KR20130015976A - Apparatus and method for detecting a vehicle - Google Patents

Abstract

PURPOSE: A method and an apparatus for detecting a vehicle first apply simple characteristics which show the general characteristic of the vehicle to learning, and then apply characteristics of high accuracy expressing detailed information of the vehicle from detected candidate vehicles to the learning. CONSTITUTION: By the usage of the first algorithm, vehicle area candidates showing vehicle form are detected from an image in which photographed vehicle detection apparatus(S120). For the detection of the vehicle area candidates, at least one of rectangle formed haar-wavelet characteristic is learned through the first algorithm. By the usage of the second algorithm, the apparatus detects a final vehicle area from the vehicle area candidates(S130). In the detection of the final vehicle area, at least one of simple characteristic from the detected vehicle area candidates is learned through the second algorithm. [Reference numerals] (AA) Start; (BB) End; (S110) Taking a video picture; (S120) Detecting vehicle area candidates through a first algorithm; (S130) Detecting the final vehicle area through a second algorithm; (S140) Outputting the final vehicle area

Description

Method and apparatus for detecting a vehicle {APPARATUS AND METHOD FOR DETECTING A VEHICLE}

The present disclosure relates to a vehicle detection method and apparatus, and more particularly, to a method and apparatus for detecting a vehicle using a camera provided in a vehicle.

The conventional vehicle detection technology is largely divided into two methods of heuristically determining and applying a feature of a vehicle or learning and applying a classifier using previously acquired training images. Here, the learning-based vehicle detection method shows an excellent performance, but a lot of error occurs when the detection time is applied to apply in real time.

An object of the present specification is to provide a high speed vehicle detection method and apparatus that can be applied in real time.

The present disclosure provides a method for detecting a vehicle, the method comprising: photographing an image;

Detecting vehicle region candidates indicating a vehicle shape from the captured image; And detecting a final vehicle region from the vehicle region candidates, wherein the detecting of the vehicle region candidates comprises: learning a at least one rectangular haar-wavelet feature through a first algorithm; The detecting of the final vehicle region may include learning at least one predetermined feature from the detected vehicle region candidates through a second algorithm.

In addition, the detecting of the vehicle area candidates and the final vehicle area is characterized by using a coarse-to-fine method.

The stepwise method also includes grouping the at least one ha-wavelet feature or the at least one predetermined feature according to a predetermined criterion; And applying the first algorithm or the second algorithm sequentially to the grouped features.

In addition, the first algorithm is characterized in that the ADABOOST algorithm.

In addition, the second algorithm may be an ADABOOST algorithm or an SVM algorithm.

In addition, the predetermined feature is characterized in that the texture or contour information of the vehicle representing the detailed characteristics of the vehicle.

In addition, the predetermined feature is characterized in that the Local Binary Pattern (LBP), LAB or Histogram of Oriented Gradients (HOG).

In addition, the present specification may further include outputting the detected final vehicle area.

In addition, the learning through the first algorithm or the second algorithm is characterized in that it is performed using previously stored learning data.

The present disclosure may further include updating the stored learning data with data being learned by the first algorithm or the second algorithm when the vehicle region candidates or the final vehicle region are not detected; And storing the updated learning data.

In addition, the present specification provides a vehicle detection apparatus, comprising: a camera unit for photographing an image; And a controller operatively connected to the camera unit, wherein the controller is configured to indicate a vehicle shape through learning using at least one rectangular haar-wavelet feature from the photographed image using a first algorithm. And controlling to detect vehicle region candidates, and detecting at least one predetermined feature from the detected vehicle region candidates by detecting a final vehicle region through learning using a second algorithm.

The control unit may control to detect the vehicle area candidates and the final vehicle area by using a coarse-to-fine method.

In addition, the control unit groups the at least one ha-wavelet feature or the at least one predetermined feature according to a predetermined criterion, and sequentially the first algorithm or the first to the grouped features. It is characterized by controlling to detect the vehicle region candidates and the final vehicle region by applying a two algorithm.

In addition, the present specification further comprises an output unit, wherein the control unit is characterized in that for controlling the output unit to output the detected final vehicle area.

The controller may control to perform learning through the first algorithm or the second algorithm using previously stored learning data.

The present disclosure may further include a memory, and the controller may be configured to convert the previously stored learning data into data being learned by the first algorithm or the second algorithm when the vehicle region candidates or the final vehicle region are not detected. And control the memory to store the updated training data.

In the present specification, a simple feature representing a general characteristic of a vehicle is first applied to learning, and then, by learning and applying highly accurate features capable of expressing detailed information of the vehicle from the detected vehicle candidates, the vehicle can be detected quickly and accurately. There is.

1 (a) is a block diagram showing the configuration of a vehicle detection apparatus according to an embodiment of the present invention, Figure 1 (b) is a conceptual diagram of a vehicle equipped with the vehicle detection device.
2 (a) and (b) is a view showing a vehicle detection method according to an embodiment of the present specification, Figure 2 (a) is a flow chart showing a vehicle detection method, Figure 2 (b) is a vehicle detection result The figure shown.
3 (a) to 3 (c) are diagrams showing features used for vehicle detection to which an embodiment of the present disclosure can be applied, and FIG. 3 (a) shows haar-wavelet features, and FIG. b) shows the LAB characteristic, and FIG. 3C shows the HOG characteristic.
4 is a diagram illustrating a vehicle detection method using a stepwise method according to an embodiment of the present disclosure, and FIG. 4 (a) illustrates vehicle area candidates and / or final vehicle area using a coarse-to-fine method. 4 (b) is a diagram illustrating a vehicle detection result through a stepwise method.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings, and the same or corresponding components will be denoted by the same reference numerals regardless of the reference numerals and redundant description thereof will be omitted.

The suffixes "module", "unit" and "unit" for components used in the following description are given or mixed in consideration of ease of specification, and do not have distinct meanings or roles by themselves. .

Hereinafter, a configuration of an apparatus for detecting an object located in front of a vehicle (hereinafter referred to as a “detecting device”) according to an exemplary embodiment of the present invention will be described with reference to FIG. 1.

Here, the vehicle detecting apparatus of FIG. 1 may be configured not only as stand alone, but also as a mobile terminal, a telematics terminal, a smart phone, and a portable detection device. (Portable Terminal), Personal Digital Assistant (PDA), Portable Multimedia Player (PMP), Notebook Computer, Tablet PC, Tablet PC, Wibro Detection Device, IPTV (Internet Protocol Television) The present invention can be applied to various detection devices such as a detection device, a television, a 3D television, an imaging device, a telematics detection device, a navigation detection device, and an audio video navigation (AVN) detection device.

1 (a) is a block diagram showing the configuration of a vehicle detection apparatus according to an embodiment of the present invention, Figure 1 (b) is a conceptual diagram of a vehicle equipped with the vehicle detection device.

As shown in FIG. 1A, the vehicle detection apparatus 100 includes a camera unit 110, a control unit 120, an output unit 130, and a memory 140. Not all components of the vehicle detecting apparatus 100 illustrated in FIG. 1A are essential components, and the detection apparatus 100 may be implemented by more components than those illustrated in FIG. 1A. The detection apparatus 100 may be implemented by fewer components.

First, the camera unit 110 includes at least one camera.

The camera processes image frames such as still images or moving images obtained by an image sensor in a video call mode or a photographing mode. The processed image frame may be displayed on the display unit. Here, the image sensor may be used in the same sense as the camera or in a broader sense than the camera.

The image frame processed by the camera may be stored in a memory or transmitted to the outside through a wireless communication unit. Two or more cameras may be provided depending on the usage environment.

The camera unit 110 may include at least one pair of cameras (eg, a stereo camera or a stereo course) spaced apart at horizontal intervals on the same central axis of any same plane of the vehicle detecting apparatus 100. A stereoscopic camera or a single camera. In this case, the fixed horizontal interval may be set in consideration of a distance between two eyes of a general person, and is set when configuring the vehicle detection apparatus 100. In addition, the camera unit 110 may be any camera unit capable of image capturing.

Also, the camera unit 110 may include a first image (eg, a left image captured by a left camera included in the pair of cameras) and a second image captured by the pair of cameras simultaneously. For example, the right image captured by the right camera included in the pair of cameras) is received.

In addition, the camera unit 110 may be an image sensor such as a CCD device, a CMOS device, or the like.

In addition, when installed in the vehicle, the camera unit 110 may be fixed to a predetermined position of the vehicle, and may be configured to photograph the front, side, rear, and the like in the driving direction, and receive the photographed image.

The controller 120 controls the overall operation of the detection device 100.

That is, the controller 120 controls to detect vehicle region candidates indicating a vehicle shape through learning using a first algorithm on at least one rectangular haar-wavelet feature from the captured image. At least one predetermined feature from the detected vehicle region candidates is controlled to detect the final vehicle region through learning using the second algorithm.

In addition, the controller 120 controls to detect the vehicle area candidates and the final vehicle area by using a coarse-to-fine method.

In addition, the control unit 120 groups the at least one ha-wavelet feature or the at least one predetermined feature according to a predetermined criterion, and sequentially the first algorithm on the grouped features. Or apply the second algorithm to detect the vehicle region candidates and the final vehicle region.

In addition, the controller 120 controls the output unit to output the detected final vehicle area.

In addition, the controller 120 controls to perform learning through the first algorithm or the second algorithm using previously stored learning data.

In addition, when the vehicle region candidates or the final vehicle region are not detected, the controller 120 updates the pre-stored learning data with data being learned by the first algorithm or the second algorithm. The memory is controlled to store training data.

The output unit 130 is for generating an output related to visual, auditory or tactile senses, and may include a display unit, a sound output module, an alarm unit, and a haptic module. In addition, the output unit 130 outputs the vehicle area finally detected from the captured image.

The display unit displays (outputs) information processed by the vehicle detecting apparatus 100. For example, when the vehicle detection apparatus is in a call mode, the UI (User Interface) or Graphic User Interface (GUI) related to the call is displayed. When the vehicle detecting apparatus 100 is in the video call mode or the photographing mode, the photographed and / or received image, the UI, and the GUI are displayed.

The display unit includes a liquid crystal display (LCD), a thin film transistor-liquid crystal display (TFT LCD), an organic light-emitting diode (OLED), a flexible display, 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 unit may also be configured as a light transmissive structure. With this structure, the user can see the object located behind the detection device body through the area occupied by the display of the detection device body.

Two or more display units may exist according to the implementation form of the vehicle detection apparatus 100. For example, a plurality of display units may be spaced apart or integrally disposed on one surface of the vehicle detection apparatus 100, or may be disposed on different surfaces.

When the display unit and a sensor for detecting a touch operation (hereinafter, referred to as a touch sensor) form a mutual layer structure (hereinafter referred to as a touch screen), the display unit 151 is an input device other than an output device. May also be used. The touch sensor may have the form of, for example, a touch film, a touch sheet, a touch pad, or the like.

The touch sensor may be configured to convert a change in pressure applied to a specific portion of the display unit or capacitance generated at a specific portion of the display unit into an electrical input signal. The touch sensor can be configured to detect not only the position and area to be touched but also the pressure at the time of touch.

If there is a touch input to the touch sensor, the corresponding signal (s) is sent to the touch controller. The touch controller processes the signal (s) and then transmits the corresponding data to the controller 120. As a result, the controller 120 may determine which area of the display unit is touched.

The proximity sensor may be disposed in an inner region of the detection device surrounded by the touch screen or near the touch screen. The proximity sensor refers to a sensor that detects the presence or absence of an object approaching a predetermined detection surface or a nearby object without mechanical contact using the force of an electromagnetic field or infrared rays. The proximity sensor has a longer life span than the contact sensor and its utilization is also high.

Examples of the proximity sensor include a transmission photoelectric sensor, a direct reflection photoelectric sensor, a mirror reflection photoelectric sensor, a high frequency oscillation proximity sensor, a capacitive proximity sensor, a magnetic proximity sensor, and an infrared proximity sensor. And to detect the proximity of the pointer by the change of the electric field along the proximity of the pointer when the touch screen is electrostatic. In this case, the touch screen (touch sensor) may be classified as a proximity sensor.

Hereinafter, for convenience of explanation, the act of allowing the pointer to be recognized without being in contact with the touch screen so that the pointer is located on the touch screen is referred to as a "proximity touch", and the touch The act of actually touching the pointer on the screen is called "contact touch." The position where the pointer is proximately touched on the touch screen means a position where the pointer is vertically corresponding to the touch screen when the pointer is touched.

The proximity sensor detects a proximity touch and a proximity touch pattern (e.g., a proximity touch distance, a proximity touch direction, a proximity touch speed, a proximity touch time, a proximity touch position, a proximity touch movement state, and the like). Information corresponding to the detected proximity touch operation and the proximity touch pattern may be output on the touch screen.

The sound output module may output audio data received from a wireless communication unit or stored in a memory in a call signal reception, a call mode or a recording mode, a voice recognition mode, a broadcast reception mode, and the like. The sound output module may output a sound signal related to a function (eg, a call signal reception sound, a message reception sound, etc.) performed by the detection device 100. Such an audio output module may include a receiver, a speaker, a buzzer, and the like.

The alarm unit outputs a signal for notifying occurrence of an event of the vehicle detection apparatus 100. Examples of events generated by the detection apparatus include call signal reception, message reception, key signal input, and touch input. The alarm unit may output a signal for notifying occurrence of an event by vibration, in addition to a video signal or an audio signal. The video signal or the audio signal may also be output through the display unit or the audio output module, so that they may be classified as part of the alarm unit.

The haptic module generates various tactile effects that a user can feel. Vibration is a representative example of the haptic effect generated by the haptic module. The intensity and pattern of vibration generated by the haptic module can be controlled. For example, different vibrations may be synthesized and output or sequentially output.

In addition to vibration, the haptic module can be operated by a pin array that vertically moves with respect to the contact skin surface, the effect of stimulation such as the blowing force or suction force of air through the injection or inlet, grazing to the skin surface, contact of the electrode, and electrostatic force. In addition, various tactile effects can be generated, such as an effect of reproducing a feeling of cold and heat using an endothermic or heat generating element. The haptic module can deliver the haptic effect through direct contact, and can also be implemented to allow the user to feel the haptic effect through muscle sensations such as fingers or arms. Two or more haptic modules may be provided according to a configuration aspect of the portable vehicle detecting apparatus 100.

The memory 140 may store a program for the operation of the controller 120 and temporarily store input / output data (for example, a phone book, a message, a still image, a video, etc.). The memory 140 may store data regarding vibration and sound of various patterns output when a touch input on the touch screen is performed.

The memory 140 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), RAM (Random Access Memory, RAM), Static Random Access Memory (SRAM), Read-Only Memory (ROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Programmable Read-Only Memory (PROM), Magnetic Memory, Magnetic It may include a storage medium of at least one type of disk, optical disk. The detection apparatus 100 may operate in association with a web storage that performs a storage function of the memory 140 on the Internet.

In addition, in addition to the above-mentioned components, in order to apply an embodiment disclosed herein, the vehicle detecting apparatus 100 may further include the following components.

That is, the vehicle detecting apparatus 100 may include a communication unit, a microphone, a sensing unit, a user input unit, an interface unit, a power supply unit, and the like.

First, the vehicle detection apparatus 100 may further include a communication unit that performs a communication function with an arbitrary detection apparatus or a server under the control of the controller 120. In this case, the communication unit may include a wired / wireless communication module. Here, the wireless Internet technology, Wireless LAN (WLAN), Wi-Fi (Wi-Fi), Wireless Broadband (Wibro), WiMAX (World Interoperability for Microwave Access (Wimax), High-speed downlink packet access (High) Speed Downlink Packet Access (HSDPA), IEEE 802.16, Long Term Evolution (LTE), and Wireless Mobile Broadband Service (WMBS), and the like, and may include the short range communication. The technology may include Bluetooth, Radio Frequency Identification (RFID), Infrared Data Association (IrDA), Ultra Wideband (UWB), ZigBee, and the like. The wired communication technology may include universal serial bus (USB) communication.

The communication unit may include CAN communication, vehicle Ethernet, flexray, LIN (Local Interconnect Network), and the like, for communication with any vehicle provided with the detection apparatus 100.

The microphone receives an external sound signal by a microphone in a call mode, a recording mode, a voice recognition mode, etc., and processes it into electrical voice data. The processed voice data may be converted into a form transmittable to the mobile communication base station through the mobile communication module and output in the call mode. Various noise reduction algorithms may be implemented in the microphone to remove noise generated while receiving an external sound signal.

The sensing unit detects the current state of the vehicle detecting apparatus 100 such as the opening / closing state of the vehicle detecting apparatus 100, the position of the vehicle detecting apparatus 100, the presence or absence of a user contact, the orientation of the detecting apparatus, the acceleration / deceleration of the detecting apparatus, and the like. A sensing signal for controlling the operation of the vehicle detecting apparatus 100 is generated.

For example, when the vehicle detecting apparatus 100 is in the form of a slide phone, it may sense whether the slide phone is opened or closed. In addition, it may be sensed whether the power supply of the power supply unit, whether the external unit of the interface unit is coupled. The sensing unit may include a proximity sensor.

The user input unit generates input data for the user to control the operation of the detection device. The user input unit may include a key pad dome switch, a touch pad (static pressure / capacitance), a jog wheel, a jog switch, and the like.

The interface unit serves as a passage with all external devices connected to the vehicle detecting apparatus 100. The interface unit receives data from an external device, receives power, transfers the power to each component inside the vehicle detecting apparatus 100, or transmits the data inside the vehicle detecting apparatus 100 to an external device. For example, wired / wireless headset ports, external charger ports, wired / wireless data ports, memory card ports, ports for connecting devices with identification modules, audio input / output (I / O) ports, The video input / output (I / O) port and the earphone port may be included in the interface unit.

The identification module is a chip that stores various types of information for authenticating the use authority of the vehicle detecting apparatus 100. The identification module includes a user identification module (UIM) and a subscriber identification module.

(Subscriber Identify Module, SIM), Universal Subscriber Identity Module (USIM) and the like. Devices with identification modules (hereinafter referred to as "identification devices") can be manufactured in a smart card format. Therefore, the identification device may be connected to the vehicle detection apparatus 100 through a port.

When the vehicle detecting apparatus 100 is connected to an external cradle, the interface unit may be a passage through which power from the cradle is supplied to the vehicle detecting apparatus 100, or various command signals input from the cradle may be input by a user. It may be a passage that is delivered to the mobile detection device. Various command signals or power input from the cradle may be operated as signals for recognizing that the mobile detection apparatus is correctly mounted on the cradle.

The power supply unit receives an external power source and an internal power source under the control of the controller 120 to supply power for operation of each component.

The various embodiments described herein may be embodied 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 processors, controllers, micro-controllers, microprocessors, and electrical units for performing other functions.

In some cases, the embodiments described herein may be implemented by the controller 120 itself.

According to the software implementation, embodiments such as the procedures and functions described herein may be implemented as separate software modules. Each of the software modules may perform one or more of the functions and operations described herein. Software code may be implemented in software applications written in a suitable programming language. The software code may be stored in a memory and executed by the controller 120.

Hereinafter, a vehicle detection method proposed in the present specification will be described with reference to FIGS. 2 to 4.

2 (a) and (b) is a view showing a vehicle detection method according to an embodiment of the present specification, Figure 2 (a) is a flow chart showing a vehicle detection method, Figure 2 (b) is a vehicle detection result The figure shown.

Referring to FIG. 2 (a), the vehicle detecting apparatus 100 captures an image through a camera unit.

Here, the vehicle detecting apparatus captures an image of the front of the vehicle including the vehicle detecting apparatus in real time through the camera unit (S110).

Thereafter, the vehicle detecting apparatus detects vehicle region candidates indicating a vehicle shape from the captured image (S120). Here, the vehicle detecting apparatus learns and applies at least one rectangular wave-wavelet feature using a first algorithm to detect vehicle region candidates.

Here, the rectangular wave-shaped feature represents a component related to the vehicle, and there may be rectangular wave-shaped features of various sizes. In addition, the first algorithm may be an ADABOOST algorithm.

Here, learning and applying using the ADABOOST algorithm means detecting vehicle region candidates in the captured image through an AdaBoost learning classifier.

The Adaboost learning classifier combines weak classifiers with fast weighting into a weighted sum to form a strong classifier with good classification ability. This algorithm detects target objects.

That is, the vehicle detection apparatus detects vehicle region candidates by applying rectangular wavelet features to the Adaboost algorithm learning in the captured image. That is, the vehicle detecting apparatus filters the non-vehicle region from the captured image through the above process. Here, the vehicle detection apparatus may filter about 90% of the non-vehicle region through the above process.

Thereafter, the vehicle detecting apparatus detects a final vehicle region by learning and applying at least one predetermined feature from the detected vehicle region candidates through a second algorithm (S130).

Here, the predetermined feature represents a texture or contour of the vehicle, and may be Local Binary Pattern (LBP), LAB, or Histogram of Oriented Gradients (HOG).

The local binary pattern (LBP) is a method of expressing a texture of a local image, that is, dividing the image into regions having an arbitrary size and position with respect to each input image, and then, based on the median value of the divided image region, By binarizing a value, it is a method of expressing a texture that is robust to changes in lighting.

Also, the second algorithm may be an ADABOOST algorithm or an SVM algorithm.

Here, a detailed description of the target target learning method and the target target classification method using the SVM algorithm is described in the literature [Christopher JC Burges, "A totorial on Support Vector Machines for Pattern Recognition", Data mining and Knowledge Discovery 2 , 121-167, 1998.

Here, the vehicle detecting apparatus may detect the vehicle region candidates and / or the final vehicle region through a coarse-to-fine method. Hereinafter, a vehicle detection method using a coarse-to-fine method will be described in detail with reference to FIG.

Here, the vehicle detection apparatus may additionally perform the step of detecting vehicle candidates in addition to the step of learning and applying through the first algorithm and the second algorithm to detect the vehicle more accurately from the captured image.

That is, the vehicle detecting apparatus may detect the vehicle more accurately through an algorithm for learning application of steps 3 to N using additional features related to the vehicle. Here, the additional steps may be added in the step of detecting vehicle area candidates or may be added in the step of detecting the final vehicle candidate.

Thereafter, the vehicle detecting apparatus outputs the detected final vehicle region through an output unit (S140).

In addition, the vehicle detection apparatus may perform the learning application through the first algorithm or the second algorithm by using the learning data previously stored in the memory.

In addition, when the vehicle region candidates or the final vehicle region are not detected, the vehicle detecting apparatus may update the learning data previously stored in the memory with data being learned by the first algorithm or the second algorithm.

Therefore, the vehicle detecting apparatus performs vehicle detection through the updated learning data and stores the updated learning data in a memory.

As shown in FIG. 2 (b), a result of detecting a vehicle through a rectangular wave-shaped feature and predetermined features (eg, LAB, HOG, LBP) is shown (200).

3 (a) to 3 (c) are diagrams showing features used for vehicle detection to which an embodiment of the present disclosure can be applied, and FIG. 3 (a) shows haar-wavelet features, and FIG. b) shows the LAB characteristic, and FIG. 3C shows the HOG characteristic.

First, the low-wavelet feature used to detect vehicle region candidates may be represented by a template 310 in the form of a (rectangular) rectangle, as shown in FIG. In addition, the lower wavelet feature is composed of two or three black and white squares and may be inclined horizontally, vertically or diagonally. Each of these low-wavelet features represents a component of vehicle region candidates.

Referring to FIG. 3B, an example of a LAB feature used to detect a final vehicle area is shown. As shown in FIG. 3B, in order to detect a feature corresponding to a specific part of the rear surface of the vehicle, the feature value is extracted by dividing the specific part into nine squares 320.

Referring to FIG. 3C, an example of a HOG feature used to detect a final vehicle area is shown. As shown in FIG. 3 (c), it can be seen that the final vehicle area is detected using more complex (or specific) features of the vehicle.

4 is a diagram illustrating a vehicle detection method using a stepwise method according to an embodiment of the present disclosure, and FIG. 4 (a) illustrates vehicle area candidates and / or final vehicle area using a coarse-to-fine method. 4 (b) is a diagram illustrating a vehicle detection result through a stepwise method.

Steps S110 and S140 are the same as those of FIG.

Referring to FIG. 4 (a), the vehicle detecting apparatus groups the at least one haar-wavelet feature or the at least one predetermined feature according to a predetermined criterion (S121, S131), and the grouping. Vehicle region candidates or final vehicle region may be detected by sequentially applying the first algorithm or the second algorithm to the specified features (S122 and S132).

Here, the predetermined feature refers to more complex features of the components of the vehicle capable of expressing texture or contour information of the vehicle area.

In addition, the predetermined criterion may be an element that exists according to the position of looking at the vehicle, an element that exists only in the vehicle, an element that exists only in a specific vehicle, and the like.

Referring to FIG. 4 (b), it can be seen that the wavelet features are grouped to represent the results of detecting vehicle region candidates sequentially using the first algorithm.

Embodiments and modifications described above may be combined. Accordingly, the embodiments may not be implemented alone, but may be implemented in combination as necessary. Such a combination can be easily implemented by those skilled in the art after reading the present specification, and the combination will not be described in detail below. However, even if not described, it is not to be excluded from the present invention, it should be construed as being included in the scope of the present invention.

Embodiments and modifications described above may be implemented through various means. For example, embodiments of the present invention may be implemented by hardware, firmware, software, or a combination thereof.

For a hardware implementation, the method according to embodiments of the present invention may be implemented in one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs) , Field programmable gate arrays (FPGAs), processors, controllers, microcontrollers, microprocessors, and the like.

In the case of an implementation by firmware or software, the method according to embodiments of the present invention may be implemented in the form of a module, a procedure or a function for performing the functions or operations described above. The software code may be stored in a memory unit and driven by a processor (or a controller). The memory unit may be located inside or outside the processor, and may exchange data with the processor by various well-known means.

For example, a method in accordance with the present invention may be implemented in software (e.g., flash memory, hard disk, etc.) that can be stored in a storage medium And may be implemented as codes or instructions within the program.

In addition, the controller 130 may include an application-specific integrated circuit (ASIC), another chipset, a logic circuit, and / or a data processing device.

When the embodiment is implemented in software, the above-described techniques may be implemented with modules (processes, functions, and so on) that perform the functions described above. The module may be stored in a memory and executed by a controller.

In addition, it is to be noted that the technical terms used herein are merely used to describe particular embodiments and are not intended to limit the present invention. It is also to be understood that the technical terms used herein are to be interpreted in a sense generally understood by a person skilled in the art to which the present invention belongs, Should not be construed to mean, or be interpreted in an excessively reduced sense. In addition, when the technical terms used herein are incorrect technical terms that do not accurately express the spirit of the present invention, they should be replaced with technical terms that can be understood correctly by those skilled in the art. In addition, the general terms used in the present invention should be interpreted according to a predefined or prior context, and should not be construed as being excessively reduced.

Also, the singular forms "as used herein include plural referents unless the context clearly dictates otherwise. In the present application, the term "comprising" or "comprising" or the like should not be construed as necessarily including the various elements or steps described in the specification, Or may be further comprised of additional components or steps.

Furthermore, terms including ordinals such as first, second, etc. used in this specification can be used to describe various elements, but the elements should not be limited by the terms. The terms are used only for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, the first component may be referred to as a second component, and similarly, the second component may also be referred to as a first component.

When an element is referred to as being "connected" or "connected" to another element, it may be directly connected or connected to the other element, but other elements may be present in between. On the other hand, when an element is referred to as being "directly connected" or "directly connected" to another element, it should be understood that there are no other elements in between.

The embodiments described above are the components and features of the present invention are combined in a predetermined form. Each component or feature shall be considered optional unless otherwise expressly stated. Each component or feature may be implemented in a form that is not combined with other components or features. It is also possible to construct embodiments of the present invention by combining some of the elements and / or features. The order of the operations described in the embodiments of the present invention may be changed. Some configurations or features of certain embodiments may be included in other embodiments, or may be replaced with corresponding configurations or features of other embodiments. It is clear that the claims that are not expressly cited in the claims may be combined to form an embodiment or be included in a new claim by an amendment after the application.

100: vehicle detection device
110:
120: control unit
130:
140: Memory

Claims (20)

In the method for detecting a vehicle,
Capturing an image;
Detecting vehicle region candidates indicating a vehicle shape from the captured image; And
Detecting a final vehicle region from the vehicle region candidates,
The detecting of the vehicle area candidates may include:
Learning at least one rectangular wave-wavelet feature through a first algorithm,
Detecting the final vehicle area,
Learning at least one predetermined feature from the detected vehicle region candidates through a second algorithm. The method of claim 1, wherein detecting the vehicle region candidates and the final vehicle region comprises:
A vehicle detection method characterized by using a coarse-to-fine method. The method of claim 2, wherein the stepwise method is
Grouping the at least one haar-wavelet feature or the at least one predetermined feature according to a predetermined criterion; And
And applying said first algorithm or said second algorithm sequentially to said grouped features. The method of claim 1,
And the first algorithm is an ADABOOST algorithm. The method of claim 1,
And said second algorithm is an ADABOOST algorithm or an SVM algorithm. The method of claim 1,
The predetermined feature is a vehicle detection method, characterized in that the texture or contour information of the vehicle representing the detailed characteristics of the vehicle. The method according to claim 6,
The predetermined feature is a vehicle detection method, characterized in that the Local Binary Pattern (LBP), LAB or Histogram of Oriented Gradients (HOG). The method of claim 1,
And outputting the detected final vehicle area. The method of claim 1,
Learning through the first algorithm or the second algorithm,
Vehicle detection method characterized in that performed using the stored learning data. The method of claim 9,
If the vehicle region candidates or the final vehicle region are not detected, updating the pre-stored learning data with data being learned by the first algorithm or the second algorithm; And
And storing the updated training data. In the vehicle detection device,
A camera unit for capturing an image; And
A control unit is functionally connected to the camera unit, wherein the control unit includes:
Controlling at least one rectangular ha-wavelet feature from the photographed image to detect vehicle region candidates indicating a vehicle shape through learning using a first algorithm, and at least one of the detected vehicle region candidates And detect a final vehicle region through learning using a second algorithm. The method of claim 11, wherein the control unit,
And detect the vehicle region candidates and the final vehicle region by using a coarse-to-fine method. The method of claim 12, wherein the control unit,
Grouping the at least one ha-wavelet feature or the at least one predetermined feature according to a predetermined criterion, and sequentially applying the first algorithm or the second algorithm to the grouped features And detect the vehicle region candidates and the final vehicle region. 12. The method of claim 11,
And the first algorithm is an ADABOOST algorithm. 12. The method of claim 11,
And said second algorithm is an ADABOOST algorithm or an SVM algorithm. 12. The method of claim 11,
The predetermined feature is a vehicle detection apparatus, characterized in that the texture or contour information of the vehicle representing the detailed characteristics of the vehicle. 17. The method of claim 16,
The predetermined feature is a vehicle detection device, characterized in that LBP, LAB or HOG (Histogram Of Gradients). 12. The method of claim 11,
Further comprising an output unit, The control unit,
And the output unit controls the output unit to output the detected final vehicle area. The method of claim 11, wherein the control unit,
And controlling to perform learning through the first algorithm or the second algorithm using previously stored learning data. 20. The method of claim 19,
Further comprising a memory, The control unit,
If the vehicle area candidates or the final vehicle area are not detected, the prestored learning data is updated with data being learned by the first algorithm or the second algorithm, and the memory is stored to store the updated learning data. Vehicle detection apparatus characterized in that for controlling.

Images