection district, extracts the maximum connected region of pixel count as car door Prototype drawing MB (as shown in Figure 5).

Proceed to step S7.

Step S7: the Background Bg that the car door Prototype drawing MB obtained by step S6 and step S2 obtains carries out dot product, obtain background detection figure ej (as shown in Figure 6), computing formula is as follows:

ej＝MB·×Bg (6)

Wherein, × represent point multiplication operation symbol; MB × Bg represents that car door Prototype drawing MB and Background Bg carries out point multiplication operation.

Proceed to step S8.

Step S8: the average gray value eb calculating background detection figure ej, its calculating formula is as follows:

$\mathrm{eb}=\frac{1}{45000}\underset{x=1}{\overset{180}{\mathrm{\Σ}}}\underset{y=1}{\overset{250}{\mathrm{\Σ}}}\mathrm{ej}(x,y)---\left(7\right)$

Wherein, ej (x, y) represents the gray-scale value of pixel (x, y) in background detection figure ej.

In the present embodiment, the average gray value eb=17.710 of the background detection figure ej calculated.

Proceed to step S9.

Step S9: the car door image after Real-time Collection vehicle sends, sample frequency is 0.5S/ frame, and the RBG picture frame (as shown in Figure 7) of the kth frame of collection, uses E _krepresent.In embodiment, get k=1.

Proceed to step S10.

Step S10: to the car door picture frame E gathered in step S9 _kaccording to the method for step S1, gray processing process is carried out to it, obtain the car door gray-scale map (as shown in Figure 8) of kth frame, be designated as Eh _k.

Proceed to step S11.

Step S11: the car door gray-scale map Eh that the car door Prototype drawing MB obtained by step S6 and step S10 obtains _kcarry out dot product, obtain car door detection figure e _k(as shown in Figure 9), computing formula is as follows:

e _k＝MB·×Eh _k(8)

Wherein, MB × Eh _krepresent car door Prototype drawing MB and car door gray-scale map Eh _kcarry out point multiplication operation.

Proceed to step S12.

Step S12: the calculation procedure S11 kth frame car door detection figure e obtained _kaverage gray value eh _k, its calculating formula is as follows:

${\mathrm{eh}}_k=\frac{1}{45000}\underset{x=1}{\overset{180}{\mathrm{\Σ}}}\underset{y=1}{\overset{250}{\mathrm{\Σ}}}{e}_k(x,y)---\left(9\right)$

Wherein, e _k(x, y) represents car door detection figure e _kthe gray-scale value of middle pixel (x, y).

In the present embodiment, kth frame car door detection figure e is calculated _kaverage gray value eh _k=31.284.

Proceed to step S13.

Step S13: calculate car door and detect figure e _kwith the gray difference degree ed of background detection figure ej _k, its calculating formula is as follows:

ed _k＝|eh _k-eb| (10)

In the present embodiment, car door detection figure e is calculated _kwith the gray difference degree ed of background detection figure ej _k=13.574.

Proceed to step S14.

Step S14: calculate car door and detect figure e _kwith contrast difference's degree r of background detection figure ej _k, its calculating formula is as follows:

$r_k=\frac{\underset{x=1}{\overset{180}{\mathrm{\Σ}}}\underset{y=1}{\overset{250}{\mathrm{\Σ}}}(e_k(x,y)-31.3)(\mathrm{ej}(x,y)-17.71)}{\sqrt{\left(\underset{x=1}{\overset{180}{\mathrm{\Σ}}}\underset{y=1}{\overset{250}{\mathrm{\Σ}}}{(e_k(x,y)-31.3)}^2\right)\left(\underset{x=1}{\overset{180}{\mathrm{\Σ}}}\underset{y=1}{\overset{250}{\mathrm{\Σ}}}{(\mathrm{ej}(x,y)-17.71)}^2\right)}}---\left(11\right)$

Wherein, r _kfor contrast difference's degree, its value is larger then illustrates that car door detects figure e _klarger with the diversity factor of background detection figure ej, on the contrary also anti-.

In the present embodiment, car door detection figure e is calculated _kwith contrast difference's degree r of background detection figure ej _k=0.4179.

Proceed to step S15.

Step S15: through type 12 judges opening/closing door of vehicle state.

Wherein, fl _kfor the mark of opening/closing door of vehicle state, fl _k=0 represents that car door is closed condition, fl _k=1 represents that car door is opening.In the present embodiment, gray difference degree threshold value T ₁value is 5; Contrast difference's degree threshold value T ₂value is-0.25.

In the present embodiment, due to ed _k=13.574>=5, and r _k=0.4179>=-0.25, can fl be obtained _k=1, thus judge that car door is now as opening.

Proceed to step S16.

Step S16: the opening/closing door of vehicle state storing the car door picture frame of kth frame.

Proceed to step S17.

Step S17: judge whether public transit vehicle reaches home.No, proceed to step S18; Then proceed to step S19.

Step S18: gather next frame, now k=k+1, proceed to step S9.

According to above technical scheme of the present invention, from working time and cost analysis two aspect, the relative merits of contrast the present invention program and traditional manual measurement method.

(1) working time.Be 15 minutes with a duration, totalframes be 21076 video file be example, common manual detection roughly needs to be approximately 15 minutes.Simulation process platform of the present invention is: Intel I3M350 processor, and the computing machine of 2GB internal memory utilizes MATLAB software to emulate, and measures the 281MB video file collected, the time used be 3 points 25 seconds.Visible, the processing speed of the method proposed in the present invention has obvious advantage.

(2) cost analysis.Manual measurement method needs the opening/closing door of vehicle situation in long-time continuous observation video, needs the manual work cost of at substantial, and easily causes the visual fatigue of personnel.And the method that the present invention proposes, can realize hyperchannel detects simultaneously, improves work efficiency while reducing cost of labor.In addition, because video capture device all installed by current most bus, the method that the present invention proposes can install video capture device additional, directly can also utilize existing video monitoring equipment acquisition of image data, can save the cost of invention.

Claims (10)

1. a public transit vehicle opening/closing door of vehicle state automatic testing method, is characterized in that, comprise the steps:

Step 0: gather bus and to dispatch a car the video background picture frame F of front door _t;

Step 1: video background image frame F before the bus of collection is dispatched a car _tgray processing, obtains background grey-level image frame f _t;

Step 2: obtain Background Bg by averaging method in background grey-level image frame;

Step 3: binaryzation is carried out to the Background Bg that step 2 obtains, obtains background binary figure MG;

Step 4: carry out closing operation of mathematical morphology to the background binary figure MG that step 3 obtains, obtains gradient connection layout OG;

Step 5: morphology opening operation is carried out to the gradient connection layout OG that step 4 obtains, obtains gradient map CG;

Step 6: carry out negate and star process of zero-suppressing to the gradient map CG that step 5 obtains, add up the pixel count in each connection district, extracts the maximum connected region of pixel count as car door Prototype drawing MB;

Step 7: the Background Bg that car door Prototype drawing MB step 6 obtained and step 2 obtain carries out dot product, obtain background detection figure ej, computing formula is as follows:

ej＝MB·×Bg (6)

Wherein, × represent point multiplication operation symbol;

Step 8: the average gray value eb calculating background detection figure ej;

Step 9: the car door picture frame E after Real-time Collection vehicle sends _k;

Step 10: to the car door picture frame E gathered in step 9 _kaccording to the method for step 1, gray processing process is carried out to it, obtain the car door gray-scale map of kth frame, be designated as Eh _k;

Step 11: the car door gray-scale map Eh that car door Prototype drawing MB step 6 obtained and step 10 obtain _kcarry out dot product, obtain kth frame car door detection figure e _k, computing formula is as follows:

e _k＝MB·×Eh _k(8)

Step 12: the kth frame car door detection figure e obtained in calculation procedure 11 _kaverage gray value eh _k;

Step 13: calculate car door and detect figure e _kwith the gray difference degree ed of background detection figure ej _k, its calculating formula is as follows:

ed _k＝|eh _k-eb| (10)

Step 14: calculate car door and detect figure e _kwith contrast difference's degree r of background detection figure ej _k;

Step 15: opening/closing door of vehicle state judges; Opening/closing door of vehicle state through type (12) judges;

Wherein, fl _kfor the mark of opening/closing door of vehicle state, fl _k=0 represents that car door is closed condition, fl _k=1 represents that car door is opening; T ₁for gray difference degree threshold value, T ₁span be: 3 ~ 10; T ₂for contrast difference's degree threshold value, its span is-0.5 ~ 0.

2. a kind of public transit vehicle opening/closing door of vehicle state automatic testing method according to claim 1, is characterized in that, also comprise the steps:

Step 16: the opening/closing door of vehicle state storing kth frame car door picture frame;

Step 17: whether public transit vehicle reaches home; No, proceed to step 18; Then proceed to step 19;

Step 18: gather next frame, proceed to step 9.

3. a kind of public transit vehicle opening/closing door of vehicle state automatic testing method according to claim 1 and 2, is characterized in that, in step 0: the video background picture frame of 5s before collection bus sets out, wherein, sample frequency is 0.5s/ frame, gathers 10 two field pictures altogether, is designated as F _t, wherein, t is frame number, t=1,2 ..., 10; The image size collected is M capable and N row, and set the coordinate of certain pixel as (x, y), x and y represents the row and column of pixel (x, y), satisfied: x and y is integer, and 1≤x≤M, 1≤y≤N, then t frame background image frame F _tthe RGB color value symbolically of middle pixel (x, y) is (R _t(x, y), G _t(x, y), B _t(x, y)).

4. a kind of public transit vehicle opening/closing door of vehicle state automatic testing method according to claim 3, is characterized in that, in step 1: step 0 is processed the t frame background image frame F obtained _tcarry out gray processing process, obtain background grey-level image frame and be designated as f _t, its calculating formula is as follows:

f _t(x,y)＝0.3×R _t(x,y)+0.59×G _t(x,y)+0.11×B _t(x,y) (1)

Wherein, R _t(x, y) represents background image frame F _tthe R component value of middle pixel (x, y); G _t(x, y) represents background image frame F _tthe G component value of middle pixel (x, y); B _t(x, y) represents background image frame F _tthe B component value of middle pixel (x, y); f _t(x, y) represents background grey-level image frame f _tthe gray-scale value of middle pixel (x, y).

5. a kind of public transit vehicle opening/closing door of vehicle state automatic testing method according to claim 4, is characterized in that, in step 2: the computing formula of Background Bg is shown below:

$\mathrm{Bg}(x,y)=\frac{1}{10}\underset{t=1}{\overset{10}{\mathrm{\Σ}}}{f}_t(x,y)---\left(2\right)$

Wherein, Bg (x, y) represents the gray-scale value of pixel (x, y) in Background Bg;

In step 3: carry out binaryzation to Background Bg and obtain background binary figure MG, its calculating formula is as follows:

Wherein, T is binary-state threshold, and its value is: 110 ~ 140; MG (x, y) represents the value of pixel (x, y) in background binary figure MG, MG (x, y)=0 represents that pixel (x, y) is possible car door region, MG (x, y)=1 represents that pixel (x, y) is non-car door region.

6. a kind of public transit vehicle opening/closing door of vehicle state automatic testing method according to claim 5, it is characterized in that, in step 4: closed operation background binary figure MG being carried out to mathematical morphology, obtain gradient connection layout OG, its calculating formula is as follows:

Wherein, Se is structural element, gets the square structure element of 3 × 3 ~ 10 × 10; MG o Se represents that background binary figure MG is carried out closed operation operation by construction operator Se; o represents closing operation of mathematical morphology; ⊕ represents dilation operation; represent erosion operation.

7. a kind of public transit vehicle opening/closing door of vehicle state automatic testing method according to claim 6, it is characterized in that, in step 5: mathematical morphology open operator is carried out to gradient connection layout OG, obtains gradient map CG, to eliminate the burr in gradient connection layout OG, its calculating formula is as follows:

$\mathrm{CG}=\mathrm{OG}\·\mathrm{Se}=(\mathrm{OG}\⊗\mathrm{Se})\⊕\mathrm{Se}---\left(5\right)$

Wherein, represent morphology opening operation, OGSe represents that gradient connection layout OG is carried out opening operation operation by construction operator Se.

8. a kind of public transit vehicle opening/closing door of vehicle state automatic testing method according to claim 3, is characterized in that, in step 8: the calculating formula of average gray value eb is as follows:

$\mathrm{eb}=\frac{1}{\mathrm{MN}}\underset{x=1}{\overset{M}{\mathrm{\Σ}}}\underset{y=1}{\overset{N}{\mathrm{\Σ}}}\mathrm{ej}(x,y)---\left(7\right)$

Wherein, ej (x, y) represents the gray-scale value of pixel (x, y) in background detection figure ej.

9. a kind of public transit vehicle opening/closing door of vehicle state automatic testing method according to claim 3, is characterized in that, in step 12: average gray value eh _kcalculating formula as follows:

${\mathrm{eh}}_k=\frac{1}{\mathrm{MN}}\underset{x=1}{\overset{M}{\mathrm{\Σ}}}\underset{y=1}{\overset{N}{\mathrm{\Σ}}}{e}_k(x,y)---\left(9\right)$

Wherein, e _k(x, y) represents kth frame car door detection figure e _kthe gray-scale value of middle pixel (x, y).

10. a kind of public transit vehicle opening/closing door of vehicle state automatic testing method according to claim 3, is characterized in that, in step 14: contrast difference's degree r _kcalculating formula as follows:

$r_k=\frac{\underset{x=1}{\overset{M}{\mathrm{\Σ}}}\underset{y=1}{\overset{N}{\mathrm{\Σ}}}(e_k(x,y)-{\mathrm{eh}}_k)(\mathrm{ej}(x,y)-\mathrm{eb})}{\sqrt{\left(\underset{x=1}{\overset{M}{\mathrm{\Σ}}}\underset{y=1}{\overset{N}{\mathrm{\Σ}}}{(e_k,(x,y)-{\mathrm{eh}}_k)}^2\right)\left(\underset{x=1}{\overset{M}{\mathrm{\Σ}}}\underset{y=1}{\overset{N}{\mathrm{\Σ}}}{(\mathrm{ej}(x,y)-\mathrm{rb})}^2\right)}}---\left(11\right)$

Wherein, r _kfor contrast difference's degree, its value is larger then illustrates that car door detects figure e _klarger with the diversity factor of background detection figure ej, on the contrary also anti-.