CN107545225A

CN107545225A - A kind of method, apparatus and electronic equipment for detecting vehicle carried driving person's unlawful practice

Info

Publication number: CN107545225A
Application number: CN201610540740.3A
Authority: CN
Inventors: 朱丽
Original assignee: Hangzhou Hikvision Digital Technology Co Ltd
Current assignee: Hangzhou Hikvision Digital Technology Co Ltd
Priority date: 2016-06-23
Filing date: 2016-07-06
Publication date: 2018-01-05
Anticipated expiration: 2036-07-06
Also published as: CN107545225B

Abstract

Embodiments of the invention disclose a kind of method, apparatus and electronic equipment for detecting vehicle carried driving person's unlawful practice, are related to traffic violations detection technique, can lift the accuracy of vehicle carried driving person's unlawful practice detection.The method of detection vehicle carried driving person's unlawful practice includes：In the image of shooting, steering wheel is identified；The unlawful practice region of vehicle carried driving person is determined according to the steering wheel identified；Using the unlawful practice region of determination, the unlawful practice of vehicle carried driving person is detected.Described device includes being used for the module for realizing methods described.The present invention is applied to carry out unlawful practice detection to the vehicle carried driving person of public transport.

Description

A kind of method, apparatus and electronic equipment for detecting vehicle carried driving person's unlawful practice

Technical field

The present invention relates to traffic violations detection technique, more particularly to a kind of method for detecting vehicle carried driving person's unlawful practice, Device and electronic equipment.

Background technology

Safety is the most basic demand of people's trip, wherein, the most important condition gone on a journey safely is to need vehicle carried driving person to abide by The relevant regulations for keeping traffic law are driven, for example, fastening seat belts, not taking phone in driving procedure, in driver's cabin Interior non-smoking etc..Thus, operation behavior of the vehicle carried driving person in driving procedure how is monitored, detects whether it has unlawful practice Etc. dangerous driving behavior, turn into the important research focus gone on a journey safely.

With developing rapidly for computer vision technique, embedded technology, the network communications technology and camera technique, utilize The vehicle-mounted driver's unlawful practice of intelligent transportation system (ITS, Intelligent Transportation System) automatic detection Have become a study hotspot in current intelligent transportation.

Analyzed more than, dangerous driving behavior is to cause a traffic accident that one of most important reason occurs, and how to utilize skill Art means, effectively prevent that harm is extremely urgent caused by dangerous driving behavior.Public transport vehicle carried driving person makees The main security director that is taken public transport for oneself and people trip is, it is necessary to by covenant of works beam, during driving Should not have do not fasten the safety belt, phone with mobile telephone, the hazardous act such as smoking.

At present, for the detection of vehicle carried driving person's unlawful practice, mainly include：Safety belt detection, take phone detection with And smoking detection, wherein, safety belt detection is to carry out straight-line detection using Hough transform, specific as follows：It is first depending on shooting The position on video images detection vehicle carried driving person head, is extended according to head position, and be expanded region, in extended area It is middle determine safety belt may where region, then using Probabilistic Hough Transform algorithm, it is determined that safety belt may where Region carries out safety belt detection.

Phone detection is taken, on the basis of the position on vehicle carried driving person head is detected, is calculated using Adaboost iteration Method, the position of mobile phone in the video image of shooting is detected, if it is determined that mobile phone is in vehicle carried driving person's ear Near, then judge that the car is in phone in driver and takes state.But this method, detected and moved using Adaboost iterative algorithms Phone, because different illumination, different cameral exposure parameter, vehicle carried driving person's difference take the posture of phone, different model movement , it is necessary to the sample training Adaboost detectors of a large amount of mobile phones in varied situations be carried out, if sample when phone This is inadequate, and the accuracy of testing result is relatively low.

Smoking detection is similar with taking phone detection, and the position on head is examined using Adaboost iterative algorithms Survey, if the cigarette of strip feature is near vehicle carried driving person's mouth, then judge that the vehicle carried driving person is in smoking state.Should Method equally exists the relatively low technical problem of accuracy of testing result.

This method is when applied to public transport scene, if the volume of the flow of passengers is larger and when passenger is more, in shooting Multiple heads are may detect that in video image, so as to which the positioning to vehicle carried driving person forms interference, and then have influence on safety belt Detect, take the degree of accuracy of the unlawful practice detection such as phone detection and smoking detection so that the degree of accuracy of unlawful practice detection It is relatively low.

The content of the invention

In view of this, the embodiment of the present invention provides a kind of method, apparatus and electronics for detecting vehicle carried driving person's unlawful practice Equipment, the accuracy of vehicle carried driving person's unlawful practice detection can be lifted, to solve existing detection vehicle carried driving person row in violation of rules and regulations For method in, under the scene that multiple heads are detected in the video image of shooting, vehicle carried driving person's unlawful practice detection The problem of degree of accuracy is relatively low.

In a first aspect, the embodiment of the present invention provides a kind of method for detecting vehicle carried driving person's unlawful practice, including：

In the image of shooting, steering wheel is identified；

The unlawful practice region of vehicle carried driving person is determined according to the steering wheel identified；

Using the unlawful practice region of determination, the unlawful practice of vehicle carried driving person is detected.

With reference in a first aspect, in the first embodiment of first aspect, described in the image of shooting, the side of identifying Include to disk：

Rim detection is carried out to described image, obtains edge-detected image；

In the edge-detected image, steering wheel profile is detected using Probabilistic Hough Transform algorithm.

It is described to described in second of embodiment of first aspect with reference to the first embodiment of first aspect Image carries out rim detection, and obtaining edge-detected image includes：

Described image is converted into gray scale image, the gray scale image is handled using Canny edge detection algorithms, Obtain edge-detected image.

It is described using general in the third embodiment of first aspect with reference to the first embodiment of first aspect Rate Hough transform algorithm detects that steering wheel profile includes：

The circular arc in the edge-detected image is extracted, using the circular arc of Probabilistic Hough Transform algorithm Detection and Extraction, and is thrown Shadow counts the votes in Hough spaces into same Hough spaces, chooses the most point of votes as the center of circle；

The center of circle is calculated to the distance of circular arc, obtains radius；

If the radius being calculated meets the steering wheel radius threshold pre-set, the center of circle is the steering wheel center of circle, The radius is steering wheel radius.

With reference in a first aspect, in the 4th kind of embodiment of first aspect, the unlawful practice region is vehicle carried driving For member's shoulder to the region of huckle, the steering wheel that the foundation identifies determines that the unlawful practice region of vehicle carried driving person includes：

The steering wheel or the boundary rectangle of the steering wheel that will identify that translate up the first distance threshold pre-set, Vehicle carried driving person's shoulder is obtained to the region of huckle.

With reference to the 4th kind of embodiment of first aspect, in the 5th kind of embodiment of first aspect, the detection car The unlawful practice for carrying driver detects for safety belt, and the safety belt detection includes：

According to the vehicle carried driving person's shoulder identified to the region of huckle, safety belt detection zone is determined；

It is determined that safety belt detection zone at, detection vehicle carried driving person whether fasten seat belts.

With reference to the 5th kind of embodiment of first aspect, in the 6th kind of embodiment of first aspect, it is described it is determined that Safety belt detection zone at, detect whether that safety belt be present includes：

It is determined that safety belt detection zone in, according to brightness of image, obtain adaptive Canny threshold values；

According to the Canny threshold values of acquisition, Canny rim detections are carried out to safety belt detection zone, obtain safety belt detection Edge image in region；

In edge image in obtained safety belt detection zone, using Probabilistic Hough Transform algorithm, the peace is detected Entirely with the straightway in the edge image in detection zone；

According to the straightway detected, the distance between straight line slope over 10 and each straightway are calculated, if two straightways Slope is approximately the same, and the distance between two straightway is less than the safety belt distance threshold pre-set, determines vehicle carried driving Member fastens seat belts；It is approximately the same if there is no two straight line slope over 10, and the distance between two straightway is not less than advance The safety belt distance threshold of setting, determines that vehicle carried driving person does not fasten seat belts.

With reference in a first aspect, in the 7th kind of embodiment of first aspect, the unlawful practice region is vehicle carried driving Member's head zone, the steering wheel that the foundation identifies determine that the unlawful practice region of vehicle carried driving person includes：

The steering wheel that will identify that obtains scaled down version steering wheel, by the scaled down version according to the scale smaller pre-set Steering wheel translates up the second distance threshold value pre-set, obtains vehicle carried driving person's head zone；Or it will identify that Steering wheel boundary rectangle according to the scale smaller pre-set, obtain scaled down version boundary rectangle, the scaled down version is external Rectangle translates up the second distance threshold value pre-set, obtains vehicle carried driving person's head zone.

With reference to the 7th kind of embodiment of first aspect, in the 8th kind of embodiment of first aspect, the detection car Carry driver unlawful practice for take phone detection, it is described take phone detection include：

It is determined that vehicle carried driving person's head zone, detect whether mobile phone be present.

With reference to the 8th kind of embodiment of first aspect, in the 9th kind of embodiment of first aspect, it is described it is determined that Vehicle carried driving person's head zone, detect whether that mobile phone be present includes：

According to the hand textural characteristics storehouse pre-set, it is special to extract the hand texture included in vehicle carried driving person's head zone Sign；

The disaggregated model that the hand textural characteristics input of extraction is pre-set, if the disaggregated model output result table It is bright hand-characteristic to be present, confirm that vehicle carried driving person takes phone, if the disaggregated model output result shows hand not be present Feature, confirm that vehicle carried driving person does not take phone.

With reference to the 7th kind of embodiment of first aspect, in the tenth kind of embodiment of first aspect, the detection car The unlawful practice for carrying driver detects for smoking, and the smoking detection includes：

According to the steering wheel and vehicle carried driving person's head zone, smoking detection zone is determined；

It is determined that smoking detection zone, detect whether smoking phenomenon be present.

It is described true in a kind of the tenth embodiment of first aspect with reference to the tenth kind of embodiment of first aspect Fixed smoking detection zone, detect whether that smoking phenomenon be present includes：

Detect the imperfect profile included in the steering wheel profile；

Connect first line in the imperfect profile starting point and the center of circle, and, connect the imperfect profile stop with Second line in the center of circle；

If the angle between the first line and the second line exceedes the angle threshold pre-set, extraction is described imperfect The hand textural characteristics of cigarette are included in profile；

The smoking identification and classification model that the input of the hand textural characteristics comprising cigarette of extraction is pre-set, if described Smoking identification and classification model output result shows cigarette in hand-characteristic be present, confirms that vehicle carried driving person has smoking phenomenon, such as Smoking identification and classification model output result shows cigarette in hand-characteristic not be present described in fruit, confirms that vehicle carried driving person is not present and inhales Cigarette phenomenon.

It is described true in the 12nd kind of embodiment of first aspect with reference to the tenth kind of embodiment of first aspect Fixed smoking detection zone, detect whether that smoking phenomenon be present includes：

Extract the mouth region in vehicle carried driving person's head zone；

Texture feature extraction is carried out to mouth region, if including elongated rectangular area in the textural characteristics of extraction, Confirm vehicle carried driving person's smoking, if not including elongated rectangular area in the textural characteristics of extraction, confirm vehicle carried driving person Non- smoking.

With reference to first aspect, first aspect the first to any embodiment in the 12nd kind, in first aspect In 13rd kind of embodiment, methods described also includes：

In the image of shooting, head image is identified, is matched with the head image storehouse pre-set, is obtained and know Vehicle carried driving person's head image that the head image not gone out matches, according to vehicle carried driving person head figure in the head image storehouse As the mapping relations with vehicle carried driving person's information, vehicle carried driving person's information corresponding to the head image identified is obtained；

According to obtained vehicle carried driving person's information, prompt, and/or, record vehicle-mounted corresponding to vehicle carried driving person's information The unlawful practice of driver.

Second aspect, the embodiment of the present invention provide a kind of device for detecting vehicle carried driving person's unlawful practice, including：Steering wheel Identification module, unlawful practice area determination module and unlawful practice detection module, wherein,

Steering wheel identification module, in the image of shooting, identifying steering wheel；

Unlawful practice area determination module, for determining the unlawful practice area of vehicle carried driving person according to the steering wheel identified Domain；

Unlawful practice detection module, for using the unlawful practice region determined, detecting the unlawful practice of vehicle carried driving person.

With reference to second aspect, in the first embodiment of second aspect, the steering wheel identification module includes：Edge Detection unit, contour detecting unit and steering wheel acquiring unit, wherein,

Edge detection unit, for carrying out rim detection to described image, obtain edge-detected image；

Contour detecting unit, in the edge-detected image, outgoing direction to be detected using Probabilistic Hough Transform algorithm Plate wheel is wide；

Steering wheel acquiring unit, for obtaining the steering wheel of steering wheel profile described in inscribe.

With reference to the first embodiment of second aspect, in second of embodiment of second aspect, the edge inspection Surveying unit includes：Conversion subunit and rim detection subelement, wherein,

Conversion subunit, for described image to be converted into gray scale image；

Rim detection subelement, for the gray scale image to be handled using Canny edge detection algorithms, obtain side Edge detection image.

With reference to the first embodiment of second aspect, in the third embodiment of second aspect, the profile inspection Surveying unit includes：Center of circle determination subelement, radius determination subelement and outline specifying unit, wherein,

Center of circle determination subelement, for extracting the circular arc in the edge-detected image, utilize Probabilistic Hough Transform algorithm The circular arc of Detection and Extraction, and project in same Hough spaces, the votes in statistics Hough spaces, it is most to choose votes Point as the center of circle；

Radius determination subelement, for calculating the center of circle to the distance of circular arc, obtain radius；

Outline specifying unit, if the radius being calculated meets the steering wheel radius threshold pre-set, the center of circle For the steering wheel center of circle, the radius is steering wheel radius.

With reference to second aspect, in the 4th kind of embodiment of second aspect, the unlawful practice area determination module bag Include：

First shift unit, steering wheel or the boundary rectangle of the steering wheel for will identify that translate up to be set in advance The first distance threshold put, vehicle carried driving person's shoulder is obtained to the region of huckle.

With reference to the 4th kind of embodiment of second aspect, in the 5th kind of embodiment of second aspect, the violation row Include for detection module：Safety belt detection zone determining unit and safety belt detection unit, wherein,

Safety belt detection zone determining unit, for identifying vehicle carried driving person's shoulder to the region of huckle, according to identification The vehicle carried driving person's shoulder gone out determines safety belt detection zone to the region of huckle；

Safety belt detection unit, for it is determined that safety belt detection zone at, detection vehicle carried driving person whether fasten peace Full band.

With reference to the 5th kind of embodiment of second aspect, in the 6th kind of embodiment of second aspect, the safety belt Detection unit includes：It is single that threshold value obtains subelement, rim detection subelement, line segment detection subelement and safety belt detection Member, wherein,

Threshold value obtain subelement, for it is determined that safety belt detection zone in, according to brightness of image, obtain adaptive Canny threshold values；

Rim detection subelement, for the Canny threshold values according to acquisition, Canny edges are carried out to safety belt detection zone Detection, obtains the edge image in safety belt detection zone；

Line segment detection subelement, in the edge image in obtained safety belt detection zone, using probability Hough transform algorithm, detect the straightway in the edge image in the safety belt detection zone；

Safety belt detection sub-unit, for according to the straightway that detects, calculate straight line slope over 10 and each straightway it Between distance, if two straight line slope over 10 are approximately the same, and the distance between two straightway is less than the safety belt that pre-sets Distance threshold, determine that vehicle carried driving person fastens seat belts；It is approximately the same if there is no two straight line slope over 10, and two straightway The distance between not less than the safety belt distance threshold that pre-sets, determine that vehicle carried driving person does not fasten seat belts.

With reference to second aspect, in the 7th kind of embodiment of second aspect, the unlawful practice area determination module bag Include：

Second shift unit, the steering wheel for will identify that obtain scaled down version side according to the scale smaller pre-set To disk, the scaled down version steering wheel is translated up to the second distance threshold value pre-set, obtains the vehicle carried driving person head Region；Or it is external to obtain scaled down version according to the scale smaller pre-set for the boundary rectangle of the steering wheel for will identify that Rectangle, the scaled down version boundary rectangle is translated up to the second distance threshold value pre-set, obtain vehicle carried driving person's head Portion region.

With reference to the 7th kind of embodiment of second aspect, in the 8th kind of embodiment of second aspect, the violation row Include for detection module：Head zone recognition unit and mobile phone detection unit, wherein,

Head zone recognition unit, for identifying steering wheel, vehicle carried driving person head is determined according to the steering wheel identified Region；

Mobile phone detection unit, for it is determined that vehicle carried driving person's head zone, detect whether mobile phone be present.

With reference to the 8th kind of embodiment of second aspect, in the 9th kind of embodiment of second aspect, the mobile electricity Words detection unit includes：Hand texture feature extraction subelement and mobile phone detection sub-unit, wherein,

Hand texture feature extraction subelement, for according to the hand textural characteristics storehouse pre-set, extracting vehicle carried driving The hand textural characteristics included in member's head zone；

Mobile phone detection sub-unit, for the hand textural characteristics of extraction to be inputted into the disaggregated model pre-set, such as Disaggregated model output result shows hand-characteristic be present described in fruit, confirms that vehicle carried driving person takes phone, if the classification mould Type output result shows hand-characteristic not be present, confirms that vehicle carried driving person does not take phone.

With reference to the 7th kind of embodiment of second aspect, in the tenth kind of embodiment of second aspect, the violation row Include for detection module：Head zone recognition unit, smoking detection zone determination subelement and smoking detection unit, wherein,

Smoking detection zone determination subelement, for according to the steering wheel and vehicle carried driving person's head zone, Determine smoking detection zone；

Smoking detection unit, for it is determined that smoking detection zone, detect whether smoking phenomenon be present.

With reference to the tenth kind of embodiment of second aspect, in a kind of the tenth embodiment of second aspect, the smoking Detection unit includes：Discontinuous detection sub-unit, connection subelement, extraction subelement and smoking detection sub-unit, wherein,

Discontinuous detection sub-unit, for detecting the imperfect profile included in the steering wheel profile；

Subelement is connected, for connecting the first line of the imperfect profile starting point and the center of circle, and, described in connection not Integrity profile stop and second line in the center of circle；

Subelement is extracted, if the angle between the first line and the second line exceedes the angle threshold pre-set, is carried Take the hand textural characteristics that cigarette is included in the imperfect profile；

Smoking detection sub-unit, the smoking for the input of the hand textural characteristics comprising cigarette of extraction to be pre-set are sentenced Other disaggregated model, if the smoking identification and classification model output result shows cigarette in hand-characteristic be present, confirm vehicle-mounted drive There is smoking phenomenon in the person of sailing, if the smoking identification and classification model output result shows cigarette in hand-characteristic not be present, really Recognize vehicle carried driving person and smoking phenomenon not be present.

With reference to a kind of the tenth embodiment of second aspect, in the 12nd kind of embodiment of second aspect, the suction Cigarette detection unit includes：Mouth determination subelement and smoking determination subelement, wherein,

Mouth determination subelement, for extracting the mouth region in vehicle carried driving person's head zone；

Smoking determination subelement, for carrying out texture feature extraction to mouth region, if wrapped in the textural characteristics of extraction Containing elongated rectangular area, vehicle carried driving person's smoking is confirmed, if not including elongated rectangle in the textural characteristics of extraction Region, confirm the non-smoking of vehicle carried driving person.

With reference to second aspect, second aspect the first to any embodiment in the 12nd kind, in second aspect In 13rd kind of embodiment, described device also includes：Vehicle carried driving person's data obtaining module and unlawful practice processing module, Wherein,

Vehicle carried driving person's data obtaining module, in the image of shooting, identifying head image, with pre-setting Head image storehouse is matched, vehicle carried driving person's head image that the head image for obtaining and identifying matches, according to described in Vehicle carried driving person's head image and the mapping relations of vehicle carried driving person's information in head image storehouse, obtain the head identified Vehicle carried driving person's information corresponding to image；

Unlawful practice processing module, and/or, record described vehicle-mounted for according to obtained vehicle carried driving person's information, prompting The unlawful practice of vehicle carried driving person corresponding to driver information.

The third aspect, the embodiment of the present invention provide a kind of electronic equipment, and the electronic equipment includes：Housing, processor, deposit Reservoir, circuit board and power circuit, wherein, circuit board is placed in the interior volume that housing surrounds, and processor and memory are set On circuit boards；Power circuit, for each circuit or the device power supply for above-mentioned electronic equipment；Memory is used to store and can hold Line program code；The executable program code that processor is stored by reading in memory is run and executable program code pair The program answered, the method for performing foregoing any described detection vehicle carried driving person's unlawful practice.

A kind of method, apparatus and electronic equipment for detecting vehicle carried driving person's unlawful practice provided in an embodiment of the present invention, leads to Cross in the image of shooting, identify steering wheel；The unlawful practice region of vehicle carried driving person is determined according to the steering wheel identified； Using the unlawful practice region of determination, the unlawful practice of vehicle carried driving person is detected, vehicle carried driving person's unlawful practice inspection can be lifted The accuracy of survey, to solve in the method for existing detection vehicle carried driving person's unlawful practice, detected in the video image of shooting To under the scene on multiple heads, vehicle carried driving person's unlawful practice detection the degree of accuracy it is relatively low the problem of.

Brief description of the drawings

In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing There is the required accompanying drawing used in technology description to be briefly described, it should be apparent that, drawings in the following description are only this Some embodiments of invention, for those of ordinary skill in the art, on the premise of not paying creative work, can be with Other accompanying drawings are obtained according to these accompanying drawings.

Fig. 1 is the method flow schematic diagram that embodiments of the invention one detect vehicle carried driving person's unlawful practice；

Fig. 2 is the apparatus structure schematic diagram that embodiments of the invention two detect vehicle carried driving person's unlawful practice；

Fig. 3 is the structural representation of electronic equipment one embodiment of the present invention.

Embodiment

The embodiment of the present invention is described in detail below in conjunction with the accompanying drawings.

It will be appreciated that described embodiment is only part of the embodiment of the present invention, rather than whole embodiments.Base Embodiment in the present invention, those of ordinary skill in the art obtained under the premise of creative work is not made it is all its Its embodiment, belongs to the scope of protection of the invention.

Fig. 1 is the method flow schematic diagram that embodiments of the invention one detect vehicle carried driving person's unlawful practice, such as Fig. 1 institutes Show, the method for the present embodiment can include：

Step 101, in the image of shooting, steering wheel is identified；

In the present embodiment, as an alternative embodiment, image is to utilize the vehicle-mounted phase for being arranged on precalculated position in driver's cabin Machine carries out the image that detection collects to vehicle carried driving person's behavior.

As another alternative embodiment, in-vehicle camera after detection is carried out to vehicle carried driving person's behavior and collects image, Using the communication module pre-set, image is sent to graphical analysis server, it is unified on each by graphical analysis server The image of biography is analyzed and processed.

In the present embodiment, as an alternative embodiment, in the image of shooting, identify that steering wheel includes：

A11, rim detection is carried out to described image, obtains edge-detected image；

In the present embodiment, as an alternative embodiment, using Canny edge detection algorithms, edge is carried out to described image Detection.Wherein, Canny edge detection algorithms are a kind of multistage edge detection operators, are set according to Image Edge-Detection optiaml ciriterion Meter, has good edge detection performance, more and more extensive application has been obtained in image procossing.It the advantage is that, Canny Edge detection algorithm is not susceptible to noise jamming, is able to detect that real weak edge；There is higher brightness in gradient image Gradient is more likely edge, and in order to accurately limit the Grad scope at edge, Canny edge detection algorithms pass through two Threshold value realizes that two different threshold values are respectively used to detect strong edge and weak edge, and when weak edge is connected with strong edge When, just weak edge is included in the output image.

As an alternative embodiment, rim detection is carried out to described image, obtaining edge-detected image includes：

In the present embodiment, as an alternative embodiment, after edge-detected image is obtained, obtained edge can also be examined Altimetric image is filtered processing.

A12, in the edge-detected image, steering wheel profile is detected using Probabilistic Hough Transform algorithm.

In the present embodiment, in the Canny edges in edge-detected image, the Canny edges of steering wheel are circular for one, tool There is obvious feature.

In the present embodiment, as an alternative embodiment, the circle in Canny edges is detected using Probabilistic Hough Transform algorithm Shape, if there is steering wheel, a fitting circle can be obtained.

In the present embodiment, steering wheel profile includes the steering wheel center of circle and steering wheel radius.As an alternative embodiment, profit Detect that steering wheel profile includes with Probabilistic Hough Transform algorithm：

In the present embodiment, as an alternative embodiment, steering wheel diameter is typically in 80~130 pixels.

In the present embodiment, as an alternative embodiment, this method can also include：

A13, obtain the steering wheel of steering wheel profile described in inscribe.

In the present embodiment, steering wheel is the external square box of a steering wheel.Detect to intend using Probabilistic Hough Transform algorithm After closing circle, the center of circle and fitting diameter of a circle of the fitting circle in the edge-detected image can be obtained, inscribe can be obtained The external square box of steering wheel of the steering wheel profile.So, it is easy to subsequent treatment.

Step 102, the unlawful practice region of vehicle carried driving person is determined according to the steering wheel identified；

In the present embodiment, as an alternative embodiment, because the circle of steering wheel has obvious feature, and driver is The personnel of sole operation steering wheel, there is more defined location relation with steering wheel.Thus, by orientation disk, and then It can determine vehicle carried driving person region, unlawful practice region to be included in vehicle carried driving person region.So, in public transport When middle passenger is more, by the detection of the image to shooting, the head image that detects may be caused more, it is vehicle-mounted so as to influence The positioning of driver, but in public transport, steering wheel only has one and has obvious feature, therefore, can basis The band of position of steering wheel determines vehicle carried driving person region, so as to lift the accuracy of the vehicle carried driving person of positioning, and then Lift the degree of accuracy of vehicle carried driving person's unlawful practice detection.

In the present embodiment, different unlawful practices, different unlawful practice regions is corresponding with.For example, disobeyed for safety belt The detection of rule behavior, corresponding unlawful practice region are vehicle carried driving person's shoulder to the region of huckle；Disobeyed for taking phone The detection of rule behavior, corresponding unlawful practice region are vehicle carried driving person's head zone；It is right for the detection of smoking unlawful practice The unlawful practice region answered includes：Steering wheel and vehicle carried driving person's head zone.Steering wheel can be entered according to advance setting After row scaling, then corresponding distance is translated up, obtain each unlawful practice region.

Step 103, using the unlawful practice region of determination, the unlawful practice of vehicle carried driving person is detected.

In the present embodiment, the unlawful practice of detection vehicle carried driving person includes：Safety belt detection, take phone detection and inhale Cigarette detects, wherein,

Unlawful practice region is vehicle carried driving person's shoulder to the region of huckle, and the steering wheel that the foundation identifies determines The unlawful practice region of vehicle carried driving person includes：

The safety belt detection includes：

B11, according to the vehicle carried driving person's shoulder identified to the region of huckle, determine safety belt detection zone；

In the present embodiment, according to the steering wheel identified, the i.e. external square box of steering wheel, upward direction translation is pre-set Distance, obtain safety belt detection zone (region of vehicle carried driving person's shoulder to huckle), as an alternative embodiment, in advance The distance of setting is the height of steering wheel, i.e., by the height of the external square box upward direction translation direction disk of steering wheel, obtains car Carry pilot harness's detection zone.

B12, it is determined that safety belt detection zone at, detection vehicle carried driving person whether fasten seat belts.

In the present embodiment, as an alternative embodiment, it is determined that safety belt detection zone at, detect whether safety be present Band includes：

B111, it is determined that safety belt detection zone in, according to brightness of image, obtain adaptive Canny threshold values；

B112, according to the Canny threshold values of acquisition, Canny rim detections are carried out to safety belt detection zone, obtain safety belt Edge image in detection zone；

In the present embodiment, as an alternative embodiment, the edge image in safety belt detection zone can be filtered Processing, for example, carrying out mean filter processing.

B113, in the edge image in obtained safety belt detection zone, using Probabilistic Hough Transform algorithm, detection The straightway in edge image in the safety belt detection zone；

B114, according to the straightway detected, the distance between straight line slope over 10 and each straightway are calculated, if two is straight Line segment slope is approximately the same, and the distance between two straightway is less than the safety belt distance threshold pre-set, determines vehicle-mounted Driver fastens seat belts；It is approximately the same if there is no two straight line slope over 10, and the distance between two straightway is not less than The safety belt distance threshold pre-set, determines that vehicle carried driving person does not fasten seat belts.

In the present embodiment, after straight line is detected, the line segment detected is carried out according to the characteristics of conventional belt type Filtering, filtering are not belonging to the straight line of safety belt, the straight-line detection result filtered, according to the straight-line detection result of filtering, from And obtain the conclusion finally whether fastened the safety belt.For example, if two straight line slope over 10 were approximately the same, i.e. the difference of two straight line slope over 10 In the range of the slope error pre-set, and between two straightway not less than the safety belt distance threshold pre-set, Show safety belt feature be present in the safety belt detection zone, determine that vehicle carried driving person has fastened seat belts, row in violation of rules and regulations be present For.

As an alternative embodiment, unlawful practice region is vehicle carried driving person's head zone, the side that the foundation identifies The unlawful practice region for determining vehicle carried driving person to disk includes：

The steering wheel that will identify that obtains scaled down version steering wheel, by the scaled down version according to the scale smaller pre-set Steering wheel translates up the second distance threshold value pre-set, obtains vehicle carried driving person's head zone.

It is described take phone detection include：

C11, vehicle carried driving person's head zone is determined according to the steering wheel identified；

In the present embodiment, according to obtained steering wheel, diminution processing (contracting is carried out to steering wheel according to the ratio pre-set As low as head zone size), and the steering wheel that diminution is obtained translates up the second distance threshold value pre-set, obtains vehicle-mounted Driver takes the region of phone detection, i.e. vehicle carried driving person's head zone.

Alternatively, the boundary rectangle for the steering wheel that can be will identify that obtains scaled down version according to the scale smaller pre-set Boundary rectangle, the scaled down version boundary rectangle is translated up to the second distance threshold value pre-set, obtains the vehicle carried driving Member's head zone.

In the present embodiment, second distance threshold value can be come according to the distance of each vehicle carried driving person's head distance steering wheel of statistics It is determined that.

C12, it is determined that vehicle carried driving person's head zone, detect whether mobile phone be present.

In the present embodiment, user is when taking phone, due to needing hand held mobile phone, thus, in vehicle carried driving person's head Portion's opening position, there can be obvious hand-characteristic, the hand of user can be as the target of detection.

In the present embodiment, as an alternative embodiment, it is determined that vehicle carried driving person's head zone, detect whether to exist and move Mobile phone includes：

C111, according to the hand textural characteristics storehouse pre-set, extract the hand included in vehicle carried driving person's head zone Textural characteristics；

C112, the disaggregated model that the hand textural characteristics input of extraction is pre-set, if the disaggregated model exports As a result show hand-characteristic be present, confirm that vehicle carried driving person takes phone, if the disaggregated model output result shows not deposit In hand-characteristic, confirm that vehicle carried driving person does not take phone.

In the present embodiment, first to Objective extraction feature, it is then sent through grader and is classified.As an optional implementation Example, textural characteristics include：Local binary patterns (LBP, Local Binary Pattern) feature, histograms of oriented gradients (HOG, Histogram of Oriented Gradient) feature, Haar features, EGL features etc..Disaggregated model connects for training The detection disaggregated model made a phone call, including：SVMs (SVM, Support Vector Machine) grader, nerve net Network grader etc..

As an alternative embodiment, unlawful practice region is steering wheel or vehicle carried driving person's head zone, and smoking detection is wrapped Include：

D11, vehicle carried driving person's head zone is determined according to the steering wheel identified；

D12, according to the steering wheel and vehicle carried driving person's head zone, determine smoking detection zone；

In the present embodiment, two smoking detection zones be present, first, steering wheel, in steering wheel, if smoking phenomenon, Cigarette can be clipped in vehicle carried driving person's hand；Second, vehicle carried driving person's head zone, if smoking phenomenon, cigarette can be contained in vehicle-mounted Driver is in the mouth.Therefore, smoking detection zone includes：Steering wheel and vehicle carried driving person's head zone.

D13, it is determined that smoking detection zone, detect whether smoking phenomenon be present.

In the present embodiment, smoking detection zone is steering wheel, it is determined that smoking detection zone, detect whether smoking be present Phenomenon includes：

D111, detect the imperfect profile included in the steering wheel profile；

D112, first line in the imperfect profile starting point and the center of circle is connected, and, connect the imperfect profile and stop Point and second line in the center of circle；

D113, if angle between the first line and the second line exceedes the angle threshold pre-set, described in extraction The hand textural characteristics of cigarette are included in imperfect profile；

D114, the smoking identification and classification model that the input of the hand textural characteristics comprising cigarette of extraction is pre-set, such as Smoking identification and classification model output result shows cigarette in hand-characteristic be present described in fruit, confirms that vehicle carried driving person has smoking and showed As if the smoking identification and classification model output result shows cigarette in hand-characteristic not be present, confirming vehicle carried driving person not Smoking phenomenon be present.

In the present embodiment, for steering wheel, the circle detected can be verified according to the method for foregoing described detection circle It is whether complete, if continuous imperfect region more than 40 degree, can directly judge position of the continuous imperfect region for hand.Example Such as, if being tested with continuous imperfect region on round edge, the of the continuous imperfect region starting point and the center of circle is connected One line, and, the continuous imperfect region stop and second line in the center of circle are connected, is not had between the first line and the second line Detect that Hough circles for continuous imperfect region, are pre-set if the angle between the first line and the second line exceedes Angle threshold, for example, 40 degree, then judge that there is hand-characteristic in the continuous imperfect region.

Then the continuous imperfect region Land use models recognition methods is classified, extraction LBP features etc., then utilized Neutral net is classified, and finally gives whether steering wheel has cigarette.

As another alternative embodiment, smoking detection zone is vehicle carried driving person's head zone, it is determined that smoking detection Region, detect whether that smoking phenomenon be present includes：

D211, extract the mouth region in vehicle carried driving person's head zone；

D212, texture feature extraction is carried out to mouth region, if including elongated rectangle in the textural characteristics of extraction Region, confirm vehicle carried driving person's smoking, if not including elongated rectangular area in the textural characteristics of extraction, confirm vehicle-mounted drive The non-smoking of the person of sailing.

In the present embodiment, for vehicle carried driving person's head zone, in the present embodiment, under the scene that cigarette is contained in the mouth, Cigarette has obvious contour feature, and therefore, the present embodiment is directed to vehicle carried driving person's head zone, directly carries out contours extract, Then elongated rectangular area is detected whether, if so, showing that vehicle carried driving person has smoking phenomenon.

As an alternative embodiment, this method also includes：

Step 104, in the image of shooting, head image is identified, is matched with the head image storehouse pre-set, Vehicle carried driving person's head image that the head image for obtaining and identifying matches, according to vehicle carried driving in the head image storehouse Member's head image and the mapping relations of vehicle carried driving person's information, obtain vehicle carried driving person corresponding to the head image identified Information.

In the present embodiment, matched by head image, it may be determined that specific vehicle carried driving person's information.

As another alternative embodiment, this method also includes：

Step 105, according to obtained vehicle carried driving person's information, prompt, and/or, record vehicle carried driving person's information pair The unlawful practice of the vehicle carried driving person answered.

, can be according to the processing plan pre-set after detecting that vehicle carried driving person has unlawful practice in the present embodiment Slightly, respective handling is carried out to vehicle carried driving person.For example, prompt message is sent to vehicle carried driving person, to prompt vehicle carried driving person to disobey Rule.

In the present embodiment, by checking the steering wheel in image, determined according to the band of position of steering wheel vehicle-mounted The band of position of driver；Then unlawful practice region delimited, whether detection vehicle carried driving person has unlawful practice.For example, according to Steering wheel, vehicle carried driving person head approximate region is determined, to the position of the range check hand, then detect whether to take phone； Or in the position of steering wheel and vehicle carried driving person's head zone detection hand, it is determined whether smoking.So as to be determined according to steering wheel The band of position of position vehicle carried driving person, the problem of public transport vehicle carried driving person positions can be efficiently solved, is prevented public Passenger excessively influences vehicle carried driving person's positioning result in the vehicles；Further, the vehicle carried driving person area according to registration Domain positioning unlawful practice region, can automatic detection go out do not fasten the safety belt, take the abnormal unlawful practice such as phone, smoking, obtain Comparatively ideal abnormal unlawful practice Detection results, the accuracy of vehicle carried driving person's unlawful practice detection is improved, for example, in car Driver head's region detection is carried, and then detects whether to take the method for phone, flase drop can be effectively reduced；Moreover, this implementation Example, to helping public transit system to establish effective supervisory system, reduces the workload of public transport supervision in public transportation system, Cost of labor, lifting operating efficiency are reduced, public transport vehicle carried driving person specification oneself behavior is supervised, safeguards vast trip Very big progradation is served in the safety of person, it is possible to prevente effectively from because the unlawful practice of vehicle carried driving person's exception and caused by hand over Interpreter's event, safety guarantee is provided for trip.

Fig. 2 is the apparatus structure schematic diagram that embodiments of the invention two detect vehicle carried driving person's unlawful practice, such as Fig. 2 institutes Show, the device of the present embodiment can include：Steering wheel identification module 21, unlawful practice area determination module 22 and unlawful practice Detection module 23, wherein,

Steering wheel identification module 21, in the image of shooting, identifying steering wheel；

In the present embodiment, image is to vehicle carried driving person's behavior using the in-vehicle camera installed in precalculated position in driver's cabin Carry out the image that detection collects.

As an alternative embodiment, steering wheel identification module 21 includes：Edge detection unit, contour detecting unit and side To disk acquiring unit (not shown), wherein,

In the present embodiment, using Canny edge detection algorithms, rim detection is carried out to described image.

In the present embodiment, as an alternative embodiment, edge detection unit includes：Conversion subunit and rim detection Unit, wherein,

In the present embodiment, as an alternative embodiment, contour detecting unit includes：Center of circle determination subelement, radius determine Subelement and outline specifying unit, wherein,

Unlawful practice area determination module 22, for determining the unlawful practice of vehicle carried driving person according to the steering wheel identified Region；

In the present embodiment, as an alternative embodiment, unlawful practice area determination module 22 includes：

First shift unit (not shown), for the steering wheel that will identify that or the boundary rectangle of the steering wheel to It is upper to translate the first distance threshold pre-set, vehicle carried driving person's shoulder is obtained to the region of huckle.

As another alternative embodiment, unlawful practice area determination module 22 includes：

Unlawful practice detection module 23, for utilizing the unlawful practice region determined, the violation row of detection vehicle carried driving person For.

In the present embodiment, the unlawful practice of detection vehicle carried driving person includes：Safety belt detection, take phone detection and inhale Cigarette detects.

As an alternative embodiment, unlawful practice detection module 23 includes：Safety belt detection zone determining unit and peace Full band detection unit (not shown), wherein,

In the present embodiment, as an alternative embodiment, safety belt detection unit includes：Threshold value obtains subelement, edge inspection Subelement, line segment detection subelement and safety belt detection sub-unit are surveyed, wherein,

As another alternative embodiment, unlawful practice detection module 23 includes：Head zone recognition unit and mobile electricity Detection unit is talked about, wherein,

In the present embodiment, according to obtained steering wheel, diminution processing is carried out to steering wheel according to the ratio pre-set, and The steering wheel that diminution obtains is translated up to the second distance threshold value pre-set, vehicle carried driving person is obtained and takes phone detection Region.

In the present embodiment, as an alternative embodiment, mobile phone detection unit includes：Hand texture feature extraction is single Member and mobile phone detection sub-unit, wherein,

In the present embodiment, as an alternative embodiment, textural characteristics include：Local binary patterns feature, direction gradient are straight Square figure feature, Haar features, EGL features etc..Disaggregated model includes：Support vector machine classifier, neural network classifier etc..

As yet another alternative embodiment, unlawful practice detection module 23 includes：Head zone recognition unit, smoking detection zone Domain determination subelement and smoking detection unit, wherein,

In the present embodiment, smoking detection unit includes：Discontinuous detection sub-unit, connection subelement, extraction subelement with And smoking detection sub-unit, wherein,

As another alternative embodiment, smoking detection zone is vehicle carried driving person's head zone, and smoking detection unit includes： Mouth determination subelement and smoking determination subelement, wherein,

In the present embodiment, as an alternative embodiment, the device can also include：Vehicle carried driving person's data obtaining module 24 And unlawful practice processing module 25, wherein,

Vehicle carried driving person's data obtaining module 24, in the image of shooting, identifying head image, with pre-setting Head image storehouse matched, vehicle carried driving person's head image that the head image for obtaining and identifying matches, according to institute Vehicle carried driving person's head image and the mapping relations of vehicle carried driving person's information in head image storehouse are stated, obtain the head identified Vehicle carried driving person's information corresponding to portion's image；

Unlawful practice processing module 25, and/or, the car is recorded for according to obtained vehicle carried driving person's information, prompting Carry the unlawful practice of vehicle carried driving person corresponding to driver information.

The device of the present embodiment, it can be used for the technical scheme for performing embodiment of the method shown in Fig. 1, its realization principle and skill Art effect is similar, and here is omitted.

It should be noted that herein, such as first and second or the like relational terms are used merely to a reality Body or operation make a distinction with another entity or operation, and not necessarily require or imply and deposited between these entities or operation In any this actual relation or order.Moreover, term " comprising ", "comprising" or its any other variant are intended to Nonexcludability includes, so that process, method, article or equipment including a series of elements not only will including those Element, but also the other element including being not expressly set out, or it is this process, method, article or equipment also to include Intrinsic key element.In the absence of more restrictions, the key element limited by sentence " including one ... ", it is not excluded that Other identical element in the process including the key element, method, article or equipment also be present.

Each embodiment in this specification is described by the way of related, identical similar portion between each embodiment Divide mutually referring to what each embodiment stressed is the difference with other embodiment.

For device embodiment, because it is substantially similar to embodiment of the method, so the comparison of description is simple Single, the relevent part can refer to the partial explaination of embodiments of method.

Expression or logic and/or step described otherwise above herein in flow charts, for example, being considered use In the order list for the executable instruction for realizing logic function, may be embodied in any computer-readable medium, for Instruction execution system, device or equipment (such as computer based system including the system of processor or other can be held from instruction The system of row system, device or equipment instruction fetch and execute instruction) use, or combine these instruction execution systems, device or set It is standby and use.For the purpose of this specification, " computer-readable medium " can any can be included, store, communicate, propagate or pass Defeated program is for instruction execution system, device or equipment or the dress used with reference to these instruction execution systems, device or equipment Put.The more specifically example (non-exhaustive list) of computer-readable medium includes following：Electricity with one or more wiring Connecting portion (electronic installation), portable computer diskette box (magnetic device), random access memory (RAM), read-only storage (ROM), erasable edit read-only storage (EPROM or flash memory), fiber device, and portable optic disk is read-only deposits Reservoir (CDROM).In addition, computer-readable medium, which can even is that, to print the paper of described program thereon or other are suitable Medium, because can then enter edlin, interpretation or if necessary with it for example by carrying out optical scanner to paper or other media His suitable method is handled electronically to obtain described program, is then stored in computer storage.

It should be appreciated that each several part of the present invention can be realized with hardware, software, firmware or combinations thereof.

In the above-described embodiment, multiple steps or method can use storage to be performed in memory and by suitable instruction The software or firmware that system performs are realized.If for example, being realized with hardware, with another embodiment, can use Any one of following technology well known in the art or their combination are realized：With for realizing logic work(to data-signal The discrete logic of the logic gates of energy, there is the application specific integrated circuit of suitable combinational logic gate circuit, programmable gate Array (PGA), field programmable gate array (FPGA) etc..

The embodiment of the present invention also provides a kind of electronic equipment, and the electronic equipment includes the dress described in foregoing any embodiment Put.

Fig. 3 is the structural representation of electronic equipment one embodiment of the present invention, it is possible to achieve is implemented shown in Fig. 1-2 of the present invention The flow of example, as shown in figure 3, above-mentioned electronic equipment can include：Housing 31, processor 32, memory 33, circuit board 34 and electricity Source circuit 35, wherein, circuit board 34 is placed in the interior volume that housing 31 surrounds, and processor 32 and memory 33 are arranged on circuit On plate 34；Power circuit 35, for each circuit or the device power supply for above-mentioned electronic equipment；Memory 33 is used to store and can hold Line program code；Processor 32 is run and executable program generation by reading the executable program code stored in memory 33 Program corresponding to code, the method for performing detection vehicle carried driving person's unlawful practice described in foregoing any embodiment.

Processor 32 to the specific implementation procedures of above-mentioned steps and processor 32 by run executable program code come The step of further performing, the description of Fig. 1-2 illustrated embodiments of the present invention is may refer to, will not be repeated here.

The electronic equipment exists in a variety of forms, includes but is not limited to：

(1) mobile communication equipment：The characteristics of this kind equipment is that possess mobile communication function, and to provide speech, data Communicate as main target.This Terminal Type includes：Smart mobile phone (such as iPhone), multimedia handset, feature mobile phone, and it is low Hold mobile phone etc..

(2) super mobile personal computer equipment：This kind equipment belongs to the category of personal computer, there is calculating and processing work( Can, typically also possess mobile Internet access characteristic.This Terminal Type includes：PDA, MID and UMPC equipment etc., such as iPad.

(3) portable entertainment device：This kind equipment can show and play content of multimedia.The kind equipment includes：Audio, Video player (such as iPod), handheld device, e-book, and intelligent toy and portable car-mounted navigation equipment.

(4) server：The equipment for providing the service of calculating, the composition of server are total including processor, hard disk, internal memory, system Line etc., server is similar with general computer architecture, but due to needing to provide highly reliable service, therefore in processing energy Power, stability, reliability, security, scalability, manageability etc. require higher.

(5) other electronic equipments with data interaction function.

Those skilled in the art are appreciated that to realize all or part of step that above-described embodiment method carries Suddenly it is that by program the hardware of correlation can be instructed to complete, described program can be stored in a kind of computer-readable storage medium In matter, the program upon execution, including one or a combination set of the step of embodiment of the method.

For convenience of description, it is to be divided into various units/modules with function to describe respectively to describe apparatus above.Certainly, exist The function of each unit/module can be realized in same or multiple softwares and/or hardware when implementing of the invention.

As seen through the above description of the embodiments, those skilled in the art can be understood that this

Invention can add the mode of required general hardware platform to realize by software.Based on such understanding, the present invention The part that is substantially contributed in other words to prior art of technical scheme can be embodied in the form of software product, should Computer software product can be stored in storage medium, such as ROM/RAM, magnetic disc, CD, including some instructions are causing One computer equipment (can be personal computer, server, or network equipment etc.) perform each embodiment of the present invention or Method described in some parts of person's embodiment.

The foregoing is only a specific embodiment of the invention, but protection scope of the present invention is not limited thereto, any Those familiar with the art the invention discloses technical scope in, the change or replacement that can readily occur in, all should It is included within the scope of the present invention.Therefore, protection scope of the present invention should be defined by scope of the claims.

Claims

A kind of 1. method for detecting vehicle carried driving person's unlawful practice, it is characterised in that including：

In the image of shooting, steering wheel is identified；

The unlawful practice region of vehicle carried driving person is determined according to the steering wheel identified；

Using the unlawful practice region of determination, the unlawful practice of vehicle carried driving person is detected.
2. the method for detection vehicle carried driving person's unlawful practice according to claim 1, it is characterised in that described in shooting In image, identify that steering wheel includes：

Rim detection is carried out to described image, obtains edge-detected image；

In the edge-detected image, steering wheel profile is detected using Probabilistic Hough Transform algorithm.
3. the method for detection vehicle carried driving person's unlawful practice according to claim 2, it is characterised in that described to the figure As carrying out rim detection, obtaining edge-detected image includes：

Described image is converted into gray scale image, the gray scale image is handled using Canny edge detection algorithms, obtained Edge-detected image.
4. the method for detection vehicle carried driving person's unlawful practice according to claim 2, it is characterised in that described to utilize probability Hough transform algorithm detects that steering wheel profile includes：

The circular arc in the edge-detected image is extracted, using the circular arc of Probabilistic Hough Transform algorithm Detection and Extraction, and is projected to In same Hough spaces, the votes in Hough spaces are counted, choose the most point of votes as the center of circle；

The center of circle is calculated to the distance of circular arc, obtains radius；

If the radius being calculated meets the steering wheel radius threshold pre-set, the center of circle is the steering wheel center of circle, described Radius is steering wheel radius.
5. the method for detection vehicle carried driving person's unlawful practice according to claim 1, it is characterised in that the unlawful practice Region be vehicle carried driving person's shoulder to the region of huckle, the steering wheel that the foundation identifies determines the violation of vehicle carried driving person Behavior region includes：

The steering wheel or the boundary rectangle of the steering wheel that will identify that translate up the first distance threshold pre-set, obtain Vehicle carried driving person's shoulder is to the region of huckle.
6. the method for detection vehicle carried driving person's unlawful practice according to claim 5, it is characterised in that the detection is vehicle-mounted The unlawful practice of driver detects for safety belt, and the safety belt detection includes：

According to the vehicle carried driving person's shoulder identified to the region of huckle, safety belt detection zone is determined；

It is determined that safety belt detection zone at, detection vehicle carried driving person whether fasten seat belts.
7. it is according to claim 6 detection vehicle carried driving person's unlawful practice method, it is characterised in that it is described it is determined that At safety belt detection zone, detect whether that safety belt be present includes：

It is determined that safety belt detection zone in, according to brightness of image, obtain adaptive Canny threshold values；

According to the Canny threshold values of acquisition, Canny rim detections are carried out to safety belt detection zone, obtain safety belt detection zone Interior edge image；

In edge image in obtained safety belt detection zone, using Probabilistic Hough Transform algorithm, the safety belt is detected The straightway in edge image in detection zone；

According to the straightway detected, the distance between straight line slope over 10 and each straightway are calculated, if two straight line slope over 10 It is approximately the same, and the distance between two straightway is less than the safety belt distance threshold pre-set, determines that vehicle carried driving person is Good safety belt；It is approximately the same if there is no two straight line slope over 10, and the distance between two straightway is not less than pre-setting Safety belt distance threshold, determine that vehicle carried driving person does not fasten seat belts.
8. the method for detection vehicle carried driving person's unlawful practice according to claim 1, it is characterised in that the unlawful practice Region is vehicle carried driving person's head zone；The steering wheel that the foundation identifies determines the unlawful practice region bag of vehicle carried driving person Include：

The steering wheel that will identify that obtains scaled down version steering wheel, by the scaled down version direction according to the scale smaller pre-set Disk translates up the second distance threshold value pre-set, obtains vehicle carried driving person's head zone；Or

The boundary rectangle for the steering wheel that will identify that obtains scaled down version boundary rectangle, by institute according to the scale smaller pre-set State scaled down version boundary rectangle and translate up the second distance threshold value pre-set, obtain vehicle carried driving person's head zone.
9. the method for detection vehicle carried driving person's unlawful practice according to claim 8, it is characterised in that the detection is vehicle-mounted The unlawful practice of driver for take phone detection, it is described take phone detection include：

It is determined that vehicle carried driving person's head zone, detect whether mobile phone be present.
10. it is according to claim 9 detection vehicle carried driving person's unlawful practice method, it is characterised in that it is described it is determined that Vehicle carried driving person's head zone, detect whether that mobile phone be present includes：

According to the hand textural characteristics storehouse pre-set, the hand textural characteristics included in vehicle carried driving person's head zone are extracted；

The disaggregated model that the hand textural characteristics input of extraction is pre-set, if the disaggregated model output result shows to deposit In hand-characteristic, confirm that vehicle carried driving person takes phone, if the disaggregated model output result shows hand-characteristic not be present, Confirm that vehicle carried driving person does not take phone.
11. the method for detection vehicle carried driving person's unlawful practice according to claim 8, it is characterised in that the detection car The unlawful practice for carrying driver detects for smoking, and the smoking detection includes：

According to the steering wheel and vehicle carried driving person's head zone, smoking detection zone is determined；

It is determined that smoking detection zone, detect whether smoking phenomenon be present.
12. it is according to claim 11 detection vehicle carried driving person's unlawful practice method, it is characterised in that it is described it is determined that Smoking detection zone, detect whether that smoking phenomenon be present includes：

Detect the imperfect profile included in the steering wheel profile；

First line in the imperfect profile starting point and the center of circle is connected, and, connect the imperfect profile stop and the center of circle The second line；

If the angle between the first line and the second line exceedes the angle threshold pre-set, the imperfect profile is extracted In include the hand textural characteristics of cigarette；

The smoking identification and classification model that the input of the hand textural characteristics comprising cigarette of extraction is pre-set, if the smoking Identification and classification model output result shows cigarette in hand-characteristic be present, confirms that vehicle carried driving person has smoking phenomenon, if institute State smoking identification and classification model output result and show cigarette in hand-characteristic not be present, confirm that vehicle carried driving person does not have smoking and showed As.
13. it is according to claim 11 detection vehicle carried driving person's unlawful practice method, it is characterised in that it is described it is determined that Smoking detection zone, detect whether that smoking phenomenon be present includes：

Extract the mouth region in vehicle carried driving person's head zone；

Texture feature extraction is carried out to mouth region, if including elongated rectangular area in the textural characteristics of extraction, confirmed Vehicle carried driving person's smoking, if not including elongated rectangular area in the textural characteristics of extraction, confirm that vehicle carried driving person does not inhale Cigarette.
14. the method for detection vehicle carried driving person's unlawful practice according to any one of claim 1 to 13, it is characterised in that Methods described also includes：

In the image of shooting, head image is identified, is matched with the head image storehouse pre-set, is obtained and identify Vehicle carried driving person's head image for matching of head image, according to vehicle carried driving person's head image in the head image storehouse with The mapping relations of vehicle carried driving person's information, obtain vehicle carried driving person's information corresponding to the head image identified；

According to obtained vehicle carried driving person's information, prompt, and/or, record vehicle carried driving corresponding to vehicle carried driving person's information The unlawful practice of member.
A kind of 15. device for detecting vehicle carried driving person's unlawful practice, it is characterised in that including：Steering wheel identification module, in violation of rules and regulations row For area determination module and unlawful practice detection module, wherein,

Steering wheel identification module, in the image of shooting, identifying steering wheel；

Unlawful practice area determination module, for determining the unlawful practice region of vehicle carried driving person according to the steering wheel identified；

Unlawful practice detection module, for using the unlawful practice region determined, detecting the unlawful practice of vehicle carried driving person.
16. the device of detection vehicle carried driving person's unlawful practice according to claim 15, it is characterised in that the steering wheel Identification module includes：Edge detection unit, contour detecting unit and steering wheel acquiring unit, wherein,

Edge detection unit, for carrying out rim detection to described image, obtain edge-detected image；

Contour detecting unit, in the edge-detected image, direction plate wheel to be detected using Probabilistic Hough Transform algorithm It is wide；

Steering wheel acquiring unit, for obtaining the steering wheel of steering wheel profile described in inscribe.
17. the device of detection vehicle carried driving person's unlawful practice according to claim 16, it is characterised in that the edge inspection Surveying unit includes：Conversion subunit and rim detection subelement, wherein,

Conversion subunit, for described image to be converted into gray scale image；

Rim detection subelement, for the gray scale image to be handled using Canny edge detection algorithms, obtain edge inspection Altimetric image.
18. the device of detection vehicle carried driving person's unlawful practice according to claim 16, it is characterised in that the profile inspection Surveying unit includes：Center of circle determination subelement, radius determination subelement and outline specifying unit, wherein,

Center of circle determination subelement, for extracting the circular arc in the edge-detected image, detected using Probabilistic Hough Transform algorithm The circular arc of extraction, and project in same Hough spaces, the votes in statistics Hough spaces, choose the most point of votes As the center of circle；

Radius determination subelement, for calculating the center of circle to the distance of circular arc, obtain radius；

Outline specifying unit, if the radius being calculated meets the steering wheel radius threshold pre-set, the center of circle is side To the disk center of circle, the radius is steering wheel radius.
19. the device of detection vehicle carried driving person's unlawful practice according to claim 15, it is characterised in that the violation row Include for area determination module：

First shift unit, steering wheel or the boundary rectangle of the steering wheel for will identify that translate up what is pre-set First distance threshold, vehicle carried driving person's shoulder is obtained to the region of huckle.
20. the device of detection vehicle carried driving person's unlawful practice according to claim 19, it is characterised in that the violation row Include for detection module：Safety belt detection zone determining unit and safety belt detection unit, wherein,

Safety belt detection zone determining unit, for identifying vehicle carried driving person's shoulder to the region of huckle, according to what is identified Vehicle carried driving person's shoulder determines safety belt detection zone to the region of huckle；

Safety belt detection unit, for it is determined that safety belt detection zone at, detection vehicle carried driving person whether fasten seat belts.
21. the device of detection vehicle carried driving person's unlawful practice according to claim 20, it is characterised in that the safety belt Detection unit includes：It is single that threshold value obtains subelement, rim detection subelement, line segment detection subelement and safety belt detection Member, wherein,

Threshold value obtain subelement, for it is determined that safety belt detection zone in, according to brightness of image, obtain adaptive Canny threshold values；

Rim detection subelement, for the Canny threshold values according to acquisition, Canny rim detections are carried out to safety belt detection zone, Obtain the edge image in safety belt detection zone；

Line segment detection subelement, in the edge image in obtained safety belt detection zone, being become using probability Hough Scaling method, detect the straightway in the edge image in the safety belt detection zone；

Safety belt detection sub-unit, for according to the straightway detected, calculating between straight line slope over 10 and each straightway Distance, if two straight line slope over 10 are approximately the same, and the distance between two straightway is less than the safety belt distance pre-set Threshold value, determine that vehicle carried driving person fastens seat belts；It is approximately the same if there is no two straight line slope over 10, and between two straightway Distance be not less than the safety belt distance threshold that pre-sets, determine that vehicle carried driving person does not fasten seat belts.
22. the device of detection vehicle carried driving person's unlawful practice according to claim 15, it is characterised in that the violation row Include for area determination module：

Second shift unit, for the steering wheel that will identify that according to the scale smaller pre-set, scaled down version steering wheel is obtained, The scaled down version steering wheel is translated up to the second distance threshold value pre-set, obtains vehicle carried driving person's head zone； Or the boundary rectangle of the steering wheel for will identify that obtains scaled down version boundary rectangle according to the scale smaller pre-set, The scaled down version boundary rectangle is translated up to the second distance threshold value pre-set, obtains the vehicle carried driving person header area Domain.
23. the device of detection vehicle carried driving person's unlawful practice according to claim 22, it is characterised in that the violation row Include for detection module：Head zone recognition unit and mobile phone detection unit, wherein,

Head zone recognition unit, for identifying steering wheel, vehicle carried driving person's head zone is determined according to the steering wheel identified；

Mobile phone detection unit, for it is determined that vehicle carried driving person's head zone, detect whether mobile phone be present.
24. the device of detection vehicle carried driving person's unlawful practice according to claim 23, it is characterised in that the mobile electricity Words detection unit includes：Hand texture feature extraction subelement and mobile phone detection sub-unit, wherein,

Hand texture feature extraction subelement, for according to the hand textural characteristics storehouse pre-set, extracting vehicle carried driving person's head The hand textural characteristics included in portion region；

Mobile phone detection sub-unit, for the disaggregated model for pre-setting the hand textural characteristics input of extraction, if institute State disaggregated model output result and show hand-characteristic be present, confirm that vehicle carried driving person takes phone, if the disaggregated model is defeated Go out result and show hand-characteristic not be present, confirm that vehicle carried driving person does not take phone.
25. the device of detection vehicle carried driving person's unlawful practice according to claim 22, it is characterised in that the violation row Include for detection module：Head zone recognition unit, smoking detection zone determination subelement and smoking detection unit, wherein,

Head zone recognition unit, for identifying steering wheel, vehicle carried driving person's head zone is determined according to the steering wheel identified；

Smoking detection zone determination subelement, for according to the steering wheel and vehicle carried driving person's head zone, it is determined that Smoking detection zone；

Smoking detection unit, for it is determined that smoking detection zone, detect whether smoking phenomenon be present.
26. the device of detection vehicle carried driving person's unlawful practice according to claim 25, it is characterised in that the smoking inspection Surveying unit includes：Discontinuous detection sub-unit, connection subelement, extraction subelement and smoking detection sub-unit, wherein,

Discontinuous detection sub-unit, for detecting the imperfect profile included in the steering wheel profile；

Subelement is connected, for connecting the first line of the imperfect profile starting point and the center of circle, and, connection is described imperfect Profile stop and second line in the center of circle；

Subelement is extracted, if the angle between the first line and the second line exceedes the angle threshold pre-set, extracts institute State the hand textural characteristics that cigarette is included in imperfect profile；

Smoking detection sub-unit, the smoking for the input of the hand textural characteristics comprising cigarette of extraction to be pre-set, which differentiates, to divide Class model, if the smoking identification and classification model output result shows cigarette in hand-characteristic be present, confirm vehicle carried driving person Smoking phenomenon be present, if the smoking identification and classification model output result shows cigarette in hand-characteristic not be present, confirm car Carry driver and smoking phenomenon not be present.
27. the device of detection vehicle carried driving person's unlawful practice according to claim 25, it is characterised in that the smoking inspection Surveying unit includes：Mouth determination subelement and smoking determination subelement, wherein,

Mouth determination subelement, for extracting the mouth region in vehicle carried driving person's head zone；

Smoking determination subelement, for carrying out texture feature extraction to mouth region, if included in the textural characteristics of extraction Elongated rectangular area, confirm vehicle carried driving person's smoking, if not including elongated rectangular area in the textural characteristics of extraction, Confirm the non-smoking of vehicle carried driving person.
28. the device of detection vehicle carried driving person's unlawful practice according to any one of claim 15 to 27, it is characterised in that Described device also includes：Vehicle carried driving person's data obtaining module and unlawful practice processing module, wherein,

Vehicle carried driving person's data obtaining module, in the image of shooting, identifying head image, with the head pre-set Image library is matched, vehicle carried driving person's head image that the head image for obtaining and identifying matches, according to the head Vehicle carried driving person's head image and the mapping relations of vehicle carried driving person's information in image library, obtain the head image identified Corresponding vehicle carried driving person's information；

Unlawful practice processing module, and/or, the vehicle carried driving is recorded for according to obtained vehicle carried driving person's information, prompting The unlawful practice of vehicle carried driving person corresponding to member's information.
29. a kind of electronic equipment, it is characterised in that the electronic equipment includes：Housing, processor, memory, circuit board and electricity Source circuit, wherein, circuit board is placed in the interior volume that housing surrounds, and processor and memory are set on circuit boards；Power supply Circuit, for each circuit or the device power supply for above-mentioned electronic equipment；Memory is used to store executable program code；Processing The executable program code that device is stored by reading in memory runs program corresponding with executable program code, for holding The method of detection vehicle carried driving person's unlawful practice described in the foregoing any claim 1-14 of row.