KR20220014083A - Road detection method and apparatus using angle of polarization in infrared imaging - Google Patents

Abstract

A road detection method and apparatus are disclosed. The road detection method includes the steps of: (a) generating four polarization images having different angles by preprocessing an input image; (b) calculating a degree of polarization and a polarization angle using the four polarization images; and (c) generating a precise road map by combining the polarization degree, the polarization angle, and edge information, and detecting and dividing the road in the input image using the precise road map.

Description

Road detection method and apparatus using angle of polarization in infrared imaging

The present invention relates to a method and apparatus for detecting a road area using zero distribution of polarization angles of an infrared image.

Road sensing is an important function in traffic safety and intelligent traffic systems such as driver assistance systems (ADAS). Vision-based road detection methods include monocular and stereo imaging technologies and provide high-resolution color, texture and lane marking information for road detection. Although stereo imaging technology can acquire depth information, there is a problem that normal operation is impossible in low-light conditions, in the dark or in strong lighting changes such as headlights of vehicles entering at night.

Korean Patent Publication No. 10-1766239 (2017.08.02.)

An object of the present invention is to provide a road area detection method and apparatus capable of stably detecting a road area in an image during the day and at night by using polarization information of an infrared image.

Another object of the present invention is to provide a road area detection method and apparatus capable of providing information necessary for autonomous driving by enabling automatic identification and detection of a road area even when lighting is insufficient, such as at night.

According to one aspect of the present invention, there is provided a method for detecting a road area using a polarization angle of an infrared image.

According to an embodiment of the present invention, the method comprising: (a) preprocessing an input image to generate four polarized images having different angles; (b) calculating a polarization degree and a polarization angle using the polarization images corresponding to the four angles; and (c) generating a precise road area map by combining the polarization degree, the polarization angle, and edge information, and detecting and dividing the road area in the input image using the precise road area map. A detection method may be provided.

The input image is an image obtained through a Division of Focal Plane (DoFP) image sensor, and in step (a), the input image is subjected to polarization demosaicking (PDM) at 0 degrees, 45 degrees, 90 degree and 135 degree polarization images can be generated.

Before step (b), deriving a first Stokes parameter for the total intensity of light using the 0 degree 0 degree, 45 degree, 90 degree and 135 degree polarization image; deriving a second Stokes parameter for an amount of linearly polarized light using the 0 degree and 90 degree polarization images; and deriving a third Stokes parameter for the amount of linearly polarized light using the 45 degree and 135 degree polarization images.

The step (b) may include deriving a degree of polarization using the first Stokes parameter to the third Stokes parameter; and deriving a polarization angle using the second Stokes parameter and the third Stokes parameter.

The step (c) may include: obtaining a schematic road area map using the zero distribution constraint of the polarization angle, and deriving a horizon using the rough road area map; determining a region of interest using the horizon, and generating a region map with precision by combining the polarization degree, the polarization angle, and edge information to the region of interest; and detecting a road area from the input image by using the precision road area map and then dividing the detected road area.

Before deriving the horizon, noise may be removed by performing a morphological open operation on the rough road area map.

According to another aspect of the present invention, an apparatus for detecting a road area using a polarization angle of an infrared image is provided.

According to an embodiment of the present invention, there is provided a pre-processing unit that pre-processes an input image to generate four polarized images having different angles from each other; a calculation unit for calculating a polarization degree and a polarization angle using the four polarized images; and a road area detection unit generating a precise road area map by combining the polarization degree, the polarization angle, and edge information, and detecting and dividing the road area in the input image using the precise road area map. A device may be provided.

A Division of Focal Plane (DoFP) image sensor configured to acquire the input image may be further included.

The preprocessor may generate 0 degree, 45 degree, 90 degree, and 135 degree polarization images on the input image through a polarization demosaicking (PDM) process.

The calculator derives a first Stokes parameter for the total intensity of light using the 0 degree to 135 degree polarization image, and a second Stokes for the amount of linear polarization using the 0 degree and 90 degree polarization images s parameter is derived, and a third Stokes parameter for the amount of linear polarization may be derived using the 45 degree and 135 degree polarization images.

The calculator may derive a degree of polarization using the first Stokes parameter to the third Stokes parameter, and derive a polarization angle using the second Stokes parameter and the third Stokes parameter.

The road area detection unit obtains a rough road area map using the zero distribution constraint of the polarization angle, derives a horizon by using the rough road area map, determines a region of interest using the horizon, , a precision road area map is generated by combining the polarization degree, the polarization angle, and edge information with the ROI, and a road area is detected from the input image using the precision road area map and then segmented.

The road area detector may remove noise by performing a morphological open operation on the rough road area map.

By providing a method and apparatus for detecting a road area using an infrared polarization angle according to an embodiment of the present invention, there is an advantage in that it is possible to stably detect a road area in an image during the day and at night using polarization information of an infrared image. .

In addition, the present invention also has an advantage in that it is possible to automatically identify and detect a road area even when lighting is insufficient, such as at night, to provide information necessary for autonomous driving.

1 is a flowchart illustrating a road area detection method according to an embodiment of the present invention.
2 is a view for explaining the acquisition of a polarized image according to an embodiment of the present invention.
3 is a diagram illustrating an image and a polarization angle image obtained by a DoFP image sensor according to an embodiment of the present invention.
4 is a diagram illustrating a zero distribution characteristic according to an embodiment of the present invention.
5 is a view showing an intensity image according to an embodiment of the present invention.
6 is a view showing a polarization degree image according to an embodiment of the present invention.
7 is a view showing a polarization angle image according to an embodiment of the present invention.
8 is a view showing a schematic road area map according to an embodiment of the present invention;
9 is a view showing a precision road area map according to an embodiment of the present invention.
10 is a view showing a road area detection result according to an embodiment of the present invention.
11 is a block diagram schematically illustrating an internal configuration of an apparatus for detecting a road area according to an embodiment of the present invention;

As used herein, the singular expression includes the plural expression unless the context clearly dictates otherwise. In this specification, terms such as "consisting of" or "comprising" should not be construed as necessarily including all of the various components or various steps described in the specification, some of which components or some steps are It should be construed that it may not include, or may further include additional components or steps. In addition, terms such as "...unit" and "module" described in the specification mean a unit that processes at least one function or operation, which may be implemented as hardware or software, or a combination of hardware and software. .

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

1 is a flowchart illustrating a method for detecting a road area according to an embodiment of the present invention, FIG. 2 is a diagram illustrating acquisition of a polarized image according to an embodiment of the present invention, and FIG. 3 is an embodiment of the present invention It is a diagram illustrating an image and a polarization angle image obtained by a DoFP image sensor according to an embodiment, and FIG. 4 is a diagram illustrating a zero distribution characteristic according to an embodiment of the present invention, and FIG. 5 is a diagram of the present invention shows an intensity image according to an embodiment of the present invention, FIG. 6 shows a polarization degree image according to an embodiment of the present invention, FIG. 7 shows a polarization angle image according to an embodiment of the present invention, and FIG. 8 shows the present invention shows a schematic road area map according to an embodiment of the present invention, FIG. 9 shows a precision road area map according to an embodiment of the present invention, and FIG. 10 shows a road area detection result according to an embodiment of the present invention.

In step 110, the road area detection apparatus 100 acquires an image.

According to an embodiment of the present invention, the road area detection apparatus 100 may acquire an image using a Division of Focal Plane (DoFP) image sensor.

As shown in Fig. 2(a), since the road plane is assumed to be parallel to the Y-axis in the image coordinate system, the zero distribution of the polarization angle means that the road is partially linearly polarized parallel to the xz plane of the road.

Fig. 2(b) shows that the thermal radiation reaching the DoFP LWIR image sensor includes reflected radiation R and emitted radiation E, both of which are expressed as the sum of orthogonally polarized components.

According to an embodiment of the present invention, the DoFP image sensor includes a 4-direction micro-polarization array filter. Here, the four directions may be 0 degrees, 45 degrees, 90 degrees, and 135 degrees.

An example of a road image obtained through such a DoFP image sensor is illustrated in FIG. 3 .

In operation 115 , the road area detecting apparatus 100 reconstructs a polarized image having polarization components of various angles through a polarization demosaicing (PDM) process for an image obtained through the DoFP image sensor.

For example, 0 degree, 45 degree, 90 degree, and 135 degree polarization images may be reconstructed.

In operation 120 , the road area detecting apparatus 100 calculates first, second, and third Stokes coefficients using the polarized image.

)를 계산할 수 있다(도 5 참조). 또한, 도로 영역 검출 장치(100)는 0도와 90도 편광 영상의 차이를 도출하여 제2 스토크스 계수(

)를 계산할 수 있으며, 45도와 135도 편광 영상의 차이를 도출하여 제3 스토크스 계수(

)를 계산할 수 있다.For example, the road area detection apparatus 100 uses the first Stokes coefficient (

) can be calculated (see Fig. 5). In addition, the road area detection apparatus 100 derives the difference between the 0 degree and 90 degree polarized images to obtain a second Stokes coefficient (

) can be calculated, and the third Stokes coefficient (

) can be calculated.

If this is expressed as an equation, it can be expressed as in Equations 1 to 2.

)는 빛의 전체 강도를 나타낸다. Here, the first Stokes coefficient (

) represents the total intensity of light.

는 0도 편광 영상을 나타내고,

는 45도 편광 영상을 나타내며,

는 90도 편광 영상을 나타내고,

는 135도 편광 영상을 나타낸다. here,

represents a 0 degree polarized image,

represents a 45 degree polarized image,

represents a 90 degree polarized image,

represents a 135 degree polarized image.

)와 제3 스토크스 계수(

)는 선형 편광의 양을 나타낸다. That is, the second Stokes coefficient (

) and the third Stokes coefficient (

) represents the amount of linearly polarized light.

In operation 125, the road area detecting apparatus 100 calculates a polarization degree and a polarization angle using at least two of the first to third Stokes coefficients, respectively.

For example, the road area detecting apparatus 100 may derive the degree of polarization using Equation 4 below. 6 illustrates an image of the degree of polarization.

Also, the road area detecting apparatus 100 may derive the polarization angle using Equation 5. 7 illustrates a polarization angle image.

The zero distribution is based on the theory that the heat dissipation is partially linearly polarized parallel to the plane of incidence. In the case of the road area, most of the polarization angles are close to "0" (refer to FIG. 4).

내의 값을 가지며, 도로 영역에서의 편광 각도는 거의 대부분이 "0"이다. As a result of analyzing the example of FIG. 3 , as shown in FIG. 4 , about 96% of the pixels of the polarization angle image are

, and most of the polarization angles in the road area are “0”.

If this is expressed as an equation, it is as shown in equation (6).

는 도로 영역내의 픽셀(x) 집합을 나타낸다. here,

denotes a set of pixels (x) in the road area.

Accordingly, the road area can be detected by finding pixels whose polarization angle is almost 0 in the image using the zero distribution of the polarization angle.

It will be understood more clearly by the following description.

)을 획득하고, 이를 이용하여 지평선을 구할 수 있다.In step 130, the road area detection apparatus 100 uses a polarization angle constraint to obtain a schematic road area map (

) and use it to find the horizon.

)에서 원근법을 적용하여 도로가 점차 좁아지는 소실점을 구한 후 소실점을 이용하여 지평선을 구하며, 지평선을 이용하여 관심 영역을 설정하여 도로 영역 검출에 불필요한 계산을 줄여 계산량을 감소시킬 수도 있다. 개략적인 도로 영역 맵은 도 8에 도시되어 있다. For example, the road area detection apparatus 100 may provide a schematic road area map (

) by applying the perspective method to find the vanishing point at which the road gradually narrows, then using the vanishing point to find the horizon, and by using the horizon to set the region of interest, it is also possible to reduce the amount of calculation by reducing unnecessary calculations for road region detection. A schematic road area map is shown in FIG. 8 .

)는 수학식 7과 같이 계산될 수 있다. For example, a schematic road area map (

) can be calculated as in Equation 7.

는 0.01로 설정되며,

는 사전 제로 분포에 기초한 실험에서 0으로 설정되었다. 이어, 도로로서 임계값을

로 정하고, 작은 노이즈 영역을 제거하기 위해 형태학적 개방 연산(morphological open operation)을 수행할 수 있다. here,

is set to 0.01,

was set to zero in the experiment based on the prior zero distribution. Then, the threshold as a road

, and a morphological open operation can be performed to remove a small noise region.

This can be expressed as Equation (8).

는 형태학적 개방 연산을 나타내며, X는 구조적 엘리먼트를 나타낸다.

는 개략적인 도로 검출 결과를 제공하고, 지평선은 도로 소실점에 위치한다. 도로 소실점은 각각의 행에서 도로가 속하는 픽셀의 개수가 아래에서 위쪽으로 감소한다. here,

denotes a morphological open operation, and X denotes a structural element.

gives a rough road detection result, and the horizon is located at the road vanishing point. In the road vanishing point, the number of pixels to which the road belongs in each row decreases from bottom to top.

)의 1차원 투영은 수학식 9과 같이 도출될 수 있다. Schematic road area map (

) can be derived as in Equation (9).

는 개략적인 도로 영역 맵(

)에서 각 행에서 도로에 속하는 픽셀의 개수를 나타낸다. Here, W represents the width of the image,

is a schematic road area map (

) indicates the number of pixels belonging to the road in each row.

를 이용하여 지평선을 찾을 수 있다. 이를 수학식으로 나타내면, 수학식 10과 같다. The road area detection device 100 is

can be used to find the horizon. If this is expressed as an equation, it is the same as in equation (10).

과

는 행(row) 번호를 나타내고,

는 직선과

에서 x축에서 교차되는 행(row)을 나타낸다. 본 발명의 일 실시예에서는 두개의 행이 있는 두개의 픽셀들을 모두 선택할 수 있다. 또한,

이며, 대괄호안의 값이 사실이면

는 1이며 그렇지 않은 경우 0이다. here,

class

represents the row number,

is a straight line and

represents a row that intersects on the x-axis. In an embodiment of the present invention, both pixels in two rows may be selected. Also,

, and if the value in square brackets is true

is 1, otherwise 0.

It can be seen that the largest or dominant value in H is a few pixels away from the ground horizon. Therefore, in order to overcome this problem, a simple post-correction as shown in Equation 11 is performed.

는 H의 정밀한 누적 에너지를 나타내고,

으로 설정될 수 있다.

에서 가장 큰 에너지를 찾아 지평선으로 설정할 수 있다. here,

represents the precise cumulative energy of H,

can be set to

Find the greatest energy in , and set it to the horizon.

)을 생성하고, 이를 이용하여 영상내 도로 영역을 검출한 후 검출된 도로 영역을 영상 분할하여 화면을 통해 출력할 수 있다. In step 135, the road area detection apparatus 100 determines a region of interest using the horizon, and combines the polarization degree, polarization angle, and edge information with the region of interest to obtain a precision road region map (

), detects a road area in an image using this, and divides the detected road area into an image to output it through a screen.

This will be described in more detail.

)는 수학식 12를 이용하여 획득될 수 있다. 정밀 도로 영역 맵을 도출한 결과가 도 9에 도시되어 있다. According to an embodiment of the present invention, the precision road area map (

) can be obtained using Equation 12. The result of deriving the precision road area map is shown in FIG. 9 .

이며, here,

is,

이고,

이다.

ego,

to be.

는 편광 각도 제약을 나타내고,

는 편광 정도 제약을 나타내고,

는 에지 제약을 나타낸다. 또한,

은 편광 정도의 에지 정보를 나타내고,

는 편광각도의 에지 정보를 나타내고,

는 강도 영상의(즉, 제1 스토크스 파라미터 영상)의 에지 정보를 나타내며,

및

는 가중치를 나타낸다. 또한,

및

는 각각 제약 조건의 가중치를 나타내며,

와

는 각각 개략적인 도로 영역 맵(

)에서 도려 영역의 각도와 편광도의 통계에 의해 획득되는 도로 영역에서의 편광 각도(AoP)와 편광 정도(DoP)의 도미넌트 값을 나타낸다. Also,

represents the polarization angle constraint,

represents the degree of polarization constraint,

represents the edge constraint. Also,

represents the edge information of the degree of polarization,

represents the edge information of the polarization angle,

represents the edge information of the intensity image (that is, the first Stokes parameter image),

and

represents the weight. Also,

and

represents the weight of each constraint,

Wow

are each a schematic road area map (

) represents the dominant values of the polarization angle (AoP) and polarization degree (DoP) in the road area obtained by statistics of the angle and polarization degree of the cut-out area.

와

보다 큰 값을 적용한다. 편광 정도 영상, 편광 각도 영상 및 강도 영상(제1 스토크스 파라미터 영상)의 에지는 도로 경계 정보를 제공하므로 강한 가장자리에 더 많은 퍼니시먼트를 제공하여 도로를 다른 영역과 분리하는데 이용한다. For the constraint of the polarization angle and polarization degree image, the shadow or wet area of the road has a smaller polarization angle and polarization degree value.

Wow

A larger value is applied. The edges of the polarization degree image, the polarization angle image, and the intensity image (first Stokes parameter image) provide road boundary information, so more furnishings are provided to the strong edge and used to separate the road from other areas.

)의 감지된 총 영역의 2% 이하보다 작은 영역(

, K(i)에 속하는 픽셀의 수)들은 제거한다. 이후, 도로 영역 검출 장치(100)는 가장 큰 영역을 도로로 인식하여 도로 영역의 평균 편광 정도 및 강도 값을 계산할 수 있다. 평균 편광 정도 또는 강도값이 기 설정된 임계값 이상인 조각들을 제거할 수 있다. 이후, 도로 영역 검출 장치(100)는 작은 구멍을 메워서 정밀 도로 영역(

)을 획득할 수 있다. 최종 획득된 정밀 도로 영역 검출 결과는 도 10에 도시된 바와 같다. In the precision road area map, pixels are binarized by reference values, and m separate areas can be obtained. The road area detection apparatus 100 detects the entire detected area (

) less than 2% of the total detected area (

, the number of pixels belonging to K(i)) are removed. Thereafter, the road area detecting apparatus 100 may calculate the average polarization degree and intensity value of the road area by recognizing the largest area as a road. Fragments having an average polarization degree or intensity value greater than or equal to a preset threshold may be removed. Thereafter, the road area detection device 100 fills in a small hole to provide a precision road area (

) can be obtained. The final obtained precision road area detection result is shown in FIG. 10 .

11 is a block diagram schematically illustrating an internal configuration of an apparatus for detecting a road area according to an embodiment of the present invention.

Referring to FIG. 11 , an apparatus 100 for detecting a road area according to an embodiment of the present invention includes a DoFP image sensor 1110 , a preprocessor 1115 , a calculation unit 1120 , a road area detection unit 1125 , and a memory ( 1130 ) and a processor 1135 .

In FIG. 11 , it is assumed that the road area detection apparatus 100 includes the DoFP image sensor 1110 , but the description is focused thereon. However, the road area detection apparatus 100 does not include the DoFP image sensor 1110 and does not include the DoFP image sensor 1110 . It goes without saying that an input image photographed through the image sensor 1110 may be acquired.

The DoFP image sensor 1110 is a means for generating an input image by photographing a target.

The preprocessor 1115 is a means for preprocessing the input image captured by the DoFP image sensor 1110 to reconstruct four polarized images having different angles.

For example, the preprocessor 1115 may reconstruct polarized images having different angles from each other through a polarization demosaicking (PDM) process. The preprocessor 1115 may reconstruct, for example, 0 degree, 45 degree, 90 degree, and 135 degree polarization images by preprocessing the input image.

The calculator 1120 is a means for calculating a polarization degree and a polarization angle using four polarized images, respectively.

In more detail, the calculator 1120 derives the first Stokes parameter for the total intensity of light using the 0 degree to 135 degree polarization image, and uses the 0 degree and 90 degree polarization images to calculate the amount of linear polarization. A second Stokes parameter is derived, and a third Stokes parameter with respect to the amount of linear polarization may be derived using the 45 degree and 135 degree polarization images.

Next, a polarization degree and a polarization angle may be derived using the first Stokes parameter to the third Stokes parameter, respectively. Since this is the same as that described above, the overlapping description will be omitted.

The road area detection unit 1125 may detect and divide the road area using the degree of polarization and the polarization angle. In more detail, the road area detection unit 1125 may generate a precise road area map by combining the polarization degree, the polarization angle, and the edge information. Next, the road area detection unit 1125 may detect and segment the road area in the input image by using the precise road area map. Since this is the same as that described with reference to FIGS. 1 to 10 , a redundant description will be omitted.

The memory 1130 is a means for storing various instructions (program codes) necessary for performing the method for detecting a road area using an infrared polarization angle according to an embodiment of the present invention.

The processor 1135 includes internal components (eg, the DoFP image sensor 1110 , the preprocessor 1115 , the calculator 1120 , the road It is a means for controlling the area detection unit 1125, the memory 1130, etc.).

The apparatus and method according to an embodiment of the present invention may be implemented in the form of program instructions that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the computer readable medium may be specially designed and configured for the present invention, or may be known and available to those skilled in the art of computer software. Examples of the computer-readable recording medium include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic such as floppy disks. - Includes magneto-optical media and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like.

The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

So far, the present invention has been looked at focusing on the embodiments thereof. Those of ordinary skill in the art to which the present invention pertains will understand that the present invention can be implemented in a modified form without departing from the essential characteristics of the present invention. Therefore, the disclosed embodiments should be considered in an illustrative rather than a restrictive sense. The scope of the present invention is indicated in the claims rather than the foregoing description, and all differences within the scope equivalent thereto should be construed as being included in the present invention.

100: road area detection device
1110: DoFP image sensor
1115: preprocessor
1120: calculator
1125: road area detection unit
1130: memory
1135: Processor

Claims (14)

(a) generating four polarized images having different angles by preprocessing the input image;
(b) calculating a degree of polarization and a polarization angle using the four polarized images; and
(c) generating a precise road area map by combining the polarization degree, the polarization angle, and edge information, and detecting and dividing the road area in the input image using the precise road area map. Way.
According to claim 1,
The input image is an image acquired through a Division of Focal Plane (DoFP) image sensor,
The step (a) is,
and generating 0 degree, 45 degree, 90 degree and 135 degree polarized images by using the input image through a polarization demosaicking (PDM) process.
3. The method of claim 2,
Before step (b),
deriving a first Stokes parameter for the total intensity of light using the 0 degree, 45 degree, 90 degree and 135 degree polarization images;
deriving a second Stokes parameter for an amount of linearly polarized light using the 0 degree and 90 degree polarization images;
and deriving a third Stokes parameter for an amount of linearly polarized light using the 45 degree and 135 degree polarized images.
4. The method of claim 3,
The step (b) is,
deriving a degree of polarization using the first Stokes parameter to the third Stokes parameter; and
and deriving a polarization angle using the second Stokes parameter and the third Stokes parameter.
According to claim 1,
Step (c) is,
obtaining a schematic road area map using the zero distribution constraint of the polarization angle, and deriving a horizon using the rough road area map;
determining a region of interest using the horizon, and generating a region map with precision by combining the polarization degree, the polarization angle, and edge information to the region of interest; and
and dividing the road area from the input image by using the precision road area map.
6. The method of claim 5,
Before deriving the horizon,
and removing noise by performing a morphological open operation on the schematic road area map.
A computer-readable recording medium product on which a program code for performing the method according to any one of claims 1 to 6 is recorded.
a preprocessor for preprocessing the input image to generate four polarized images having different angles;
a calculation unit for calculating a polarization degree and a polarization angle using the four polarized images; and
and a road area detection unit generating a precise road area map by combining the polarization degree, the polarization angle, and edge information, and detecting and dividing the road area in the input image using the precise road area map. .
9. The method of claim 8
The apparatus for detecting a road area further comprising a Division of Focal Plane (DoFP) image sensor configured to acquire the input image.
9. The method of claim 8
The preprocessor is
and generating 0 degree, 45 degree, 90 degree and 135 degree polarized images by using the input image through a polarization demosaicking (PDM) process.
11. The method of claim 10,
The calculator is
Using the 0 degree, 45 degree, 90 degree and 135 degree polarization images, a first Stokes parameter for the total intensity of light is derived, and the first Stokes parameter for the amount of linear polarization using the 0 degree and 90 degree polarization images A road area detection apparatus, comprising deriving 2 Stokes parameters, and deriving a third Stokes parameter with respect to the amount of linear polarization using the 45 and 135 degree polarized images.
12. The method of claim 11,
The calculator is
Road area detection, characterized in that the degree of polarization is derived using the first Stokes parameter to the third Stokes parameter, and the polarization angle is derived using the second Stokes parameter and the third Stokes parameter. Device.
9. The method of claim 8,
The road area detection unit,
obtaining a schematic road area map using the zero distribution constraint of the polarization angle, deriving a horizon using the schematic road area map, determining a region of interest using the horizon, and adding the A road area detection device, characterized in that the polarization degree, the polarization angle, and edge information are combined to generate a precision road area map, and the road area is detected from the input image by using the precision road area map and then divided.
14. The method of claim 13,
The road area detection unit,
and removing noise by performing a morphological open operation on the schematic road area map.

Images