KR20190086334A - Method and apparatus for tracking led in captured image based on probability - Google Patents

Abstract

Disclosed are a method and apparatus for tracking a light emitting diode (LED) in a captured image based on probability. According to the present invention, the method comprises the steps of: configuring pixel information corresponding to a current frame based on an optical flow and a pixel intensity distribution of all pixels in the current frame among a plurality of frames included in a captured image; calculating a conditional probability that an i^th pixel belongs to an LED when the i^th pixel in the current frame corresponds to i^th pixel information in the pixel information; and determining an LED pixel belonging to the LED among all the pixels of the current frame based on the calculated conditional probability.

Description

≪ Desc / Clms Page number 1 > METHOD AND APPARATUS FOR TRACKING LED IN CAPTURED IMAGE BASED ON PROBABILITY < RTI ID = 0.0 >

The following embodiments relate to a method for tracking an LED on a probability basis in a captured image.

Vehicular visible light communication (V2LC) is a promising application for visible light communication (VLC) because it is equipped with high power front and rear LED lights that can be used to transmit data to the vehicle. However, there are few practical applications of V2LC due to the existing problems when receiving LED optical signals. A receiver in a VLC system typically consists of a photodiode (PD) or an image sensor. PD, however, is known to be sensitive to ambient light and is not suitable for outdoor applications such as V2LC. The image sensor can handle ambient light well, including sunlight, but there is a problem when using an image sensor for V2LC. For example, the frame rate of a camera is generally much less than that required in terms of data rate. This problem is likely to be solved as more and more high-speed cameras are introduced into the market at lower prices. However, even when shooting a series of LED images using a high-speed camera, it is not easy to track the position of the LEDs in each image. This is because it is necessary to track the position of the LED quickly and accurately. There may also be difficulties associated with vehicle movement.

Other methods of using image processing for vehicle detection include recognizing the entire vehicle in consideration of various vehicle characteristics such as color, symmetry structure, shadows, corners, edges, and various types of textures. This approach can be more complex than required for simpler problems of detecting the vehicle's LED light. In addition, an application according to this method may have many frames to be processed per second, and therefore it may be difficult to achieve a sufficient processing speed.

In general, LEDs are tracked according to the intensity of the image. However, when the vehicle is moving, motion blur may appear in their images, which may make it difficult to detect the LED using only pixel intensity. More specifically, if the LED image is blurred due to vehicle motion, all the pixels belonging to the LED and the background around the LED are blurred together and have almost the same pixel intensity. Since there are more background pixels than LED pixels, background pixels are more likely to be misjudged as LED pixels. Therefore, as the vehicle speed increases, more frames may be missed. For example, there is an algorithm that tracks LED arrays in a driving situation. This algorithm analyzes the pattern of the LED array. However, the mechanism for sensing a single LED in an LED array depends only on the pixel strength of all the frames.

According to one embodiment, a method of tracking an LED in a shot image is based on an optical flow and a pixel intensity distribution on all pixels in a current frame among a plurality of frames included in the shot image Constructing pixel information corresponding to the current frame; Calculating a conditional probability that the i-th pixel belongs to an LED when the i-th pixel in the current frame corresponds to i-th pixel information in the pixel information; And determining an LED pixel belonging to an LED among all the pixels of the current frame based on the calculated conditional probability.

According to one embodiment, an apparatus for tracking an LED in a captured image includes a photographing unit for photographing an LED to generate a photographing image; And constructing pixel information corresponding to the current frame based on an optical flow and a pixel intensity distribution of all pixels in a current frame among a plurality of frames included in the captured image, Calculating a conditional probability that the i-th pixel belongs to the LED when the i-th pixel in the current frame corresponds to the i-th pixel information in the pixel information, and calculating, based on the calculated conditional probability, And an LED tracing unit for determining the LED pixels to which the LEDs belong.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a view of a vehicle visible light communication system according to one embodiment.
2 illustrates an operation of an LED tracking unit according to an embodiment;
3 is a diagram illustrating a process of generating an optical flow according to an exemplary embodiment;
4 is a diagram of an effective motion vector map according to one embodiment.
5 illustrates a concept of probability based LED tracking in accordance with one embodiment.
6 is an operational flow diagram illustrating a method of tracking an LED in a captured image according to one embodiment.

It is to be understood that the specific structural or functional descriptions of embodiments of the present invention disclosed herein are presented for the purpose of describing embodiments only in accordance with the concepts of the present invention, May be embodied in various forms and are not limited to the embodiments described herein.

Embodiments in accordance with the concepts of the present invention are capable of various modifications and may take various forms, so that the embodiments are illustrated in the drawings and described in detail herein. However, it is not intended to limit the embodiments according to the concepts of the present invention to the specific disclosure forms, but includes changes, equivalents, or alternatives falling within the spirit and scope of the present invention.

The terms first, second, or the like may be used to describe various elements, but the elements should not be limited by the terms. The terms may be named for the purpose of distinguishing one element from another, for example without departing from the scope of the right according to the concept of the present invention, the first element being referred to as the second element, Similarly, the second component may also be referred to as the first component.

It is to be understood that 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, . 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. Expressions that describe the relationship between components, for example, "between" and "immediately" or "directly adjacent to" should be interpreted as well.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The singular expressions include plural expressions unless the context clearly dictates otherwise. In this specification, the terms " comprises " or " having ", and the like, are used to specify one or more of the features, numbers, steps, operations, But do not preclude the presence or addition of steps, operations, elements, parts, or combinations thereof.

Unless defined otherwise, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Terms such as those defined in commonly used dictionaries are to be interpreted as having a meaning consistent with the meaning of the context in the relevant art and, unless explicitly defined herein, are to be interpreted as ideal or overly formal Do not.

Hereinafter, embodiments will be described in detail with reference to the accompanying drawings. However, the scope of the patent application is not limited or limited by these embodiments. Like reference symbols in the drawings denote like elements.

1 is a view illustrating a vehicle visible light communication system according to an exemplary embodiment of the present invention.

Referring to FIG. 1, an apparatus 110 for tracking an LED (Light Emitting Diode) in a captured image includes a photographing unit 111, an LED tracking unit 112, a position estimating unit 113, a visible light communication unit 114, (115).

The photographing unit 111 can operate as a receiver in a vehicle visible light communication system. The photographing unit 111 may include at least one camera each including an image sensor, and according to an embodiment, at least one camera may include a high-speed camera. The image sensor may comprise a complementary metal-oxide semiconductor (CMOS) sensor.

The photographing unit 111 can photograph an LED in the surrounding vehicle 120 or the streetlight 130 to generate a photographed image. The shot image may comprise a series of images, or frames, comprising an LED. The device 110 may estimate the position of the vehicle 100 based on the LED in the shot image, or may perform visible light communication. For example, the peripheral vehicle 120 and the streetlight 130 can transmit a visible light signal including data through the LED, and the device 110 photographs the light through the photographing unit 111 to receive data in the visible light signal can do. The apparatus 110 can estimate the position of the vehicle 100 by constructing a geometric relationship between the image coordinates of the LED in the shot image and the world coordinates of the streetlight 130. [ The world coordinates of the streetlight 130 may be determined based on the visible light signal transmitted by the streetlight 130.

For this operation, it is necessary to detect and continuously track the LEDs in the shot image. For example, when the position of the LED is detected in the captured image, data included in the light can be extracted from the captured image. In general, LEDs can be detected according to pixel intensity. However, when the vehicle is traveling, there is movement in the LED image, which may make it difficult to detect the LED. At high speeds, some frames may become so high that they may not be able to detect the LEDs and extract the information contained in these frames.

According to the embodiment, the LED tracker 112 tracks the LED using not only the pixel intensity but also the optical flow of the LED and the statistical information obtained from the previous frames. Based on this information, the conditional probability that the pixel belongs to the LED is calculated, and the position of the LED can be determined based on this probability.

When the LED is detected in the shot image, position estimation or visible light communication based thereon can be performed. For example, the position estimating unit 113 may receive the image coordinates of the LEDs in the shot image from the LED tracking unit 112, thereby estimating the position of the vehicle 100. The visible light communication unit 114 receives the image coordinates of the LED in the photographed image from the LED tracking unit 112 and extracts the visible light signal transmitted from the surrounding vehicle 120 or the streetlight 130 in the photographed image . The visible light communication unit 114 can transmit the visible light signal to the surrounding vehicle 120 through the LED 115.

Since vehicles 100 and 120 are equipped with high-power LEDs on the front and rear surfaces that can be used to transmit data, vehicular visible light communication (V2LC) is a type of visible light communication (VLC) It is a promising application. However, there are few practical applications of V2LC due to the existing problems when receiving LED optical signals. The photographing unit 111 corresponding to the receiver of the VLC system generally comprises a photodiode (PD) or an image sensor. PD, however, is known to be sensitive to ambient light and is not suitable for outdoor applications such as V2LC.

The image sensor can handle ambient light well, including sunlight, but there is a problem when using an image sensor for V2LC. For example, the frame rate of a camera is generally much less than that required in terms of data rate. This problem is likely to be solved as more and more high-speed cameras are introduced into the market at lower prices. However, even when shooting a series of LED images using a high-speed camera, it is not easy to track the position of the LEDs in each image. This is because it is necessary to track the position of the LED quickly and accurately. There may also be difficulties associated with vehicle movement.

Other methods of using image processing for vehicle detection include recognizing the entire vehicle in consideration of various vehicle characteristics such as color, symmetry structure, shadows, corners, edges, and various types of textures. This approach can be more complex than required for simpler problems of detecting the vehicle's LED light. In addition, an application according to this method may have many frames to be processed per second, and therefore it may be difficult to achieve a sufficient processing speed.

In general, LEDs are tracked according to the intensity of the image. However, when the vehicle 100 receiving the visible light signal or the peripheral vehicle 120 transmitting the visible light signal moves, a motion blur may appear in the photographed image generated by the photographing unit 111, Thereby making it difficult to detect the LED. More specifically, if the LED image is blurred due to vehicle motion, all the pixels belonging to the LED and the background around the LED are blurred together and have almost the same pixel intensity. Since there are more background pixels than LED pixels, background pixels are more likely to be misjudged as LED pixels. Therefore, as the vehicle speed increases, more frames may be missed. For example, there is an algorithm that tracks LED arrays in a driving situation. This algorithm analyzes the pattern of the LED array. However, the mechanism for sensing a single LED in an LED array depends only on the pixel strength of all the frames.

The embodiments below propose a probability-based model for tracking the position of LEDs in a series of images. The proposed algorithm computes not only the pixel intensity but also the optical flow of the LEDs generated in the continuous frame and the statistical information of the previous frame in which the LED is still clearly visible, to calculate the probability that the pixel belongs to the LED or background. Then, based on this probability, the LED pixels in the image are determined. If the LED is blurred due to vehicle motion, the pixel intensity of the LED pixel and surrounding background pixel may become similar or equal. In this case, however, the optical flow of the LED can still be configured. Since the statistical information of the previous frame is also taken into account to calculate the probability, the probability of detecting the LED pixel is still higher than the probability of detecting the nearby background. In other words, even if the LED is blurred, the LED can be accurately detected.

2 is a diagram illustrating an operation of the LED tracking unit according to an exemplary embodiment.

Referring to FIG. 2, the LED tracker 210 may generate LED pixel information in the shot image based on the shot image. The LED pixel information may include position information, or coordinates, of the LED pixel.

The LED tracking unit 210 can track the LED pixels in the captured image based on the optical flow and the pixel intensity distribution. More specifically, the LED tracking unit 210 constructs pixel information corresponding to the current frame based on the optical flow and the pixel intensity distribution relating to all pixels in the current frame among a plurality of frames included in the shot image, Calculate the conditional probability that the i-th pixel belongs to the LED when the ith i-th pixel corresponds to the i-th pixel information in the pixel information, and determine an LED pixel belonging to the LED among all pixels of the current frame based on the calculated conditional probability have.

Depending on the pixel intensity, the background pixel and the LED pixel can be distinguished. The pixel intensity may be compared to, for example, a threshold value to determine whether the pixel is an LED pixel or a background pixel. According to an embodiment, the pixel intensity is used as one of the parameters for calculating the conditional probability that the pixel is an LED pixel.

로 나타낼 수 있다. 일반적인 상황에서,

은 이미지 센서의 중심에서 픽셀 강도가 가장 높고 이미지 센서의 외곽으로 갈수록 점차 감소하는 이미지 센서에 관한 가우스 분포를 갖는 것으로 가정될 수 있다.the intensity of the i-th pixel in the n-th frame

. Under normal circumstances,

Can be assumed to have a Gaussian distribution with respect to the image sensor having the highest pixel intensity at the center of the image sensor and gradually decreasing toward the outer edge of the image sensor.

The optical flow will be described with reference to Figs. 3 and 4 below.

3 is a diagram illustrating a process of generating an optical flow according to an exemplary embodiment of the present invention.

Because vehicle motion can be reflected in the optical signal taken by the image sensor, an optical flow can be considered in detecting the LED. An optical flow is a pattern that represents the apparent motion of an object, a surface, and an edge of a visual scene created by the relative movement between the camera and the scene. Thus, an instant optical flow map can be generated that shows the direction of motion using two successive images. Vehicle motion can be measured in an image taking into account two factors: the direction of motion and its magnitude. Assuming that the time interval between two consecutive frames is very short, the pixel intensities of the same LEDs in two frames are therefore approximately the same.

3, an optical flow may be generated based on a comparison between a pixel 301 in the current frame 320 and a corresponding pixel 302 in the reference image 310. The position of the pixel 301 in the current frame 320 may be compared to the corresponding pixel 302 in the reference frame 310 to calculate the direction and magnitude of that pixel 301's motion. For example, the reference frame 310 may be a previous frame of the current frame, a frame in which relatively few motion blur is included in the previous frames, or a frame in which the LED is relatively accurately detected. In Fig. 3, the motion direction is represented by the angle [theta], and the motion size is represented by r.

Once the motion vector for all the pixels of the image is calculated, an effective motion vector map is defined that is expected to contain pixels belonging to the vehicle. It is assumed that the pixels belonging to the background are stable or only slightly changed between consecutive frames. The effective motion vector map is then generated by comparing the combination of r and θ with a given threshold. If the combination does not exceed the threshold, this pixel is considered stable. The effective motion vector for every single pixel can be calculated according to Equation (1), and an optical flow or an effective motion vector map can be constructed based on the calculated effective motion vectors.

의 메트릭이고, 하위 인덱스 i는 이미지의 i번째 픽셀을 나타내며, 상위 인덱스 n은 일련의 이미지들 중 n번째 이미지를 나타낸다. 여기서 이미지는 프레임을 의미할 수 있다.Where δ is

The lower index i represents the i-th pixel of the image, and the upper index n represents the n-th image of the series of images. Here, the image may mean a frame.

4 is a diagram illustrating an effective motion vector map according to an embodiment. The effective motion vector map 400 may be generated in two consecutive frames where the dots in the valid motion vector map 400 may correspond to background pixels and the arrows may correspond to pixels belonging to the vehicle.

5 is a diagram illustrating the concept of probability based LED tracking according to an embodiment.

The optical flow and the pixel intensity distribution are independent of each other. To locate these LEDs using these, you need to combine them into a single framework that uses both features as inputs. The Bayesian framework proposed in the embodiments below uses the optical flow and pixel intensity distribution as input to calculate the probability that the pixel belongs to the background or LED.

Given the optical flow and intensity for every pixel in the image, the probability that the pixel belongs to the LED is calculated. This is a conditional probability that can be computed using the Bayes' rule.

및

가 있다.

은 픽셀이 LED에 속하는 것을 나타내고

는 픽셀이 LED에 속하지 않는 것을 나타낸다. 확률 함수

는 이벤트의 전체 집합을 [0, 1]로 매핑한다. n번째 프레임에서 i번째 픽셀의 정보를

라고 정의한다. 이 정보는 위에서 언급 한 두 가지 특징, 즉 운동 방향 θ 및 그 크기 r을 포함하는 옵티컬 플로우, 및 픽셀 강도 분포 f를 포함한다. n번째 프레임에서, 이벤트 w1 또는 w2가 발생할 조건부 확률은 주어진 n번째 프레임까지의 일련의 이전 프레임들로부터의 정보를 고려하여 베이즈의 규칙에 따라 수학식 2와 같이 계산된다.First, a complete event set is defined to determine where the pixel belongs. The pixel has two states

And

.

Indicates that the pixel belongs to the LED

Indicates that the pixel does not belong to the LED. Probability function

Maps the entire set of events to [0, 1]. the information of the i-th pixel in the n-th frame

. This information includes the two features mentioned above, the optical flow including the direction of motion [theta] and its magnitude r, and the pixel intensity distribution f. In the nth frame, the conditional probability of occurrence of the event w1 or w2 is calculated according to the Bayes rule, taking into account information from a series of previous frames up to the given nth frame, as shown in equation (2).

이다.here,

to be.

를 계산하려면 두 종류의 확률

및

를 계산해야 한다. θ, r 및 f가 통계적으로 독립적이라고 보면 확률

는 수학식 3과 같이 나타낼 수 있다.

To calculate the two types of probability

And

. If θ, r, and f are statistically independent,

Can be expressed by Equation (3).

는 LED 픽셀 및 배경 픽셀을 포함하는 픽셀 상태

및 픽셀 강도 분포

에 기초하며, 아래에서 제1 확률 성분으로 지칭될 수 있다.

는 옵티컬 플로우의 크기

및 픽셀 강도 분포

에 기초하며, 아래에서 제2 확률 성분으로 지칭될 수 있다.

는 옵티컬 플로우의 방향

및 픽셀 강도 분포

에 기초하며, 아래에서 제3 확률 성분으로 지칭될 수 있다.Where j = 1, 2. In Equation 3,

RTI ID = 0.0 > LED < / RTI > pixels and background pixels

And pixel intensity distribution

And may be referred to below as a first probability component.

The size of the optical flow

And pixel intensity distribution

And may be referred to below as a second probability component.

The direction of the optical flow

And pixel intensity distribution

And may be referred to below as a third probability component.

Assuming that θ, r, and f are normal distributions, the three likelihoods of Equation (3) can be expressed as Equation (4) according to the Gaussian function.

,

,

,

,

,

,

는 이전 프레임들을 통해 계산 및 업데이트된 옵티컬 플로우의 크기 r, 방향 θ, 및 픽셀 강도 분포 f 각각의 평균 및 분산을 나타낸다. 다시 말해, 제1 확률 성분

, 제2 확률 성분

, 및 제3 확률 성분

각각은 촬영 이미지 내에서 현재 프레임까지의 이전 프레임들에 관해 지속적으로 갱신된 옵티컬 플로우의 크기 r, 방향 θ 및 픽셀 강도 분포 f에 기초하여 계산될 수 있다. 이러한 연산을 위해 옵티컬 플로우의 크기 r, 방향 θ은 촬영 이미지의 각 프레임에 관해 통계 정보로서 지속적으로 계산되고, 업데이트될 수 있다.here

,

,

,

,

,

,

Represents the mean and variance of the magnitude r, the direction [theta], and the pixel intensity distribution f of the optical flow computed and updated over the previous frames, respectively. In other words, the first probability component

, The second probability component

, And the third probability component

Each can be computed based on the magnitude r, the direction [theta], and the pixel intensity distribution f of the optical flow continuously updated with respect to previous frames up to the current frame in the shot image. For such an operation, the magnitude r of the optical flow and the direction? Can be continuously calculated and updated as statistical information about each frame of the photographed image.

라 하면, 픽셀이 LED에 속할 확률

는 수학식 5와 같이 나타낼 수 있다.a set of pixels belonging to the LED in the (n-1)

, The probability that a pixel belongs to an LED

Can be expressed by Equation (5).

는 이전 프레임의 LED 픽셀의 수를 나타낸다.here

Represents the number of LED pixels of the previous frame.

The probability that the pixel belongs to the background can be expressed by Equation (6).

The set of LED pixels in the current frame can be expressed as Equation (7).

Equation (7) can be expressed by Equation (8).

The decision rule of Equation (3) is applied to all the pixels in the image to find the LED area. The exact position of the LED is determined by selecting the pixel most likely to belong to the LED according to equation (9).

The advantage of the algorithm according to the embodiment can be found from the equation (3). When the LED image is blurred due to vehicle motion, the pixels belonging to the LED and the pixels belonging to the surrounding background are mixed. Thus, pixel intensity can no longer be a crucial factor in tracking the position of the LED. This is a problem with previous tracking algorithms.

은 LED에 흐림 현상이 발생하면 서로 유사한 값을 갖는다. 그러나, 다른 두 개의 확률

및

는 여전히 LED 픽셀에 대해 더 높은 값을 갖는다. 따라서 LED가 흐려져도 LED가 여전히 검출될 수 있다. 다시 말해, 특정 픽셀이 LED에 속하고, 상기 픽셀에 흐림 현상이 발생할 때,

은 상기 픽셀이 배경에 속할 때와 유사한 값을 가질 수 있고,

및

는 상기 픽셀이 배경에 속할 때보다 큰 값을 가질 수 있다.However, the probability that the LED tracking is based on in the proposed method is calculated from the three probabilities as in Equation (3). The probability of LED pixels and surrounding background pixels

Have similar values when blurring occurs in LEDs. However, the other two probabilities

And

Lt; / RTI > still has a higher value for the LED pixel. Therefore, even if the LED is dimmed, the LED can still be detected. In other words, when a particular pixel belongs to the LED and the pixel is blurred,

May have a value similar to when the pixel belongs to the background,

And

May have a larger value than when the pixel belongs to the background.

6 is an operational flow diagram illustrating a method for tracking an LED in a captured image according to an embodiment. Referring to FIG. 6, in step 610, the device constructs pixel information corresponding to the current frame based on the optical flow and pixel intensity distribution for all pixels in the current frame among a plurality of frames included in the captured image. In step 620, the device calculates a conditional probability that the i-th pixel belongs to the LED when the i-th pixel in the current frame corresponds to the i-th pixel information in the pixel information. In step 630, the device determines an LED pixel belonging to the LED among all the pixels of the current frame based on the calculated conditional probability. The apparatus performs visible light communication based on the determined LED pixel in step 640 and estimates the position of the vehicle based on the determined LED pixel in step 650. In addition, the above description can be applied to the method of tracking the LED in the captured image, and a detailed description will be omitted.

The apparatus described above may be implemented as a hardware component, a software component, and / or a combination of hardware components and software components. For example, the apparatus and components described in the embodiments may be implemented as a computer-readable recording medium, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate array (FPGA) , A programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions. The processing device may execute an operating system (OS) and one or more software applications running on the operating system. The processing device may also access, store, manipulate, process, and generate data in response to execution of the software. For ease of understanding, the processing apparatus may be described as being used singly, but those skilled in the art will recognize that the processing apparatus may have a plurality of processing elements and / As shown in FIG. For example, the processing unit may comprise a plurality of processors or one processor and one controller. Other processing configurations are also possible, such as a parallel processor.

The software may include a computer program, code, instructions, or a combination of one or more of the foregoing, and may be configured to configure the processing device to operate as desired or to process it collectively or collectively Device can be commanded. The software and / or data may be in the form of any type of machine, component, physical device, virtual equipment, computer storage media, or device , Or may be permanently or temporarily embodied in a transmitted signal wave. The software may be distributed over a networked computer system and stored or executed in a distributed manner. The software and data may be stored on one or more computer readable recording media.

The method according to an embodiment may be implemented in the form of a program command 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, and the like, alone or in combination. The program instructions to be recorded on the medium may be those specially designed and configured for the embodiments or may be available to those skilled in the art of computer software. Examples of computer-readable media include magnetic media such as hard disks, floppy disks, and magnetic tape; optical media such as CD-ROMs and DVDs; magnetic media such as floppy disks; Magneto-optical media, and hardware devices specifically configured to store and execute program instructions such as ROM, RAM, flash memory, and the like. Examples of program instructions include machine language code such as those produced by a compiler, as well as high-level language code 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 embodiments, and vice versa.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments. For example, it is to be understood that the techniques described may be performed in a different order than the described methods, and / or that components of the described systems, structures, devices, circuits, Lt; / RTI > or equivalents, even if it is replaced or replaced.

Therefore, other implementations, other embodiments, and equivalents to the claims are also within the scope of the following claims.

Claims (10)

A method for tracking an LED in a shot image,
Constructing pixel information corresponding to the current frame based on an optical flow and a pixel intensity distribution of all pixels in a current frame among a plurality of frames included in the captured image;
Calculating a conditional probability that the i-th pixel belongs to the LED when the i-th pixel in the current frame corresponds to the i-th pixel information in the pixel information; And
Determining an LED pixel belonging to an LED of all the pixels of the current frame based on the calculated conditional probability
≪ / RTI > The method according to claim 1,
The optical flow
Wherein the determination is based on the i-th pixel of the current frame and the corresponding pixel of the i-th pixel in the reference frame included in the captured image. The method according to claim 1,
The optical flow
The magnitude and direction of the motion vector based on the i-th pixel of the current frame and the corresponding pixel of the i-th pixel in the reference frame included in the captured image. The method according to claim 1,
The conditional probability
A first probability component based on the pixel state including the LED pixel and the background pixel and the pixel intensity distribution;
A second probability component based on the magnitude of the optical flow and the pixel intensity distribution; And
And a third probability component based on the direction of the optical flow and the pixel intensity distribution
≪ / RTI > 5. The method of claim 4,
Wherein when the i-th pixel belongs to the LED and the blurring occurs in the i-th pixel, the first probability component has a value similar to that when the i-th pixel belongs to the background, and the second probability component and the 3 probability component has a larger value than when the i-th pixel belongs to the background. 5. The method of claim 4,
Wherein each of the first probability component and the third probability component is calculated based on a continuously updated optical flow and a pixel intensity distribution for previous frames in the captured image to the current frame. The method according to claim 1,
Wherein at least one of visible light communication and vehicle position estimation is performed based on the determined LED pixel. 1. An apparatus for tracking an LED in a shot image,
A photographing unit photographing an LED to generate a photographing image; And
Constructs pixel information corresponding to the current frame based on an optical flow and a pixel intensity distribution of all pixels in a current frame among a plurality of frames included in the captured image, Calculating a conditional probability that the i-th pixel belongs to the LED when the i-th pixel in the frame corresponds to i-th pixel information in the pixel information, and calculating a conditional probability that the i- LED tracing unit for determining LED pixels
/ RTI > 9. The method of claim 8,
The optical flow
Th pixel of the current frame and the corresponding pixel of the i-th pixel in the reference frame contained in the captured image. 9. The method of claim 8,
The conditional probability
A first probability component based on the pixel state including the LED pixel and the background pixel and the pixel intensity distribution;
A second probability component based on the magnitude of the optical flow and the pixel intensity distribution; And
And a third probability component based on the direction of the optical flow and the pixel intensity distribution
/ RTI >

Images