KR20160121026A - Apparatus and method for estimating distance to a pedestrian - Google Patents

Apparatus and method for estimating distance to a pedestrian Download PDF

Info

Publication number
KR20160121026A
KR20160121026A KR1020150050394A KR20150050394A KR20160121026A KR 20160121026 A KR20160121026 A KR 20160121026A KR 1020150050394 A KR1020150050394 A KR 1020150050394A KR 20150050394 A KR20150050394 A KR 20150050394A KR 20160121026 A KR20160121026 A KR 20160121026A
Authority
KR
South Korea
Prior art keywords
pedestrian
boundary
road surface
distance
estimating
Prior art date
Application number
KR1020150050394A
Other languages
Korean (ko)
Other versions
KR101684095B1 (en
Inventor
이강훈
이완재
최은진
Original Assignee
현대자동차주식회사
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 현대자동차주식회사 filed Critical 현대자동차주식회사
Priority to KR1020150050394A priority Critical patent/KR101684095B1/en
Publication of KR20160121026A publication Critical patent/KR20160121026A/en
Application granted granted Critical
Publication of KR101684095B1 publication Critical patent/KR101684095B1/en

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60R—VEHICLES, VEHICLE FITTINGS, OR VEHICLE PARTS, NOT OTHERWISE PROVIDED FOR
    • B60R21/00—Arrangements or fittings on vehicles for protecting or preventing injuries to occupants or pedestrians in case of accidents or other traffic risks
    • B60R21/34—Protecting non-occupants of a vehicle, e.g. pedestrians
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60R—VEHICLES, VEHICLE FITTINGS, OR VEHICLE PARTS, NOT OTHERWISE PROVIDED FOR
    • B60R21/00—Arrangements or fittings on vehicles for protecting or preventing injuries to occupants or pedestrians in case of accidents or other traffic risks
    • B60R21/01—Electrical circuits for triggering passive safety arrangements, e.g. airbags, safety belt tighteners, in case of vehicle accidents or impending vehicle accidents
    • B60R21/013—Electrical circuits for triggering passive safety arrangements, e.g. airbags, safety belt tighteners, in case of vehicle accidents or impending vehicle accidents including means for detecting collisions, impending collisions or roll-over
    • B60R21/0134—Electrical circuits for triggering passive safety arrangements, e.g. airbags, safety belt tighteners, in case of vehicle accidents or impending vehicle accidents including means for detecting collisions, impending collisions or roll-over responsive to imminent contact with an obstacle, e.g. using radar systems
    • G06T7/602—
    • B60R2021/34—
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10048—Infrared image

Landscapes

  • Engineering & Computer Science (AREA)
  • Mechanical Engineering (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Measurement Of Optical Distance (AREA)
  • Image Analysis (AREA)
  • Traffic Control Systems (AREA)

Abstract

The present invention relates to an apparatus and method for estimating a pedestrian distance, comprising the steps of: obtaining an input image by a far-infrared ray image sensor; setting a pedestrian zone in the input image; Estimating a symmetric boundary in the pedestrian zone as a boundary between a road surface and a pedestrian's toe when there is the road surface reflection; Estimating a pedestrian's longitudinal position based on the second estimated road surface and a pedestrian's foot boundary, calculating a pedestrian's longitudinal position based on the pedestrian's foot boundary estimated from the second estimated pedestrian's foot position, The method comprising the steps of: tracking a pedestrian distance from the vehicle based on the pedestrian distance; Quot;

Description

[0001] APPARATUS AND METHOD FOR ESTIMATING DISTANCE TO A PEDESTRIAN [0002]

The present invention relates to a pedestrian distance estimation apparatus and method for estimating and tracking a distance from a vehicle to a pedestrian based on a far-infrared ray sensor.

The existing pedestrian distance estimation method is performed by using a sensor that displays distance information of objects such as radar, Lidar, and stereo camera. However, distance measuring sensors such as radar and lidar have a disadvantage in that it is difficult to judge whether or not a pedestrian exists because of lack of information that can classify an object although the distance information of the object is known.

In addition, stereo cameras have information that can classify objects' distances and object types, but have a disadvantage in that the performance of the system is greatly degraded at nighttime or backlight conditions where illumination is low.

SUMMARY OF THE INVENTION The present invention has been made in order to solve the above problems of the prior art, and an object of the present invention is to provide a pedestrian distance estimating apparatus for estimating a distance from a vehicle to a pedestrian by detecting a boundary between a foot tip of a pedestrian and a road surface in a far-infrared ray image, And methods.

According to an aspect of the present invention, there is provided a pedestrian distance estimation method comprising: obtaining an input image by a far-infrared ray image sensor; setting a pedestrian area in the input image; And determining whether there is a road surface reflection in the pedestrian zone after the first estimation; and if there is a road surface reflection, determining a symmetric boundary in the pedestrian zone as a boundary between the road surface and the foot tip of the pedestrian Estimating a boundary between the road surface and the foot tip of the pedestrian based on the boundary between the road surface and the foot tip of the pedestrian; estimating a pedestrian longitudinal position using the second estimated road surface and the pedestrian foot boundary; Tracking a pedestrian distance from the vehicle based on the pedestrian longitudinal position, Use will be characterized by including the step of estimating a pedestrian lateral position.

The step of setting the pedestrian zone may further include setting the pedestrian zone by expanding the vertical area of the pedestrian to a predetermined magnification in the lower half direction in the input image.

According to another aspect of the present invention, there is provided a method for estimating a boundary between a road surface and a pedestrian's foot, comprising the steps of: calculating a brightness sum of pixels of each of the horizontal lines in the pedestrian area; And estimating a boundary between the road surface and the foot tip of the pedestrian.

The step of confirming the road surface reflection may further include the steps of applying a symmetric decision filter to the pedestrian zone on the basis of the boundary between the first estimated road surface and the foot tip of the pedestrian, And the step of selecting the step of selecting the step.

The step of estimating the boundary between the road surface and the foot tip of the pedestrian may include the steps of setting a toe area based on the first estimated boundary between the road surface and the foot tip of the pedestrian so as to enlarge the corresponding area at a predetermined ratio, Calculating a brightness distribution in the vertical direction of the enlarged area, and accurately estimating a boundary between the road surface and the toe of the pedestrian based on the vertical brightness distribution, .

The pedestrian distance tracking step continuously tracks the pedestrian distance using the pedestrian longitudinal position corrected for errors due to the vehicle speed and the pitch of the vehicle.

The step of tracing the pedestrian distance may further include calculating a tilt using the least square method on pedestrian distances estimated in a predetermined number of previous frames when the vehicle speed is less than a predetermined reference speed, And calculating the current pedestrian distance as an average of the pedestrian distance estimated in the current frame.

The pedestrian distance tracking step may determine a pedestrian distance of the current frame by subtracting the moving distance of the vehicle from the estimated pedestrian distance in the previous frame when the vehicle speed is equal to or higher than the predetermined reference speed.

The step of estimating the lateral position of the pedestrian is characterized by determining the lateral position of the pedestrian by modeling the relationship between the number of pixels and the pedestrian distance at which the center of the pedestrian falls from the horizontal center of the input image.

A pedestrian distance estimating apparatus according to an embodiment of the present invention includes a camera having a far infrared ray image sensor, a boundary estimator for estimating a boundary between a pedestrian's toe and a road surface in an input image acquired through the far infrared ray image sensor, And an image processor for calculating a distance from the vehicle to the pedestrian using the boundary between the road surface and the road surface.

The present invention can detect a boundary between a foot tip of a pedestrian and a road surface in a far infrared ray image and estimate the distance from the vehicle to the pedestrian using the detected boundary information.

In addition, the present invention can estimate an accurate pedestrian distance with a single image sensor in contrast to a stereo camera.

Further, the present invention corrects the distance error caused by the characteristics (e.g., road surface reflection) of the far-infrared image sensor.

Further, the present invention can calculate and correct a distance error generated from an image pixel value of an image sensor having an integer value up to a real value.

In addition, the present invention can correct the distance error caused by the pitch of the vehicle due to bumps, bumps, etc. of the road surface.

Further, since the distance can be estimated by one far-infrared sensor, the present invention can be utilized in a pedestrian collision and landing system without fusion with other distance sensors.

1 is a block diagram illustrating a pedestrian distance estimating apparatus according to an embodiment of the present invention.
FIG. 2 is a block diagram of the image processor shown in FIG. 1. FIG.
Fig. 3 is a view for explaining a boundary between a toe of a pedestrian and a road surface; Fig.
4 is a flowchart illustrating a pedestrian distance estimation method according to an embodiment of the present invention.
5 is a view for explaining a pedestrian area setting step shown in Fig.
FIG. 6 is a diagram for explaining a first-order estimation step of a road surface and a pedestrian toe boundary shown in FIG. 4; FIG.
7 is a view for explaining the road surface reflection determination step shown in Fig.
FIG. 8 is a diagram for explaining the second-order estimation step of the road surface and the toe of the foot tip shown in FIG. 4; FIG.
FIG. 9 is an exemplary view showing a screen outputting a pedestrian distance estimation result related to the present invention; FIG.

The terms "comprises", "comprising", "having", and the like are used herein to mean that a component can be implanted unless otherwise specifically stated, Quot; element ".

Also, the terms " part, "" module, " and" module ", as used herein, refer to a unit that processes at least one function or operation and may be implemented as hardware or software or a combination of hardware and software . It is also to be understood that the articles "a", "an", "an" and "the" Can be used.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

The present invention is a technology for estimating a distance of a pedestrian based on a far-infrared sensor and tracking the distance. When the center of the image sensor is parallel to the road surface and the lens focal length and the image sensor mounting height are known, the distance is estimated by searching the far infrared ray image for the boundary between the pedestrian's foot and the road surface. Particularly, in order to more accurately calculate the pedestrian distance, it is most important that the boundary between the pedestrian's foot and the road surface is sophisticated. However, due to the characteristics of the far-infrared ray image, the pedestrian is reflected on the road surface or the ankle part appears dark, and the detected foot part of the pedestrian is estimated to be long or short. In addition, the distance from the pedestrian is a real number, while the pixel coordinate value of the far-infrared sensor image is an integer, so an error occurs in the calculation process. Accordingly, the present invention proposes a method of detecting a pedestrian's foot / foot boundary and estimating and tracking a pedestrian's longitudinal / lateral distance.

FIG. 1 is a block diagram of a pedestrian distance estimating apparatus according to an embodiment of the present invention. FIG. 2 is a block diagram of the image processor shown in FIG. 1. FIG. FIG.

As shown in FIG. 1, the pedestrian distance estimating apparatus includes a camera 110, an image processing unit 120, and an output unit 130.

The camera 110 is mounted on a vehicle and photographs images of the surroundings of the vehicle. The camera 110 includes a far-infrared image sensor. The camera 110 is installed such that the center line of the far-infrared image sensor is horizontal with the road surface. In other words, the camera 110 is installed so that a straight line perpendicular to the center surface of the camera lens is horizontal to the road surface.

The image processing unit 120 checks whether a pedestrian exists in an input image (e.g., a far-infrared ray image) input from the camera 110. [ The image processing unit 120 estimates the pedestrian distance using the camera geometry model if a pedestrian exists in the input image. In other words, the image processing unit 120 calculates the pedestrian distance d using the mounting height H of the camera 110 and the focal length f of the lens, and the vertical distance y from the horizontal center line of the image in the input image to the boundary between the road surface and the foot tip of the pedestrian . At this time, the finally calculated pedestrian distance is determined as a coordinate value in the input image.

2, the image processing unit 120 includes a pedestrian toe boundary estimation module 121, a road surface reflection determination module 122, a pedestrian foot boundary precision estimation module 123, a pedestrian longitudinal position estimation module 124, A pedestrian distance tracking module 125, an error correction module 126, and a pedestrian lateral position estimation module 127.

The pedestrian toe boundary estimation module 121 sets a pedestrian area when a pedestrian is detected in the input image. At this time, the pedestrian toe boundary estimation module 121 expands the vertical region of the pedestrian to a predetermined magnification in the lower half direction.

The pedestrian toe boundary estimation module 121 first estimates the boundary between the road surface and the pedestrian's toe in the pedestrian area. The pedestrian toe boundary estimation module 121 calculates the one-dimensional brightness distribution by calculating the sum of pixel brightness values of the pedestrian area by line by line. The pedestrian toe boundary estimation module 121 estimates a point at which the brightness is suddenly brighter than the reference in the one-dimensional brightness distribution as the first boundary between the road surface and the pedestrian's foot. In other words, the pedestrian toe boundary estimation module 121 differentiates the one-dimensional brightness distribution function and estimates the point where the differential value (slope) approaches "0" as the first boundary between the road surface and the foot tip of the pedestrian.

The road surface reflection determination module 122 sets a certain region as a region of interest (ROI) based on the first estimated boundary and applies a symmetric decision filter to the ROI. The road surface reflection determination module 122 selects a point within a range of the filter application value within a reference range (e.g., 0 1). The road surface reflection determination module 122 transmits the primary boundary estimation value to the pedestrian toe boundary precision estimation module 123 if the sum of the filter application values does not fall within the reference range.

The pedestrian toe boundary precision estimation module 123 resets a predetermined area (e.g., upper and lower N pixels on the boundary reference) based on the first estimated boundary, enlarges the area to a certain magnification (for example, M times), and applies a Gaussian filter (smoothing).

The pedestrian toe boundary precision estimation module 123 calculates the vertical brightness distribution of the toe enlargement area. The pedestrian toe boundary precision estimation module 123 searches upward from the lower end of the brightness distribution and determines a starting point as a point at which the inclination becomes larger than a reference point and sets the middle point as a point where the inclination changes to 0, .

When the boundary between the road surface and the pedestrian's foot is determined, the pedestrian longitudinal position estimation module 124 may divide the coordinate system of the boundary into M and convert it into the existing coordinate system to determine the final road surface and the pedestrian's foot boundary value y to the decimal point unit.

The pedestrian distance tracking module 125 calculates the final pedestrian distance d by substituting the final road surface and the pedestrian foot boundary value y into the following equation (1).

Figure pat00001

Here, f is the focal length of the camera lens, and H is the mounting height of the camera 110. [

The pedestrian distance tracking module 125 continuously tracks the pedestrian distance based on the vehicle speed (vehicle speed).

If the vehicle speed is less than the reference speed, the pedestrian distance tracking module 125 estimates the inclination (pedestrian distance change rate) using the Least Square Method (LSM) on the pedestrian distances calculated in a certain number of previous frames. Then, the pedestrian distance tracking module 125 calculates the current pedestrian distance as an average value of the pedestrian distance calculated from the estimated slope and the image-based pedestrian distance in the current frame as shown in Equation (2).

Figure pat00002

here,

Figure pat00003
Is the pedestrian distance calculated from the current frame to the toe coordinates,
Figure pat00004
Is the pedestrian distance predicted by the previous frame information,
Figure pat00005
Is the distance of the pedestrian (object) in the nth frame.

When the vehicle speed exceeds the reference speed, the moving speed of the pedestrian (for example, 4 km / h) is smaller than the vehicle speed and can be ignored. Therefore, the pedestrian distance tracking module 125 calculates the pedestrian distance < RTI ID = 0.0 >

Figure pat00006
Distance of the vehicle from
Figure pat00007
To the pedestrian distance in the current frame
Figure pat00008
.

Figure pat00009

Where v is the difference and t is the time.

The pedestrian distance tracking module 125 updates the pedestrian distance estimated based on the far-infrared image every frame. The pedestrian distance tracking module 125 calculates and updates the pedestrian distance using Equation (4).

Figure pat00010

here,

Figure pat00011
And
Figure pat00012
Is the weight.

The error correction module 126 corrects an error due to the pitch of the vehicle (e.g., bump and road surface curvature).

The error correction module 126 determines that there is a curve on the road surface when the difference in distance between the inter-frame pedestrian (object) is larger than the speed of the vehicle. That is, the error correction module 126

Figure pat00013
It is judged that there is a curve on the road surface.

The error correction module 126 reapplies the pedestrian distance tracking module 125 to the pedestrian distance tracking module 125 when the pedestrian distance change occurs linearly with respect to the vehicle speed for a predetermined frame.

The pedestrian lateral position estimation module 127 determines the lateral position of the pedestrian by modeling the relationship between the number of pixels and the pedestrian distance of the center of the pedestrian from the horizontal center line of the image as a function Dx = F (x, y). Where x is the number of horizontal pixels vertically away from the center of the image and y is the longitudinal distance of the object.

The output unit 130 outputs one or more pieces of information such as time information, auditory information, and tactile information. The output unit 130 may include a display device and an audio output device.

The display device displays (outputs) information processed by the pedestrian distance estimating device and UI (User Interface) or GUI (Graphic User Interface) for controlling the operation of the pedestrian distance estimating device.

The display device 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, Display means such as a three-dimensional display (3D display), an electronic ink display (e-ink display), a transparent display, a touch screen and the like.

The acoustic output device also outputs an acoustic signal related to a function performed by the pedestrian distance estimation device. Such a sound output apparatus may include a receiver, a speaker, a buzzer, and the like.

FIG. 4 is a flowchart illustrating a pedestrian distance estimating method according to an embodiment of the present invention. FIG. 5 is a view for explaining a pedestrian area setting step shown in FIG. 4. FIG. FIG. 7 is a view for explaining the step of judging the road surface reflection shown in FIG. 4, and FIG. 8 is a view for explaining the step of estimating the pedestrian toe boundary second estimating step FIG. 9 is an exemplary view showing a screen outputting a pedestrian distance estimation result related to the present invention. FIG.

As shown in FIG. 4, the image processing unit 120 of the pedestrian distance estimating apparatus obtains an input image including a pedestrian through the camera 110 (S101).

The image processing unit 120 sets a pedestrian area in the input image (S103). At this time, the image processing unit 120 sets the pedestrian area by extending the vertical area of the pedestrian in the lower half direction at a predetermined magnification from the input image. For example, as shown in FIG. 5, the image processing unit 120 detects a pedestrian in an input image, and sets a pedestrian area R p by extending a vertical region of the pedestrian by a predetermined magnification in the lower half direction.

The image processing unit 120 firstly estimates the boundary between the road surface and the foot tip of the pedestrian in the pedestrian area (S105). Image processing unit 120 is a road surface and the perimeter of the pedestrian toe L b as shown in FIG. 6 a point which calculates a sum Horizontal pixel brightness per line of the pedestrian area by calculating the brightness distribution and the brightness is brighter than the reference in the brightness distribution .

The image processing unit 120 confirms whether there is road surface reflection in the pedestrian area after the first estimation (S107). As shown in FIG. 7, the image processing unit 120 sets a ROI based on a boundary L1 between a first estimated road surface and a pedestrian's foot, and applies a symmetric decision filter to the ROI . The image processing unit 120 checks whether the sum of the results of applying the symmetric decision filter to the ROI is within the reference range.

If there is a road surface reflection, the image processing unit 120 estimates a symmetric boundary by the road surface reflection (S109). The image processing unit 120 detects a point within a reference range of the sum of the resultant values obtained by applying the symmetric decision filter to the ROI and outputs it as a boundary first estimate value of the road surface and the foot tip of the pedestrian. That is, the image processing unit 120 outputs the symmetric boundary due to the road surface reflection to the boundary between the first estimated road surface and the foot tip of the pedestrian.

On the other hand, if there is no road surface reflection, the image processing unit 120 performs a second estimation (precision estimation) using the result of primary estimation of the boundary between the road surface and the pedestrian's foot (S111).

The image processing unit 120 secondarily estimates the boundary between the road surface and the foot tip of the pedestrian based on the boundary between the first estimated road surface and the foot tip of the pedestrian (S111). At this time, the image processing unit 120 sets a certain area to the foot area R f based on the first estimated boundary between the road surface and the pedestrian foot or the estimated road surface reflection, and enlarges the corresponding area at a certain ratio . Then, the image processing unit 120 calculates the brightness sum of each of the vertical lines in the enlarged area to obtain the brightness distribution. The image processing unit 120 calculates the boundary between the pedestrian's foot and the road surface based on the obtained brightness distribution.

The image processing unit 120 estimates the longitudinal position of the pedestrian based on the boundary between the second estimated road surface and the foot tip of the pedestrian (S113). That is, the image processing unit 120 calculates the distance y from the center of the input image to the boundary between the road surface estimated secondary from the center of the input image and the foot tip of the pedestrian.

The image processing unit 120 tracks the pedestrian distance using the pedestrian's longitudinal position (S115). In other words, the image processing unit 120 tracks the distance from the vehicle to the pedestrian using the mounting height and focal length of the camera 110, and the position of the pedestrian longitudinal direction. At this time, the image processing unit 120 continuously corrects the pedestrian distance by correcting the pedestrian longitudinal position (distance) error due to the noise and the pitch of the vehicle.

The image processing unit 120 estimates the lateral distance (position) of the pedestrian using the horizontal pixel difference between the center of the input image and the center of the pedestrian and the vertical distance of the pedestrian (S117).

The image processing unit 120 displays the estimated longitudinal distance and the lateral distance of the pedestrian by matching the corresponding pedestrian as shown in FIG. Then, the image processing unit 120 terminates the pedestrian distance estimation when the pedestrian (object) leaves the photographing range of the camera 110.

The embodiments described above are those in which the elements 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.

Embodiments in accordance with the present invention may be implemented by various means, for example, hardware, firmware, software, or a combination thereof. In the case of hardware implementation, an embodiment of the present invention may include 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, processors, controllers, microcontrollers, microprocessors, and the like.

In the case of an implementation by firmware or software, an embodiment of the present invention may be implemented in the form of a module, a procedure, a function, or the like which performs the functions or operations described above. The software code can be stored in a memory unit and driven by the processor. The memory unit is located inside or outside the processor, and can exchange data with the processor by various known means.

It will be apparent to those skilled in the art that the present invention may be embodied in other specific forms without departing from the spirit of the invention. Accordingly, the foregoing detailed description is to be considered in all respects illustrative and not restrictive. The scope of the present invention should be determined by rational interpretation of the appended claims, and all changes within the scope of equivalents of the present invention are included in the scope of the present invention.

100: Pedestrian distance estimation device
110: camera
120: image processor
130:

Claims (10)

Acquiring an input image with a far infrared ray image sensor,
Setting a pedestrian zone in the input image,
Estimating a boundary between the road surface and the foot tip of the pedestrian in the pedestrian zone;
Determining whether there is a road surface reflection in the pedestrian zone after the primary estimation;
Estimating a symmetric boundary in the pedestrian zone as a boundary between a road surface and a pedestrian's foot if there is the road surface reflection;
Estimating a boundary between the road surface and the foot tip of the pedestrian based on the boundary between the road surface and the toe of the pedestrian;
Estimating a pedestrian longitudinal position using the second presumed road surface and pedestrian toe boundary;
Tracking a pedestrian distance from the vehicle based on the pedestrian longitudinal position,
And estimating a pedestrian's lateral position using the pedestrian distance.
The method according to claim 1,
Wherein the pedestrian zone setting step comprises:
Wherein the pedestrian zone is set by extending a vertical region of the pedestrian in the lower half of the input image to a predetermined magnification.
The method according to claim 1,
The boundary estimating step of the boundary between the road surface and the pedestrian's foot,
Calculating a brightness sum of pixels of each of the horizontal lines in the pedestrian zone;
Detecting a point at which a brightness is brighter than a reference in a brightness distribution of the pedestrian zone based on the sum of brightness, and estimating a boundary between the road surface and a pedestrian's foot.
The method of claim 3,
In the road surface reflection confirmation step,
Applying a symmetric decision filter to the pedestrian zone on the basis of the boundary between the first estimated road surface and the foot tip of the pedestrian;
And selecting a symmetric boundary point based on the result of the application of the symmetric decision filter.
The method according to claim 1,
The boundary estimating step of the boundary between the road surface and the pedestrian's foot,
Setting a toe area on the basis of the boundary between the first estimated road surface and the foot tip of the pedestrian so as to enlarge the corresponding area at a predetermined ratio;
Applying a Gaussian filter to the enlarged toe area,
Calculating a vertical brightness distribution of the enlarged area;
And accurately estimating a boundary between the road surface and the foot tip of the pedestrian based on the vertical brightness distribution.
The method according to claim 1,
The pedestrian distance tracking step includes:
Wherein the pedestrian distance is continuously tracked using a pedestrian longitudinal position corrected for errors due to a vehicle speed and a pitch of the vehicle.
The method according to claim 6,
The pedestrian distance tracking step includes:
Calculating a slope using a least square method on pedestrian distances estimated in a certain previous frame when the vehicle speed is less than a predetermined reference speed;
And calculating a current pedestrian distance based on an average of the pedestrian distance calculated from the slope and the pedestrian distance estimated in the current frame.
The method according to claim 6,
The pedestrian distance tracking step includes:
Wherein a pedestrian distance of the current frame is determined by subtracting the moving distance of the vehicle from the estimated pedestrian distance in the previous frame when the vehicle speed is equal to or greater than a predetermined reference speed.
The method according to claim 1,
Wherein the step of estimating the pedestrian's lateral position comprises:
Wherein the lateral position of the pedestrian is determined by modeling the relationship between the number of pixels of the pedestrian falling from the horizontal center of the input image and the pedestrian distance as a function.
A camera having a far-infrared image sensor,
And an image processor for estimating a boundary between the pedestrian's foot and the road surface in the input image acquired through the far infrared ray image sensor and calculating the distance from the vehicle to the pedestrian using the estimated boundary between the toe and the road surface A pedestrian distance estimating device.
KR1020150050394A 2015-04-09 2015-04-09 Apparatus and method for estimating distance to a pedestrian KR101684095B1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
KR1020150050394A KR101684095B1 (en) 2015-04-09 2015-04-09 Apparatus and method for estimating distance to a pedestrian

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
KR1020150050394A KR101684095B1 (en) 2015-04-09 2015-04-09 Apparatus and method for estimating distance to a pedestrian

Publications (2)

Publication Number Publication Date
KR20160121026A true KR20160121026A (en) 2016-10-19
KR101684095B1 KR101684095B1 (en) 2016-12-07

Family

ID=57250480

Family Applications (1)

Application Number Title Priority Date Filing Date
KR1020150050394A KR101684095B1 (en) 2015-04-09 2015-04-09 Apparatus and method for estimating distance to a pedestrian

Country Status (1)

Country Link
KR (1) KR101684095B1 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113246931A (en) * 2021-06-11 2021-08-13 创新奇智(成都)科技有限公司 Vehicle control method and device, electronic equipment and storage medium
US11320830B2 (en) 2019-10-28 2022-05-03 Deere & Company Probabilistic decision support for obstacle detection and classification in a working area

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2001101420A (en) * 1999-09-27 2001-04-13 Toshiba Corp Device and method for detecting obstacle
JP2005079999A (en) * 2003-09-01 2005-03-24 Honda Motor Co Ltd Vehicle surroundings monitoring unit
JP2007214806A (en) * 2006-02-08 2007-08-23 Sumitomo Electric Ind Ltd Obstacle detection system and obstacle detecting method
KR20100113371A (en) * 2009-04-13 2010-10-21 주식회사 만도 Method and apparatus for guarding pedestrian using far-infra-red stereo camera

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2001101420A (en) * 1999-09-27 2001-04-13 Toshiba Corp Device and method for detecting obstacle
JP2005079999A (en) * 2003-09-01 2005-03-24 Honda Motor Co Ltd Vehicle surroundings monitoring unit
JP2007214806A (en) * 2006-02-08 2007-08-23 Sumitomo Electric Ind Ltd Obstacle detection system and obstacle detecting method
KR20100113371A (en) * 2009-04-13 2010-10-21 주식회사 만도 Method and apparatus for guarding pedestrian using far-infra-red stereo camera

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11320830B2 (en) 2019-10-28 2022-05-03 Deere & Company Probabilistic decision support for obstacle detection and classification in a working area
CN113246931A (en) * 2021-06-11 2021-08-13 创新奇智(成都)科技有限公司 Vehicle control method and device, electronic equipment and storage medium
CN113246931B (en) * 2021-06-11 2021-09-28 创新奇智(成都)科技有限公司 Vehicle control method and device, electronic equipment and storage medium

Also Published As

Publication number Publication date
KR101684095B1 (en) 2016-12-07

Similar Documents

Publication Publication Date Title
US10755417B2 (en) Detection system
US9576367B2 (en) Object detection method and device
JP4919036B2 (en) Moving object recognition device
CN110793544B (en) Method, device and equipment for calibrating parameters of roadside sensing sensor and storage medium
JP5503578B2 (en) Object detection apparatus and object detection method
US8126210B2 (en) Vehicle periphery monitoring device, vehicle periphery monitoring program, and vehicle periphery monitoring method
WO2015125298A1 (en) Local location computation device and local location computation method
US9886933B2 (en) Brightness adjustment system and method, and mobile terminal
KR101769177B1 (en) Apparatus and method for eye tracking
JP2007235950A (en) Method and device for detecting vehicle position
JP2004117078A (en) Obstacle detection device and method
JP2005072888A (en) Image projection method and image projection device
JP2011128756A (en) Object detection device
JP5874831B2 (en) Three-dimensional object detection device
JP2016200557A (en) Calibration device, distance measurement apparatus and calibration method
JP6561670B2 (en) Step detecting device and step detecting method
JP6396499B2 (en) Scale measurement of 3D information
US6697146B2 (en) Range finder for finding range by image realization
JP2021033510A (en) Driving assistance device
KR101684095B1 (en) Apparatus and method for estimating distance to a pedestrian
JP6188860B1 (en) Object detection device
US20160255266A1 (en) Focus position detection device and focus position detection method
JP2013092820A (en) Distance estimation apparatus
CN104537627A (en) Depth image post-processing method
JP6398217B2 (en) Self-position calculation device and self-position calculation method

Legal Events

Date Code Title Description
A201 Request for examination
E902 Notification of reason for refusal
E701 Decision to grant or registration of patent right
GRNT Written decision to grant
FPAY Annual fee payment

Payment date: 20191127

Year of fee payment: 4