KR101416457B1 - Road crime prevention system using recognition of opposite direction drive and pedestrian - Google Patents

Abstract

The present invention relates to a road security system to which recognition of dangerous situations of vehicles driving on the wrong side of the road and pedestrians is applied. The road security system related to an embodiment of the present invention may comprise a camera which photographs at least one first vehicle continuously; a control unit which determines whether the first vehicle is driving on the wrong side of the road by using at least one among moving direction information of the first vehicle which is photographed through the camera, the vehicle plate information, and position change information and extracts information on the vehicle number of the first vehicle if the control unit determines that the first vehicle is driving on the wrong side of the road; and a wired communication unit which transmits the vehicle number information to a preset external organization in accordance with the control of the control unit.

Description

BACKGROUND OF THE INVENTION Field of the Invention [0001] The present invention relates to a road crime prevention system,

BACKGROUND OF THE INVENTION 1. Field of the Invention [0001] The present invention relates to a road crime prevention system to which a reverse car and pedestrian danger situation recognition is applied. More specifically, the present invention relates to a road security system for continuously photographing a vehicle, judging whether the vehicle reverses by using at least one of moving direction information, license plate information and position change information of the vehicle included in the photographed image, ≪ / RTI >

Crime Prevention refers to various activities or actions that maintain the order of society in general and prevent the occurrence of crime.

Crime prevention can be divided into crime prevention activities by police, which are national air lines, and self-help crime prevention activities in the private sector.

The crime prevention activities of the police include the patrol (patrol, bike patrol, motorcycle patrol, 112 car patrol), diagnosis and improvement of the crime susceptible environment, guidance for the youth, management and control of the sex workers, firearms, And the like.

Private crime prevention activities include patrol and leading activities by autonomous crime prevention organizations in the community, and crime prevention activities by private security companies in charge of service expenses.

On the other hand, the cameras for controlling violations of the regulations of the vehicles are frequently installed on the road.

Typically, the camera can perform tail wagging, stop line violation, illegal U-turn control, and reverse run.

A method of performing a cracking process on a tail, a stop line violation, illegal U-turn, reverse run, etc. using a camera frequently parked on the road can be used.

Especially, regarding the method of controlling the reverse run, although the loop method is widely used, there is a disadvantage that the installed facility is not suitable for the old and advanced software method.

That is, in order to execute the interception of the reverse run using the software method, the vehicle is continuously photographed, and the vehicle is reversed by using at least one of the moving direction information of the vehicle, the license plate information and the position change information included in the photographed image We need an algorithm that makes it easy to judge whether the

Korean Intellectual Property Office Registration No. 10-1136021

SUMMARY OF THE INVENTION It is an object of the present invention to provide a road crime prevention system and a control method thereof,

More specifically, the present invention relates to a road security system for continuously photographing a vehicle, judging whether the vehicle reverses by using at least one of moving direction information, license plate information and position change information of the vehicle included in the photographed image, Method to the user.

It is to be understood that both the foregoing general description and the following detailed description of the present invention are exemplary and explanatory and are not intended to limit the invention to the precise form disclosed. It can be understood.

A road crime prevention system according to an embodiment of the present invention for realizing the above-mentioned object includes a camera for continuously photographing at least one first vehicle, a moving direction of the first vehicle included in the first image photographed through the camera A control unit for determining whether the first vehicle reverses by using at least one of information, license plate information, and position change information, and extracting vehicle number information of the first vehicle when it is determined that the first vehicle reverses; And a wired / wireless communication unit for transmitting the vehicle number information to a predetermined external organization under the control of the control unit.

The control unit may determine a moving direction of the first vehicle by comparing a plurality of consecutive images included in the first image, extract a road on which the first vehicle is located from the first image, And the movement direction information may be information obtained by comparing a moving direction of the first vehicle with a predetermined traveling direction of the extracted road.

The control unit may detect a number plate region of the first vehicle, compare a plurality of consecutive images contained in the first image to determine a change in position and size of the plate plate region, Determines a moving direction of the first vehicle using a change in position and size, extracts a road on which the first vehicle is located from the first image, and determines a predetermined traveling direction of the extracted road, And may be information obtained by comparing the moving direction of the first vehicle and the predetermined traveling direction of the extracted road determined using the position and size change of the license plate area.

The control unit estimates a feature vector for the first image and detects the first vehicle using an adaboost algorithm and an SVM algorithm to which the estimated feature vector is applied, Applying a mean shift algorithm to the first vehicle to track the first vehicle and to determine a direction of travel of the first vehicle using a tracking path for the first vehicle, Wherein the control unit extracts a road on which the first vehicle is located to determine a predetermined traveling direction of the extracted road, and the position change information includes a traveling direction of the first vehicle determined using the tracking route, May be information that is compared with the information.

In addition, the SVM algorithm uses a feature vector of the pre-learned vehicle and the estimated feature vector together, and the controller applies the adaboost algorithm to classify the information related to the first vehicle first , And the SVM algorithm may be applied to the classified information to detect the first vehicle.

Further, the car number information includes type information of the first vehicle, usage information, and identification information, the type information is represented by two numbers, the usage information is represented by one Hangul, It can be represented by four numbers.

The wired / wireless communication unit and the external organization communicate using at least one of a short-distance wireless communication and a long-distance wireless communication. The short-range wireless communication may be WiFi, Bluetooth, Radio Frequency Identification (RFID) (IrDA), Ultra Wideband (UWB), and ZigBee technology, and the long distance wireless communication uses at least one of a code division multiple access (CDMA), a frequency division multiple access (FDMA), a time division multiple access division multiple access (OFDMA), and single carrier frequency division multiple access (SC-FDMA) techniques.

Meanwhile, a road crushing method according to another embodiment of the present invention for realizing the above-mentioned problem is characterized by including a first step of continuously photographing at least one first vehicle, a first step of photographing at least one first vehicle, A second step of determining whether the first vehicle reverses by using at least one of movement direction information, license plate information, and position change information of the first vehicle, A third step of extracting vehicle number information of the vehicle, and a fourth step of transmitting the vehicle number information to a predetermined external organization.

The second step may include determining a moving direction of the first vehicle by comparing a plurality of consecutive images contained in the first image, extracting a road on which the first vehicle is located from the first image, And determining a predetermined traveling direction of the extracted road. The traveling direction information may be information obtained by comparing a moving direction of the first vehicle with a predetermined traveling direction of the extracted road.

The second step may include detecting a license plate area of the first vehicle, comparing a plurality of consecutive images included in the first image to determine a change in position and size of the license plate area, Determining a moving direction of the first vehicle by using a change in position and size of the license plate area, extracting a road on which the first vehicle is located in the first image, determining a predetermined traveling direction of the extracted road, And the license plate information may be information obtained by comparing the moving direction of the first vehicle and the predetermined traveling direction of the extracted road determined using the position and size change of the license plate area.

The second step may include estimating a feature vector for the first image, detecting the first vehicle using an adaboost algorithm and an SVM algorithm to which the estimated feature vector is applied, Applying a mean shift algorithm to the detected first vehicle to track the first vehicle, determining a direction of travel of the first vehicle using a tracking path for the first vehicle, The method of claim 1, further comprising: extracting a road on which the first vehicle is located from the first image and determining a predetermined traveling direction of the extracted road; And may be information obtained by comparing the moving direction and a predetermined traveling direction of the extracted road.

Further, the car number information includes type information of the first vehicle, usage information, and identification information, the type information is represented by two numbers, the usage information is represented by one Hangul, It can be represented by four numbers.

The fourth step may include at least one of a short-range wireless communication and a long-distance wireless communication. The short-range wireless communication may be WiFi, Bluetooth, RFID, infrared Wherein the long distance wireless communication is at least one of code division multiple access (CDMA), frequency division multiple access (FDMA), time division multiple access (TDMA) access, orthogonal frequency division multiple access (OFDMA), and single carrier frequency division multiple access (SC-FDMA) techniques.

The present invention can provide a user with a road crime prevention system and a control method thereof with efficient backward movement discrimination.

More specifically, the present invention relates to a road security system for continuously photographing a vehicle, judging whether the vehicle reverses by using at least one of moving direction information, license plate information and position change information of the vehicle included in the photographed image, Method can be provided to the user.

It should be understood, however, that the effects obtained by the present invention are not limited to the above-mentioned effects, and other effects not mentioned may be clearly understood by those skilled in the art to which the present invention belongs It will be possible.

BRIEF DESCRIPTION OF THE DRAWINGS The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate a preferred embodiment of the invention and, together with the description, serve to provide a further understanding of the technical idea of the invention, It should not be construed as limited.
Fig. 1 shows an example of a block diagram of a road security system according to the present invention.
Fig. 2 is a flowchart for explaining controlling the reverse run through the road security system according to the present invention.
FIG. 3 is a flowchart for explaining a specific process of using the moving direction information of the vehicle in connection with the inverse-motion intermittent process according to FIG.
FIG. 4 is a flow chart for explaining a concrete procedure of using the license plate information of the vehicle in connection with the reverse run control process according to FIG.
FIG. 5 is a flowchart for explaining a specific process using position change information of a vehicle with respect to the process of regulating the backward running according to FIG.
6A and 6B illustrate a specific example of extracting a specific object from an acquired image in connection with the present invention.
FIG. 7 illustrates another specific example of extracting a specific object from the acquired image, according to the present invention.
8A and 8B are views for explaining a concrete example of a license plate transferred to an external organ and the meaning of the included information, in the context of the present invention.

A camera frequently installed on the street to crack down on the violation of the regulations of the vehicle can carry out the reverse run enforcement. At this time, an algorithm is required to easily determine the vehicle inversely related to the vehicle and acquire the related vehicle information.

Accordingly, the present invention proposes a road crime prevention system for continuously photographing a vehicle and judging whether the vehicle reverses by using at least one of moving direction information, license plate information and position change information of the vehicle included in the photographed image .

Before describing the concrete operation of the present invention, the components of the road security system applied to the present invention will be described in detail.

The road security system 1100 includes a wired / wireless communication unit 1110, a camera 1120, a user input unit 1130, a signal lamp 1140, an output unit 1150, a memory 1160, a control unit 1180, and a power supply unit 1190, And the like.

However, the components shown in FIG. 1 are not essential, so that a road security system 1100 having more or fewer components may be implemented.

Hereinafter, the components will be described in order.

The wired and wireless communication unit 1110 may include one or more modules for enabling wireless communication between the road security system 1100 and the wireless communication system or between the road security system 1100 and the network in which the road security system 1100 is located .

For example, the wired / wireless communication unit 1110 may include a mobile communication module 1112, a wireless Internet module 1113, a short distance communication module 1114, a location information module 1115, and the like.

The mobile communication module 1112 transmits and receives radio signals to at least one of a base station, an external road security system 1100, and a server on a mobile communication network. The wireless Internet module 1113 refers to a module for wireless Internet access, and may be built in or enclosed in the road security system 1100.

WLAN (Wi-Fi), Wibro (Wireless broadband), Wimax (World Interoperability for Microwave Access), HSDPA (High Speed Downlink Packet Access) and the like can be used as the technology of the wireless Internet.

Next, the wireless Internet module 1113 refers to a module for wireless Internet access, and may be built in or enclosed in the road security system 1100.

WLAN (Wi-Fi), Wibro (Wireless broadband), Wimax (World Interoperability for Microwave Access), HSDPA (High Speed Downlink Packet Access) and the like can be used as the technology of the wireless Internet.

The short-range communication module 1114 refers to a module for short-range communication. Bluetooth, Radio Frequency Identification (RFID), Infrared Data Association (IrDA), Ultra Wideband (UWB), ZigBee, and the like can be used as the short range communication technology.

The position information module 1115 is a module for obtaining the position of the road security system 1100, and a representative example thereof is a Global Position System (GPS) module.

According to the current technology, the GPS module 1115 calculates distance information and accurate time information from three or more satellites, and then applies a trigonometric method to the calculated information to generate a three-dimensional string of latitude, longitude, The location information can be accurately calculated.

At present, a method of calculating position and time information using three satellites and correcting an error of the calculated position and time information using another satellite is widely used. In addition, the GPS module 1115 can calculate speed information by continuously calculating the current position in real time.

In addition, the wired / wireless communication unit 1110 may support wired communication.

Wired communication is a telecommunication that is made by using a wire in a transmission line.

Although there is a drawback in that wired communication is expensive to install, there is an advantage that a stable and high-quality transmission path is guaranteed as compared with wireless communication, and branching can be performed on the way.

Conventionally, all of domestic communications relies on a wired transmission line, but now, there are more and more cases in which the local communication is dependent on the wired and the out-of-area communication is wireless.

The transmission lines to be applied here include spiral, city and suburban cable, transport cable, coaxial cable, submarine cable (combination of coaxial cable and relay amplifier), and fiber optic cable.

Wired communication is collectively referred to as a telecommunication in which both transmission and reception are connected to an electric line, and thereby a signal is mediated.

A typical example is a telegraph communication. Unlike a wireless communication in which a plurality of transmissions are performed in principle for one transmission, the principle of communication in a transmission rate of 1: 1 is a wired communication method. In some cases, wireless communication is included in a part of a transmission line in accordance with the development of communication technology. Therefore, when a broadcasting station, an exchange, etc. and a terminal receiver (subscriber) are combined into a wired transmission line, it is proper to view them as wired communication.

In 1876, a practical telephone was invented by the AG Bell, and the telephone exchange service for general subscribers was put into practical use within a relatively short range. A triode tube was invented and inserted into the telephone trunk line, Has been made possible to make a long distance call.

Initially, in order to reduce the attenuation of the voice signal in the telephone trunk line, an induction coil having an induction coil was used at a certain distance (about 2 km). However, after long time, kHz carrier to a voice signal, multiplexed and transmitted.

Coaxial cable and microwave trunk line are widely used for signal multiplexing, but coaxial cable is gradually being replaced by optical fiber cable.

Referring to Fig. 1, a camera 1121 processes an image frame such as a still image or moving image obtained by an image sensor in a video communication mode or a photographing mode.

The processed image frame can be displayed on the display portion 1151. [

The image frame processed by the camera 1121 can be stored in the memory 1160 or transmitted to the outside via the wired / wireless communication unit 1110. [

At this time, two or more cameras 1121 may be provided depending on the use environment.

Typically, the camera 1121 is capable of photographing objects that can be photographed, that is, images of a pedestrian and a vehicle in an area where the camera 1121 is installed.

In addition, the control unit 180 can recognize the vehicle number using the plurality of vehicle images obtained through the camera 1121. [

At this time, the recognized plurality of car number information may be stored in the memory 1160 or transmitted to the outside through the wired / wireless communication unit 1110. [

The user input unit 1130 generates input data for controlling the operation of the road security system 1100 by the user.

The user input unit 1130 may receive from the user a signal designating two or more contents among the displayed contents according to the present invention.

A signal for designating two or more contents may be received via the touch input, or may be received via the hard key and soft key input.

The user input unit 1130 may receive an input from the user to select the one or more contents. In addition, an input for generating an icon related to a function that the road security system 1100 can perform can be received from the user.

The user input unit 1130 may include a direction key, a key pad, a dome switch, a touch pad (static / static), a jog wheel, a jog switch, and the like.

In addition, the traffic light 1140 is a device for expressing traffic conditions in color for safe traffic order.

There are three colors of red, green, and orange displayed by traffic light.

Usually, red is stopped, green is on, orange is on attention, but it is also used to mean orange on road traffic.

Originally, it is red, blue, and yellow, but the light of a light bulb has an orange color, and when it passes through the color glass, it becomes yellow and orange and blue becomes green. The driving regulations of the railway have prescribed red, green and orange.

Light intensity The light bulb uses the same wattage (W) but the light transmittance differs according to the wavelength of light.

The brightness of a traffic light is different depending on the speed of the vehicle. However, it is usually required to have a brightness that can be recognized at a distance of about 600 meters on a railroad, (On a weekly basis).

The output unit 1150 is for generating output related to visual, auditory or tactile sense and includes a display unit 1151, an acoustic output module 1152, an alarm unit 1153, a haptic module 1154, 1155, and the like.

The display unit 1151 displays (outputs) the information processed by the road security systems 1100 and 100.

For example, a UI (User Interface) or a GUI (Graphic User Interface) associated with the interception of the road security system 1100 is displayed.

In addition, the display unit 1151 according to the present invention supports 2D and 3D display modes.

The display unit 1151 drives the switch liquid crystal 1151b and the optical parallax barrier 50 and the display device 1151a under the control of the control unit 1180 in the 3D display mode, 1151a to perform a 3D display operation.

The display unit 151 may be a liquid crystal display (LCD), a thin film transistor-liquid crystal display (TFT LCD), an organic light-emitting diode (OLED) A flexible display, and a three-dimensional display (3D display).

Some of these displays may be transparent or light transmissive so that they can be seen through. This can be referred to as a transparent display, and a typical example of the transparent display is TOLED (Transparent OLED) and the like. The rear structure of the display portion 1151 may also be of a light transmission type. With this structure, the user can see the object located behind the body of the road security system 1100 through the area occupied by the display portion 1151 of the body of the road security system 1100.

The audio output module 1152 can output audio data received from the wire / wireless communication unit 1110 or stored in the memory 1160 in a recording mode, a voice recognition mode, a broadcast receiving mode, and the like.

The sound output module 1152 also outputs an acoustic signal related to the function performed by the road security system 1100. The sound output module 1152 may include a receiver, a speaker, a buzzer, and the like.

The alarm unit 1153 outputs a signal for notifying the occurrence of an event of the road security system 1100.

The alarm unit 1153 may output a signal for informing occurrence of an event in a form other than a video signal or an audio signal, for example, a vibration. In this case, the display unit 1151 and the audio output module 152 may be a type of the alarm unit 1153. The display unit 1151 and the audio output module 152 may be connected to the display unit 1151 or the audio output module 1152, .

The haptic module 154 generates various tactile effects that the user can feel.

A typical example of the haptic effect generated by the haptic module 154 is vibration. The intensity and pattern of the vibration generated by the hit module 154 can be controlled.

For example, different vibrations may be synthesized and output or sequentially output.

In addition to vibration, the haptic module 1154 may be configured to perform various functions such as a pin arrangement vertically moving with respect to the contact skin surface, a spraying force or suction force of the air through the injection port or the suction port, a touch on the skin surface, contact with an electrode, And various tactile effects such as an effect of reproducing a cold sensation using an endothermic or exothermic element can be generated.

The haptic module 1154 can be implemented not only to transmit a tactile effect through direct contact, but also to allow a user to feel a tactile effect through a muscular sensation such as a finger or an arm.

The projector module 1155 is a component for performing an image project function using the road security system 1100 and is displayed on the display unit 1151 according to a control signal of the control unit 1180 An image that is the same as or at least partially different from the image can be displayed on an external screen or a wall.

Specifically, the projector module 1155 includes a light source (not shown) that generates light (for example, laser light) for outputting an image to the outside, a light source And a lens (not shown) for enlarging and outputting the image at a predetermined focal distance to the outside. Further, the projector module 1155 may include a device (not shown) capable of mechanically moving the lens or the entire module to adjust the image projection direction.

The projector module 1155 can be divided into a CRT (Cathode Ray Tube) module, an LCD (Liquid Crystal Display) module and a DLP (Digital Light Processing) module according to the type of the display means. In particular, the DLP module may be advantageous in downsizing the projector module 1151 by enlarging and projecting an image generated by reflecting light generated from a light source on a DMD (Digital Micromirror Device) chip.

Preferably, the projector module 1155 may be provided on the side, front or rear surface of the road security system 1100 in the longitudinal direction. It goes without saying that the projector module 1155 may be provided at any position of the road security system 1100 as needed.

The memory 1160 may store a program for processing and controlling the controller 1180, and may perform functions for temporarily storing input / output data. The frequency of use of each of the data may also be stored in the memory 1160. In addition, the memory unit 1160 may store data related to vibration and sound of various patterns that are output upon touch input on the touch screen.

The memory 1160 may be a flash memory type, a hard disk type, a multimedia card micro type, a card type memory (for example, SD or XD memory, etc.) ), A random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read- A magnetic disk, an optical disk, a memory, a magnetic disk, or an optical disk. The road security system 1100 may operate in association with a web storage that performs a storage function of the memory 1160 on the Internet.

The controller 1180 typically controls the overall operation of the road security system 1100.

The control unit 1180 may include a multimedia module 1181 for multimedia playback. The multimedia module 1181 may be implemented in the control unit 1180 or may be implemented separately from the control unit 1180.

The power supply unit 1190 receives external power and internal power under the control of the controller 1180 and supplies power required for operation of the respective components.

The various embodiments described herein may be implemented in a recording medium readable by a computer or similar device using, for example, software, hardware, or a combination thereof.

According to a hardware implementation, the embodiments described herein may be implemented as application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays May be implemented using at least one of a processor, controllers, micro-controllers, microprocessors, and other electronic units for performing other functions. In some cases, The embodiments described may be implemented by the controller 1180 itself.

According to a software implementation, embodiments such as the procedures and functions described herein may be implemented in separate software modules. Each of the software modules may perform one or more of the functions and operations described herein. Software code can be implemented in a software application written in a suitable programming language. The software code is stored in the memory 1160 and can be executed by the control unit 1180. [

A concrete interruption operation based on the configuration of the above-described road crime prevention system 1100 will be described with reference to Fig.

Fig. 2 is a flowchart for explaining controlling the reverse run through the road security system according to the present invention.

Referring to FIG. 2, step (S10) of the camera 1120 continuously photographing at least one first vehicle is proceeded.

At this time, the control unit 1180 can determine whether the first vehicle reverses by using at least one of moving direction information, license plate information, and position change information of the first vehicle included in the first image photographed at step S10 (S20).

That is, the control unit 180 determines whether or not the moving direction of the first vehicle determined by using the movement direction information, the position and the size change of the license plate area, which is information obtained by comparing the moving direction of the first vehicle with the predetermined traveling direction of the extracted road, Which is a comparison between a moving direction of the first vehicle determined by using the tracking path and a predetermined traveling direction of the extracted road, based on at least one of the license plate information, It is possible to judge whether or not it is reversed.

A detailed description thereof will be given later on in Examples 1 to 3 below.

Thereafter, the control unit 1800 can extract the vehicle number information of the first vehicle if it is determined that the first vehicle is running in reverse (S30).

The vehicle number information may include the type information of the first vehicle, the usage information, and the identification information.

Finally, the control unit 1180 transmits the car number information to the predetermined external organization through the wired / wireless communication unit 1110 (S40).

At this time, the predetermined external organization may include an organization that gives penalties and fines in accordance with the control of the vehicle, and an organization that sends information on the regulation to the borrower.

Also, the wired / wireless communication unit 1110 and the external organization can communicate using at least one of short-distance wireless communication and long-distance wireless communication.

The short-range wireless communication may use at least one of WiFi, Bluetooth, Radio Frequency Identification (RFID), IrDA, Ultra Wideband (UWB), and ZigBee technology. have.

Further, the long-distance wireless communication may be performed by using a code division multiple access (CDMA), a frequency division multiple access (FDMA), a time division multiple access (TDMA), an orthogonal frequency division multiple access (OFDMA), a single carrier frequency division multiple access (SC- Technology may be used.

Hereinafter, a specific example will be described by classifying the parameter according to the type of parameter used to determine the inverse operation.

1st Example

According to one embodiment of the present invention, the control unit 180 can determine whether the vehicle is reversed through the movement direction information by comparing the moving direction of the first vehicle with the predetermined traveling direction of the extracted road.

FIG. 3 is a flowchart for explaining a specific process of using the moving direction information of the vehicle in connection with the inverse-motion intermittent process according to FIG.

Referring to FIG. 3, step S20 described in FIG. 2 is embodied according to a method of using moving direction information.

First, the control unit 1180 can determine the moving direction of the first vehicle by comparing a plurality of consecutive images included in the first image (S11a).

Then, the controller 1180 extracts the road on which the first vehicle is located from the first image (S11b), and determines a predetermined traveling direction of the extracted road (S11c).

As a result, the moving direction information used by the control unit 1180 is information obtained by comparing the moving direction of the first vehicle with the predetermined traveling direction of the extracted road.

Second Example

According to another embodiment of the present invention, the controller 180 compares the moving direction of the first vehicle determined by using the position and size change of the license plate area with the predetermined traveling direction of the extracted road, Can be determined.

FIG. 4 is a flow chart for explaining a concrete procedure of using the license plate information of the vehicle in connection with the reverse run control process according to FIG.

Referring to FIG. 4, step S20 described in FIG. 2 is embodied according to a method using license plate information.

More specifically, the step of detecting the license plate area of the first vehicle proceeds (S12a).

In addition, the control unit 1180 may compare a plurality of consecutive images included in the first image to determine a change in position and size of the license plate area (S12b).

In addition, the controller 1180 can determine the moving direction of the first vehicle using the determined position and size of the license plate area (S12c).

On the other hand, the control unit 1180 can extract the road on which the first vehicle is located from the first image (S12d), and determine the predetermined traveling direction of the extracted road (S12e).

As a result, the license plate information used by the controller 1180 is information obtained by comparing the moving direction of the first vehicle determined using the position and size change of the license plate area with the predetermined traveling direction of the extracted road.

Third Example

According to another embodiment of the present invention, the controller 180 can determine whether the vehicle is reversed by using the position change information which is obtained by comparing the moving direction of the first vehicle determined using the tracking path with the predetermined traveling direction of the extracted road .

FIG. 5 is a flowchart for explaining a specific process using position change information of a vehicle with respect to the process of regulating the backward running according to FIG.

Referring to FIG. 5, step S20 described in FIG. 2 is embodied according to a method of using position change information.

First, the control unit 1180 can estimate a feature vector for the first image (S13a).

In addition, the control unit 1180 can detect the first vehicle using the adaboost algorithm and the SVM algorithm to which the estimated feature vector is applied (S13b).

In addition, a step S13c for tracking the first vehicle is performed by applying a mean shift algorithm to the detected first vehicle.

When tracking the first vehicle, the control unit 1180 can determine the direction of movement of the first vehicle using the tracking path for the first vehicle (S13d).

In addition, the control unit 1180 extracts the road on which the first vehicle is located from the first image (S13e), and determines a predetermined traveling direction of the extracted road (S13f).

Hereinafter, a method of extracting and tracking specific objects applied in steps S13a to S13c will be described in detail with reference to FIGS. 6A to 7. FIG.

6A and 6B illustrate a specific example of extracting a specific object from an acquired image in connection with the present invention.

FIG. 7 illustrates another specific example of extracting a specific object from the acquired image, according to the present invention.

6A and 6B, it is assumed that the object object vehicle is extracted, but the contents of the present invention are not limited thereto.

That is, it is apparent that the method described below can also be applied to a method of extracting a pedestrian by the present invention other than a target object vehicle.

6A and 6B are graphs illustrating a method of applying a mean shift method to a histogram region determined as a specific vehicle using the Batacharya distance of a histogram according to an exemplary embodiment of the present invention, 6A is the n-1th sample screen, and FIG. 6B is the n-th sample screen.

The learning data extraction process extracts feature vectors from the positive data of the object to be recognized (brightness difference between the surroundings and boundary value distribution, etc.) Process.

In order to detect a vehicle as a specific object, a block or a macroblock of a predetermined size is moved on the entire image and compared with a feature vector of a previously learned vehicle block. If the feature vector of the previously learned vehicle block and the feature vector of the intra-image block are similar, it is detected as a vehicle.

The adaboost learning classification method, which is a widely used recognition and classification algorithm, is widely used in the field of detecting an object having a block form such as a pedestrian, a vehicle, and a face.

In this method, a weak classifier that classifies the learning samples well is selected, and a strong classifier is created by combining the weak classifiers. And we use this strong classifier in several steps in cascade.

That is, a method of sequentially comparing similar patterns with each other according to a defined pattern can be applied.

To do this, we first perform adaboosting on feature vectors, and then search for n feature vectors corresponding to weak classifiers. Then, for all positive and negative samples, a strong classifier is calculated using the following equation (1).

Here, α means the weight of h (x). A value of S (x) including a positive sample at an arbitrary ratio of K% is defined as a detection threshold value Th _D and an intermediate value of the minimum value of S (x) values of the negative sample and Th _D is defined as a sample removal threshold value Th _E ). For each step, a negative sample whose value is smaller than Th _E is determined as a positive negative and removed from the learning data. Negative samples do not exist or repeat the above steps as many times as the maximum number of steps predefined by the user. In this process, the lower the Th _D , the higher the recognition rate, but the more steps are needed. In this way, by removing the negative sample reliably at each step, the next step will give more weight to the uncertain sample, thus improving the final recognition rate.

At the time of object detection, adaboost detection is performed using h (x) and α of each step obtained in the learning process for the input image. At this time, the final discrimination equation to be used in the proposed method uses [Equation 2] in which the threshold value in the existing AdaBoost method is changed to Th _D.

Through the above discrimination, discrimination of the next step is performed on regions detected positively having a value of 1. This process is the same as the conventional cascade detection method. The AdaBoost is fast because it is based on pixel values. However, there is a case in which a vehicle is mistakenly recognized as a vehicle.

In order to compensate for this, we use SVM algorithm. HOG (Histogram of the Oriented Gradients) was used as the feature vector at this time.

In other words, performing adaboost detection as a classifier and performing SVM and HoG at the same time means repeating steps by performing two steps.

The HOG divides a block of a predetermined size into M small macroblocks (typically 3X3 pixels) Bm (m = 1, ..., M). The orientation of each pixel in each Bm is histogrammed to determine the most prominent direction vector Vm. Each Vm is gathered to construct an M-dimensional vector. The generated HOG is one of the most used feature vectors and shows excellent detection performance in combination with the SVM algorithm.

Extracting the HOG of each candidate vehicle through the Adaboost algorithm results in an M-dimensional space. Among the vehicle candidates distributed in the high dimension (N dimension), it is almost impossible to find an optimal linear segment to distinguish the actual vehicle, and most of them have a nonlinear segment. That is, in general, the linear discriminant condition is to find a solution to be obtained according to a given arbitrary value, and the nonlinear discriminant condition is a solution to find a solution by an arbitrary set value.

However, data can be projected onto a higher M-dimensional space by combining the N-dimensional HOG features to create a new dimension. It is easier to find the optimal linear segment to distinguish the actual vehicle among the projected vehicle candidates in the M-dimension. The SVM algorithm is to project the data to a higher dimension and find the optimal linear separation line.

However, to obtain HOG, a large amount of computation is required. Therefore, HOG is not obtained for the whole image, and SVM is performed only for those judged to be vehicle candidates through the results of the adabo boost. In this paper, we propose a vehicle recognition method with S140 step structure, which improves accuracy and computational complexity.

In addition, a method of tracking the extracted specific object may be applied together, and this will be described in detail.

That is, the tracking method for improving the object detection rate will be described in detail.

In addition, object recognition and tracking techniques are generally being studied in different areas.

The focus of this paper is on the combination of object tracking techniques as a complementary tool for finding missing vehicles in extraction.

it may happen that the vehicle detected on the (n-1) th screen is not detected on the n-th screen. However, it can be assumed that the vehicle with one frame interval is not large in brightness or color change. Therefore, by applying the object tracking technique, the vehicle detected on the (n-1) th screen can be continuously tracked without missing on the n th screen. In this embodiment, among the various object tracking methods, a section in the histogram where errors are converged using the Bhattacharyya distance of the histogram is determined as a characteristic region of the corresponding vehicle, and a mean shift method is applied to the region Applied.

FIGS. 6A and 6B are sample photographs for explaining a technique of applying a mean shift method to a histogram region judged as a vehicle using a Batacharya distance of a histogram according to an embodiment of the present invention, , FIG. 6A shows the n-1th sample screen, and FIG. 6B shows the n-th sample screen.

As shown in FIG. 6A and FIG. 6B, in order to detect a vehicle missed in the n-th screen, at a position P _{n -1} (x _{n -1} , y _{n -1} ) where the vehicle is detected in the previous frame, Begin applying. In the n-th screen _{P n -1 (x n -1,} y n -1) the same location as the _{P n (x n -1, y} n -1) obtained in the color histogram in the block with a predetermined size in advance. n-1-th screen histogram obtained in the area of the detected person in the H ₁ and a histogram H ₂ obtained from the same location of the n-th screen compared according to the following [Equation 3] Bata charya constant ρ (P _n (x _{n - 1} , y _{n -1} )).

The larger the value of the Batagarija constant, the more similar the two histograms are. The average shift algorithm is a technique for finding the position P _n (x _n , y _n ) of a block maximizing the Batachari coefficient in the n-th screen. (5), which obtains a new position P _n (x, y) of the n-th screen which maximizes the Batakari constants, by differentiating the expression (3) with respect to the position.

B _n (x _i , y _i ) is the x and y coordinate values in the image of the i-th pixel in the block of the predetermined size in the n-th screen. H ₂ is newly obtained at the new position obtained by the equation (5), and P (x, y) is obtained again through the equation (5). Repeat this process to obtain the final position P _n (x _n , y _n ) on the _nth screen. P _n (x _n , y _n ) is the position on the n th screen of the vehicle that was not detected in the above technique for detecting an object.

In addition, Fig. 7 shows another concrete example of extracting a specific object from the acquired image, in connection with the present invention.

FIG. 7A is a diagram showing a specific object result in the 1000th frame, FIG. 7B is a diagram showing a specific object result in the 1200th frame, FIG. FIG. 7D is a view showing a specific object result in the 1900th frame. FIG.

7 (a) to 7 (d) are diagrams for explaining a simple example to which the present invention is applied, and the contents of the present invention are not limited thereto.

Meanwhile, FIG. 8 is a detailed description of the car number information transmitted to the external organization through the step S40 described in FIG.

Referring to Fig. 8A, an example of a license plate is shown.

Referring to FIG. 8A, a license plate of " 43 is 651X " is disclosed.

Here "43", that is, the two numbers at the beginning indicate the type of vehicle.

Concretely, 01 ~ 69 means passenger cars, 70 ~ 79 means van, 80 ~ 97 means cargo, and 98 ~ 99 means special car.

Next, "A" means that a single Hangul means the use of the vehicle.

Specifically, "ah, ba, sa, ja" is a Hangul that is given to a business vehicle such as a taxi or a bus. Among them, "ah, ba, 31, 32, 33, and 34 "in front of" ah, bar, four, and so on " Here, "31, 32" means a private taxi, and "33, 34" means a taxi in a company.

The next four digits are the identification numbers for distinguishing the vehicles from each other.

Referring to FIG. 8B, license plates classified according to classification, color, and application preference are classified, and representative license plate information corresponding thereto is shown on the right side.

When the configuration of the present invention described above is applied, the vehicle is continuously photographed, and it is determined whether the vehicle reverses by using at least one of moving direction information, license plate information, and position change information of the vehicle included in the photographed image A road security system can be provided to the user.

The present invention can also be embodied as computer-readable codes on a computer-readable recording medium. A computer-readable recording medium includes all kinds of recording apparatuses in which data that can be read by a computer system is stored. Examples of the computer-readable recording medium include a ROM, a RAM, a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like, and may be implemented in the form of a carrier wave (for example, transmission over the Internet) .

In addition, the computer-readable recording medium may be distributed over network-connected computer systems so that computer readable codes can be stored and executed in a distributed manner. In addition, functional programs, codes, and code segments for implementing the present invention can be easily inferred by programmers of the technical field to which the present invention belongs.

It should be understood that the above-described apparatus and method are not limited to the configurations and methods of the embodiments described above, but the embodiments may be modified such that all or some of the embodiments are selectively combined .

Claims (13)

A camera for continuously photographing at least one first vehicle;
Determining whether the first vehicle reverses by using at least one of movement direction information, license plate information, and position change information of the first vehicle included in the first image photographed through the camera, A controller for extracting vehicle number information of the first vehicle when it is determined that the vehicle is running in reverse; And
And a wired / wireless communication unit for transmitting the vehicle number information to a predetermined external organization under the control of the controller,
Wherein,
Estimating a feature vector for the first image,
The first vehicle is detected using an adaboost algorithm and an SVM algorithm to which the estimated feature vector is applied,
Applying a mean shift algorithm to the detected first vehicle to track the first vehicle,
Determine a moving direction of the first vehicle using a tracking path for the first vehicle,
A road on which the first vehicle is located is extracted from the first image to determine a predetermined traveling direction of the extracted road,
Wherein the position change information is information obtained by comparing a moving direction of the first vehicle determined using the tracking path and a predetermined traveling direction of the extracted road. The method according to claim 1,
Wherein,
Determining a moving direction of the first vehicle by comparing a plurality of consecutive images included in the first image,
A road on which the first vehicle is located is extracted from the first image to determine a predetermined traveling direction of the extracted road,
Wherein the moving direction information is information obtained by comparing a moving direction of the first vehicle with a predetermined traveling direction of the extracted road. The method according to claim 1,
Wherein,
Detecting a license plate area of the first vehicle,
Comparing a plurality of consecutive images included in the first image to determine a change in position and size of the license plate area,
Determining a moving direction of the first vehicle by using a change in position and size of the license plate area,
A road on which the first vehicle is located is extracted from the first image to determine a predetermined traveling direction of the extracted road,
Wherein the license plate information is information obtained by comparing a moving direction of the first vehicle and a predetermined traveling direction of the extracted road determined using a change in position and size of the license plate area. delete 3. The method of claim 2,
The SVM algorithm uses the feature vector of the pre-learned vehicle and the estimated feature vector together,
Wherein,
Applying the adaboost algorithm to classify information associated with the first vehicle first,
And the first vehicle is detected by applying the SVM algorithm to the classified information. The method according to claim 1,
Wherein the vehicle number information includes type information, usage information, and identification information of the first vehicle,
Wherein the type information is represented by two numbers, the usage information is represented by one Hangul, and the identification information is represented by four numbers. The method according to claim 1,
The wired / wireless communication unit and the external organization communicate using at least one of wired communication, short-range wireless communication, and long-distance wireless communication,
The short-range wireless communication uses at least one of WiFi, Bluetooth, Radio Frequency Identification (RFID), infrared data association (IrDA), Ultra Wideband (UWB) ,
The long-range wireless communication may be performed using a wireless communication technique such as code division multiple access (CDMA), frequency division multiple access (FDMA), time division multiple access (TDMA), orthogonal frequency division multiple access (OFDMA), single carrier frequency division multiple access Wherein the at least one of the plurality of roads is used. A first step in which the camera continuously photographs at least one first vehicle;
A second step of determining whether the first vehicle reverses by using at least one of movement direction information, license plate information, and position change information of the first vehicle included in the first image photographed in the first step;
A third step of extracting vehicle number information of the first vehicle when it is determined that the first vehicle reverses; And
And a fourth step of transmitting the car number information to a predetermined external organization,
The second step comprises:
Estimating a feature vector for the first image;
Detecting the first vehicle using an adaboost algorithm and an SVM algorithm to which the estimated feature vector is applied;
Tracking the first vehicle by applying a mean shift algorithm to the detected first vehicle;
Determining a moving direction of the first vehicle using a tracking path for the first vehicle;
Extracting a road on which the first vehicle is located from the first image; And
And determining a predetermined traveling direction of the extracted road,
Wherein the position change information is information obtained by comparing a moving direction of the first vehicle determined using the tracking path and a predetermined traveling direction of the extracted road. 9. The method of claim 8,
The second step comprises:
Comparing a plurality of consecutive images included in the first image to determine a moving direction of the first vehicle;
Extracting a road on which the first vehicle is located from the first image; And
And determining a predetermined traveling direction of the extracted road,
Wherein the moving direction information is information obtained by comparing a moving direction of the first vehicle with a predetermined traveling direction of the extracted road. 9. The method of claim 8,
The second step comprises:
Detecting a license plate area of the first vehicle;
Comparing a plurality of consecutive images included in the first image to determine a change in position and size of the license plate area;
Determining a moving direction of the first vehicle by using a change in position and size of the license plate area;
Extracting a road on which the first vehicle is located from the first image; And
And determining a predetermined traveling direction of the extracted road,
Wherein the license plate information is information obtained by comparing a moving direction of the first vehicle and a predetermined traveling direction of the extracted road determined using a change in position and size of the license plate area. delete 9. The method of claim 8,
Wherein the vehicle number information includes type information, usage information, and identification information of the first vehicle,
Wherein the type information is represented by two numbers, the usage information is represented by one Hangul, and the identification information is represented by four numbers. 9. The method of claim 8,
The fourth step may include communicating using at least one of a short-range wireless communication and a long-range wireless communication,
The short-range wireless communication uses at least one of WiFi, Bluetooth, Radio Frequency Identification (RFID), infrared data association (IrDA), Ultra Wideband (UWB) ,
The long-range wireless communication may be performed using a wireless communication technique such as code division multiple access (CDMA), frequency division multiple access (FDMA), time division multiple access (TDMA), orthogonal frequency division multiple access (OFDMA), single carrier frequency division multiple access Wherein the at least one of the roads is used.

Images