WO2017006451A1 - 灯火器検出装置及び灯火器検出方法 - Google Patents
灯火器検出装置及び灯火器検出方法 Download PDFInfo
- Publication number
- WO2017006451A1 WO2017006451A1 PCT/JP2015/069599 JP2015069599W WO2017006451A1 WO 2017006451 A1 WO2017006451 A1 WO 2017006451A1 JP 2015069599 W JP2015069599 W JP 2015069599W WO 2017006451 A1 WO2017006451 A1 WO 2017006451A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- image
- edge portion
- lamp
- edge
- luminance
- Prior art date
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 51
- 230000001360 synchronised effect Effects 0.000 claims abstract description 55
- 239000000284 extract Substances 0.000 claims abstract description 7
- 238000009499 grossing Methods 0.000 claims description 26
- 238000003708 edge detection Methods 0.000 claims description 16
- 238000003384 imaging method Methods 0.000 claims description 11
- 238000000605 extraction Methods 0.000 claims description 5
- 230000005611 electricity Effects 0.000 abstract 1
- 235000019557 luminance Nutrition 0.000 description 70
- 238000012545 processing Methods 0.000 description 20
- 238000000034 method Methods 0.000 description 18
- 230000008569 process Effects 0.000 description 11
- 230000004044 response Effects 0.000 description 7
- 230000006399 behavior Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 230000007274 generation of a signal involved in cell-cell signaling Effects 0.000 description 6
- 230000008859 change Effects 0.000 description 4
- 230000010365 information processing Effects 0.000 description 3
- 238000004590 computer program Methods 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- 230000001629 suppression Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 239000000725 suspension Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/09—Arrangements for giving variable traffic instructions
- G08G1/095—Traffic lights
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
- G06V20/584—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads of vehicle lights or traffic lights
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
- G06V20/63—Scene text, e.g. street names
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/09—Arrangements for giving variable traffic instructions
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/09—Arrangements for giving variable traffic instructions
- G08G1/0962—Arrangements for giving variable traffic instructions having an indicator mounted inside the vehicle, e.g. giving voice messages
- G08G1/09623—Systems involving the acquisition of information from passive traffic signs by means mounted on the vehicle
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
- G06V20/582—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads of traffic signs
Definitions
- the present invention relates to a lighting device detection apparatus and a lighting device detection method.
- Patent Document 1 a traffic signal detection device that detects a traffic signal from an image captured by a camera is known (see Patent Document 1).
- a signal lamp candidate is detected from an image based on the color and shape of the signal lamp, and it is determined whether the signal lamp candidate blinks at a predetermined cycle.
- the position of the edge portion having a luminance difference equal to or larger than a predetermined value included in the image captured during the movement is also moved.
- a luminance step response occurs in the region where the edge portion has moved, and noise having a wide frequency component is generated.
- noise having a wide frequency component may be erroneously recognized as the lighting device.
- the present invention has been made in view of the above problems, and its purpose is to suppress erroneous detection of a luminance step response as a lighting device even when the edge portion moves, and to stabilize the lighting device. It is an object to provide a lamp detection device and a lamp detection method capable of detection.
- the lighting device detection apparatus detects an edge portion having a luminance difference of a predetermined value or more from a camera image, and adjusts the luminance difference of the edge portion to generate a smooth image. And the synchronous pixel from which a brightness
- the luminance gradient which is the luminance difference between adjacent pixels in the edge portion
- the luminance gradient is smoothed, so that the luminance of the step response of the luminance of the pixel due to the edge portion moving the position on the image
- the noise intensity in the power frequency band is reduced, and it is possible to suppress erroneous detection of noise due to the step response of the luminance of the pixel as a lighting device, and to detect the lighting device stably.
- FIG. 1 is a block diagram showing the overall configuration of the lighting device detection apparatus according to the first embodiment.
- 2A (a) shows a camera image
- FIG. 2A (b) shows an edge portion detected from FIG. 2A (a)
- FIG. 2A (c) adjusts the luminance of the edge portion of FIG. 2A (b).
- generated by this is shown.
- FIG. 2B (a) shows a camera image
- luminance of a part is shown.
- FIG. 3A is a diagram illustrating an example of a smoothing process for adjusting the luminance of the center pixel using the luminance of the surrounding eight pixels.
- FIGS. 3B (a) and 3 (b) are diagrams illustrating an example of smoothing processing for adjusting the luminance of the center pixel using the luminance of the surrounding six pixels, and smoothing processing of the edge portion extending in the vertical direction. is there.
- FIGS. 3C (a) and 3 (b) are diagrams illustrating an example of the smoothing process for adjusting the brightness of the center pixel using the brightness of the surrounding 6 pixels, and smoothing the edge portion extending in the horizontal direction. is there.
- FIG. 4 is a block diagram showing the detailed configuration of the synchronized image generation circuit 24 and the reference signal setting circuit 23.
- FIG. 5 is a flowchart showing an example of a lamp detection method using the lamp detection apparatus of FIG.
- FIG. 6 is a block diagram showing the overall configuration of the lighting device detection apparatus according to the second embodiment.
- FIG. 7 is a flowchart showing an example of a lamp detection method using the lamp detection apparatus of FIG.
- the traffic light detection device is mounted on a vehicle, captures an image of the surroundings of the vehicle and acquires a camera image, an image processing unit 12 that detects a lamp using the camera image acquired by the imaging unit 11, and With a map database.
- the imaging unit 11 is a digital camera equipped with a solid-state imaging device (image sensor), for example, a CCD or a CMOS, and acquires a camera image that can be image-processed.
- image sensor solid-state imaging device
- the imaging unit 11 repeatedly captures, for example, the front of the vehicle at predetermined time intervals, and acquires a plurality of continuous camera images.
- the imaging unit 11 performs imaging a plurality of times during one AC cycle of the system power supply supplied to the lamp.
- the imaging unit 11 can capture a lighting device existing in front of the vehicle in each camera image. That is, the image sensor has at least sensitivity enough to detect the lighting device.
- a “lighting device” is an artificial light source that emits light when supplied with system power, such as a traffic light signal light, a self-luminous road sign, and other electric lights (self-luminous signboard, vending machine, etc. Light emitting part).
- the image processing unit 12 receives the data of the camera image acquired by the imaging unit 11, and detects the type of the lighting device included in the camera image.
- the information on the detected lamp is transferred to another processing arithmetic device (vehicle CPU 16) mounted on the vehicle, including a controller for realizing automatic driving of the vehicle, for example.
- vehicle CPU 16 processing arithmetic device mounted on the vehicle, including a controller for realizing automatic driving of the vehicle, for example.
- the image processing unit 12 includes, for example, a microcontroller including a CPU, a memory, and an input / output unit, and configures a plurality of information processing circuits included in the lamp detection device by executing a computer program installed in advance.
- the image processing unit 12 repeatedly executes a series of information processing cycles (including edge detection processing, edge smoothing processing, and power synchronization processing) for detecting a lighting device from a camera image for each of a plurality of continuous color images.
- the image processing unit 12 may also be used as an ECU used for other control relating to the vehicle.
- the plurality of information processing circuits configured by the image processing unit 12 include an edge detection circuit 21, an edge smoothing circuit 22, a reference signal setting circuit 23, a synchronous image generation circuit 24, and a lamp detection circuit 26. It is.
- the edge detection circuit 21 detects an edge portion where the luminance difference between adjacent pixels is a predetermined value or more from the camera image. For example, the edge detection circuit 21 detects the edge portion 31a shown in FIG. 2A (b) from the camera image shown in FIG. 2A (a). In FIG. 2A (a), (0) and (255) indicate luminance values (8 bits), respectively. (0) indicates black and (255) indicates white. The grid in FIG. 2A (a) shows a pixel. The edge detection circuit 21 detects, as an edge portion 31a, a portion that produces a luminance difference equal to or greater than a predetermined threshold value between adjacent pixels. A portion where the luminances (0) and (255) are adjacent to each other in FIG.
- the edge part 31a shown in FIG. 2A (b) is configured by a total of 16 pixels that are continuous in a rectangular shape of 4 vertical pixels and 4 horizontal pixels.
- the luminance value of the pixel to which the edge portion has moved changes in a step response, so that a luminance value is generated as noise in a wide frequency band.
- the predetermined value which is a predetermined threshold, is set based on the luminance difference of the edge portion where noise that is extracted as a synchronized image occurs.
- Specific edge detection methods are not particularly limited, and existing techniques can be used. For example, a Sobel operator, a Prewitt operator, and a Canny method can be used.
- the edge smoothing circuit 22 adjusts the luminance difference of the edge part 31a in the camera image so as to generate a smooth image. For example, the edge smoothing circuit 22 daringly adjusts the luminance difference of the edge part 31a in the camera image shown in FIG. 2A (a) to generate the smoothed image shown in FIG. 2A (c). At this time, in the camera image of FIG. 2A (b), the brightness difference is daringly adjusted so that the brightness gradient, which is the brightness difference between adjacent pixels in the edge portion 31a, becomes gentle. As an example, the edge smoothing circuit 22 adjusts each luminance of a total of 16 pixels constituting the edge part 31a shown in FIG. 2A (b) by the smoothing process shown below. A small adjustment is made so that the luminance difference of the edge portion 31a does not generate noise that is extracted as a synchronized image.
- the central pixel represented as “0” corresponds to the pixel constituting the edge portion 31a.
- the luminance of the pixel is adjusted to the average value of the luminance of the surrounding eight pixels. This process is performed on each of the 16 pixels constituting the edge portion 31a.
- the luminance of the eight pixels forming the four sides of the edge portion 31a is adjusted to (192), and the luminance of the four pixels forming the four corners of the edge portion 31a is adjusted to (96). Adjusted. In this way, the brightness difference of the camera image of FIG.
- FIG. 2A (a) is adjusted to be small so that the brightness gradient, which is the brightness difference between adjacent pixels in the edge portion 31a, becomes gentle, and FIG. 2A (c) Smooth images are generated.
- a pixel group composed of two vertical pixels and two horizontal pixels that are surrounded by eight pixels whose luminances are adjusted and whose luminances are not adjusted constitutes a region 32a for detecting a lamp.
- FIGS. 3B (a) and 3 (b) show examples of smoothing edge portions extending in the vertical direction
- FIGS. 3C (a) and 3 (b) show examples of smoothing edge portions extending in the horizontal direction.
- smoothing processing may be performed by a different method between the edge portion extending in the vertical direction and the edge portion extending in the horizontal direction in the edge portion 31a in FIG. 2A (b).
- only the edge portion extending in the horizontal direction or the vertical direction may be smoothed according to the moving direction of the image.
- the synchronized image generation circuit 24 (synchronized pixel extraction circuit) extracts a synchronized pixel whose luminance changes in synchronization with the AC cycle of the power supplied to the lamp from the smooth image, and is a synchronized image made up of synchronized pixels. Is generated.
- the reference signal setting circuit 23 sets a reference signal used when extracting a synchronous pixel.
- the synchronous image generation circuit 24 includes an image memory 35, a multiplication circuit 36, and an average processing circuit 37.
- the reference signal setting circuit 23 includes a reference signal generation circuit 38 and a phase determination circuit 39.
- the reference signal generation circuit 38 acquires frequency information of the system power supply (commercial power supply) around the vehicle using the own vehicle position information 14 shown in FIG. , 100 Hz) to generate a reference signal whose intensity varies.
- the own vehicle position information 14 can be obtained from a positioning signal such as a GPS signal, or position information of a landmark reflected in a camera image.
- the multiplication circuit 36 reads out a smooth image stored in the image memory 35 and multiplies the reference signal to extract a synchronous pixel.
- the multiplication circuit 36 performs the multiplication process described above for each of the plurality of smooth images stored in the image memory 35 at the same time.
- the average processing circuit 37 calculates the average value of the multiplication results for each smoothed image by the multiplication circuit 30 and outputs the average value as a synchronized image composed of synchronized pixels.
- the phase determination circuit 39 determines whether the phase of the reference signal is synchronized with the luminance fluctuation phase of the smoothed image, and feeds back the determination result to the reference signal generation circuit 38.
- the reference signal generation circuit 38 receives this feedback and adjusts the phase of the reference signal so that the phase of the reference signal is synchronized with the luminance fluctuation of the smoothed image. Specifically, the phase of the reference signal is adjusted so that the multiplication result of the multiplication circuit 36, that is, the luminance of the synchronized pixel takes the maximum value. By this feedback control, the brightness of the synchronized image can be increased.
- the electric power supplied to the lamp is AC power obtained by full-wave rectification of commercial power.
- the brightness of the lamp that is lit by receiving power supplied from a commercial power supply changes in the same cycle as the cycle of full-wave rectified AC power (for example, 100 Hz). Synchronized pixels whose luminance changes in synchronization with the AC cycle of power supplied to the lamp can be extracted from the smooth image.
- the brightness difference at the edge portion is adjusted to be small so that the brightness gradient, which is the brightness difference between adjacent pixels in the edge portion, becomes gentle, and then the brightness changes in synchronization with the AC cycle of power. Extract pixels.
- the spatial frequency is less likely to be mixed in time-frequency synchronization detection, and noise included in the synchronized image can be reduced.
- the lighting device detection circuit 26 determines a synchronized pixel having a luminance higher than a predetermined luminance threshold as a lighting device from the synchronized image, and detects the position of the lighting device on the synchronized image. Then, the type of the lighting device is detected using at least one of the synchronization image, the edge image, and the camera image. For example, the lighting device detection circuit 26 uses pattern matching to determine whether the shape of the lighting device is similar to a self-luminous road sign. Then, the lighting device detection circuit 26 determines whether or not the hue of the position of the camera image corresponding to the position of the lighting device is similar to the lighting color of the traffic signal. With these determination functions, it is possible to determine whether the lighting device is a traffic light signal light, a self-luminous road sign, or another electric light.
- step S01 the imaging unit 11 continuously acquires camera images.
- the acquired plurality of camera images are stored in the memory.
- the process proceeds to step S03.
- the edge detection circuit 21 uses, for example, a Sobel operator or a Prewitt operator to select an edge portion 31a (FIG. 2A (b)) having a luminance difference of a predetermined value or more from the camera image (FIG. 2A (a)). To detect.
- the edge smoothing circuit 22 gradually reduces the luminance gradient, which is the luminance difference between adjacent pixels in the edge portion 31a.
- a smooth image (FIG. 2A (c)) is generated by adjusting the luminance difference of the camera image to be small.
- step S07 the synchronous image generation circuit 24 extracts, from the smooth image, a synchronous pixel whose luminance changes in synchronization with the AC cycle of the power supplied to the lamp, and generates a synchronous image including the synchronous pixels. Generate.
- the phase determination circuit 39 refers to the synchronized image and determines whether the phase of the reference signal is synchronized with the luminance fluctuation of the smoothed image. For example, when the average luminance value of the entire synchronized image is less than a predetermined value, it is determined that the synchronization is not performed (NO in step S09), and the phase determination circuit 39 feeds back the determination result to the reference signal generation circuit 38.
- the reference signal generation circuit 38 receives this feedback and adjusts the phase of the reference signal so that the phase of the reference signal is synchronized with the luminance fluctuation of the smoothed image (step S10).
- step S09 When it is determined that they are synchronized (YES in step S09), the process proceeds to step S11, and the lighting device detection circuit 26 collects a synchronized pixel (pixel group) whose luminance is higher than a predetermined luminance threshold value from the synchronized image. Is determined as a lamp, and the position of the lamp on the synchronized image is detected.
- the lighting device detection circuit 26 uses pattern matching to determine whether the shape of the lighting device (pixel group) is similar to a self-luminous road sign. If it is similar to the self-luminous road sign (YES in step S11), the lamp detection circuit 26 labels the lamp (pixel group) as a self-luminous road sign (step S15), and proceeds to step S21.
- the lamp detection circuit 26 determines whether the hue of the position of the camera image corresponding to the position of the lamp is similar to the lamp color of the traffic signal. Judgment is made (step S13). If it is similar to the traffic light color (YES in S13), the light detector detection circuit 26 labels the light device (pixel group) as a traffic light signal light (step S17), and proceeds to step S21. If it is not similar to the color of the traffic light (NO in S13), the lamp detection circuit 26 labels the lamp (pixel group) as another lamp (step S19), and proceeds to step S21.
- step S21 If the type determination of all the pixel groups determined as lamps from the synchronized image is not completed (NO in step S21), the process returns to step S11, and the type determination is performed for the pixel groups (lamps) that are not determined. carry out. After the type determination is completed for all the pixel groups (lamps) (YES in step S21), the lamp detection device transfers information on the detected lamps to the vehicle CPU 16.
- the brightness difference that is a brightness difference between adjacent pixels in the edge portion is deliberately adjusted so that the brightness difference in the edge portion is small (S05), Synchronized pixels whose luminance changes synchronously are extracted (S07).
- a luminance gradient that is a luminance difference between adjacent pixels in the edge portion becomes smooth.
- the noise When noise is generated in the frequency band of the AC power cycle supplied to the lamp, the noise is also extracted as a synchronized pixel whose luminance changes in synchronization. Therefore, by adjusting the brightness difference of the edge portion so that the brightness gradient, which is the brightness difference between adjacent pixels in the edge portion, becomes gentle, the edge portion moves by changing the position on the image. Since erroneous detection of noise due to luminance step response as a synchronization pixel is suppressed, noise included in the synchronization image is reduced, and the lamp can be detected stably.
- the extending direction (horizontal and vertical) of the edge portion to be subjected to edge detection and smoothing processing is not limited.
- the vehicle body vibrates up and down, so that the amount of movement of the edge portion is greater in the vertical direction than in the horizontal direction. Therefore, in order to reduce the image processing burden and efficiently remove noise, only the edge portions that are continuous in the horizontal direction among the edge portions may be adjusted to be small.
- the edge smoothing circuit 22 in FIG. 1 has a luminance between adjacent pixels only in the edge portion 31a that is continuous in the horizontal direction among the edge portions 31a shown in FIG. 2A (b) detected by the edge detection circuit 21.
- the smoothing process can be performed only on the edge portion 31a that is continuous in the horizontal direction among the edge portions 31a. Thereby, a smooth image as shown in FIG. 2B (c) can be generated.
- the edge detection circuit 21 in FIG. 1 may detect only the edge portion 31b that is continuous in the horizontal direction from the camera image in FIG. 2B (a).
- the edge smoothing circuit 22 can perform a smoothing process on all the edge portions 31 b detected by the edge detection circuit 21.
- a smooth image as shown in FIG. 2B (c) can be generated.
- a pixel group composed of two vertical pixels and four horizontal pixels, the luminance of which is not adjusted, constitutes a region 32b for detecting a lighting device.
- the detection area 32b in FIG. 2B (c) is wider than the detection area 32a in FIG. 2A (c) in which both horizontal and vertical edges are smoothed. This is because the edge portion extending in the vertical direction is left as it is without being smoothed.
- the luminance gradient which is the luminance difference between adjacent pixels in the edge portions that are continuous in the horizontal direction
- the edge part continued in the vertical direction can be left as a region for detecting the lamp without being smoothed. Therefore, it is possible to detect with high sensitivity even a lamp that is far from the vehicle and has low luminance while efficiently reducing noise at the edge portion.
- the image processing burden associated with edge detection or edge smoothing is reduced.
- the lamp detection device is different from FIG. 1 in that it further includes a vehicle behavior determination circuit 13.
- the vehicle behavior determination circuit 13 includes a microcontroller, for example, and configures the vehicle behavior determination circuit 13 by executing a computer program installed in advance.
- the vehicle behavior determination circuit 13 determines whether or not the speed of the position fluctuation of the edge portion in the camera image is equal to or higher than a reference value.
- the reference value is a value serving as a reference for determining whether or not the edge portion needs to be smoothed, and is a predetermined value.
- the vehicle behavior determination circuit 13 can estimate the speed of the position variation of the edge portion in the camera image based on the vehicle pitching information and suspension stroke information obtained by CAN communication from the skid prevention device (ESC).
- FIG. 7 shows an example of a lamp detection method using the lamp detection apparatus of FIG. If the speed of position fluctuation of the edge portion in the camera image is equal to or higher than the reference value (YES in step S02), edge detection (S03) and smoothing (S05) are performed in the same manner as in the first embodiment. On the other hand, if the speed of the position fluctuation of the edge portion is less than the reference value (NO in step S02), the synchronized image generation circuit 24 directly selects the synchronized image from the camera image without performing edge detection and smoothing. What is necessary is just to produce
- ⁇ Noise is likely to occur when the speed of the edge position change in the camera image increases. Therefore, when the speed of the position variation of the edge portion in the camera image is equal to or higher than the reference value, a synchronization pixel is extracted from the smooth image in order to suppress noise generation. On the other hand, if the speed is less than the reference value, noise is not generated so much, and a synchronized pixel is extracted from the camera image. In this way, the necessity for noise suppression is determined according to the speed of the position variation of the edge portion in the camera image. Therefore, the image processing burden from the camera image to the smooth image is reduced, and noise due to the edge portion can be efficiently suppressed.
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
- Closed-Circuit Television Systems (AREA)
- Traffic Control Systems (AREA)
- Photometry And Measurement Of Optical Pulse Characteristics (AREA)
- Circuit Arrangement For Electric Light Sources In General (AREA)
Abstract
Description
次に、図面を参照して、本発明の実施の形態を詳細に説明する。
第1実施形態では、エッジ部における隣接する画素の間の輝度差である輝度勾配が緩やかになるように、敢えて、エッジ部の輝度差を小さく調整したうえ(S05)で、電力の交流周期と同期して輝度が変化する同期画素を抽出する(S07)。エッジ部の隣接する画素の間の輝度差である輝度勾配が滑らかになる。エッジ部の高輝度領域がカメラ画像上を移動すると、エッジ部が移動した画素の輝度はステップ応答で変化するため、広い周波数帯域で輝度値がノイズとして発生する。移動による輝度変化が大きいほど、また、移動速度が速いほど、ノイズ量が大きくなる。灯火器に供給される電力交流周期の周波数帯域にノイズが発生すると、ノイズも同期して輝度が変化する同期画素として抽出してしまう。そのため、エッジ部における隣接する画素の間の輝度差である輝度勾配が緩やかになるようにエッジ部の輝度差を小さく調整されることによって、エッジ部が画像上の位置を移動することによる画素の輝度のステップ応答によるノイズを同期画素として誤検出することが抑制されるので、同期画像に含まれるノイズが減少し、安定して灯火器を検出することができる。
第1実施形態(図2A)では、エッジ検出及び平滑処理の対象とするエッジ部に関し、それが延在する方向(水平、垂直)を限定していない。車両が走行している実走行環境において、車体は上下に振動するため、エッジ部の移動量は、水平方向よりも垂直方向が多くなる。そこで、画像処理負担を軽減してノイズを効率的に除去するために、エッジ部のうち、水平方向に連続するエッジ部のみ、その輝度差を小さく調整するようにしてもよい。
カメラ画像におけるエッジ部の位置変動の速度が速くなると、エッジ部が移動した領域にノイズが発生しやすくなる。カメラ画像におけるエッジ部の位置変動の速度が遅ければ、ノイズは発生し難いため、エッジ部を平滑化せずにそのまま残して、検出領域を広く確保することが望ましい。第2実施形態では、カメラ画像におけるエッジ部の変動速度に応じて、ノイズ抑制の必要性を判断する灯火器検出装置及び灯火器検出方法を説明する。
13 車両挙動判断回路
21 エッジ検出回路
22 エッジ平滑回路
23 基準信号設定回路
24 同期画像生成回路(同期画素抽出回路)
26 灯火器検出回路
31a、31b エッジ部
Claims (4)
- 車両に搭載され、前記車両の周囲を撮像してカメラ画像を取得する撮像部と、
前記カメラ画像の中から、所定値以上の輝度差を有するエッジ部を検出するエッジ検出回路と、
前記カメラ画像の中の前記エッジ部の輝度差を小さく調整して平滑画像を生成するエッジ平滑回路と、
前記平滑画像の中から、灯火器に供給される電力の交流周期と同期して輝度が変化する同期画素を抽出する同期画素抽出回路と、
前記同期画素の中から、前記灯火器を検出する灯火器検出回路と、
を備えることを特徴とする灯火器検出装置。 - 前記エッジ平滑回路は、前記エッジ部のうち、水平方向に連続するエッジ部の輝度差を小さく調整することを特徴とする請求項1に記載の灯火器検出装置。
- 前記カメラ画像における前記エッジ部の位置変動の速度が基準値以上であるか否かを判断する車両挙動判断回路を更に備え、
前記同期画素抽出回路は、前記エッジ部の位置変動の速度が基準値以上である場合に、前記平滑画像の中から、交通信号機に供給される電力の交流周期と同期して輝度が変化する同期画素を抽出し、
同期画素抽出回路は、前記エッジ部の位置変動の速度が基準値未満である場合に、前記カメラ画像の中から、前記灯火器に供給される電力の交流周期と同期して輝度が変化する同期画素を抽出する
ことを特徴とする請求項1又は2に記載の灯火器検出装置。 - 車両に搭載された撮像部を用いて、前記車両の周囲を撮像してカメラ画像を取得し、
前記カメラ画像の中から、所定値以上の輝度差を有するエッジ部を検出し、
前記カメラ画像の中の前記エッジ部の輝度差を小さく調整して平滑画像を生成し、
前記平滑画像の中から、灯火器に供給される電力の交流周期と同期して輝度が変化する同期画素を抽出し、
前記同期画素の中から、前記灯火器を検出する
ことを特徴とする灯火器検出方法。
Priority Applications (10)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2015/069599 WO2017006451A1 (ja) | 2015-07-08 | 2015-07-08 | 灯火器検出装置及び灯火器検出方法 |
EP15897714.0A EP3321911B1 (en) | 2015-07-08 | 2015-07-08 | Lamp detection device and lamp detection method |
RU2018104690A RU2668885C1 (ru) | 2015-07-08 | 2015-07-08 | Устройство обнаружения лампы и способ обнаружения лампы |
JP2017527026A JP6455596B2 (ja) | 2015-07-08 | 2015-07-08 | 灯火器検出装置及び灯火器検出方法 |
CN201580081528.9A CN107851382B (zh) | 2015-07-08 | 2015-07-08 | 灯具检测装置及灯具检测方法 |
US15/742,404 US10074022B2 (en) | 2015-07-08 | 2015-07-08 | Lamp detection device and lamp detection method |
MX2018000029A MX364013B (es) | 2015-07-08 | 2015-07-08 | Dispositivo de deteccion de lampara y metodo de deteccion de lampara. |
BR112018000133-3A BR112018000133B1 (pt) | 2015-07-08 | 2015-07-08 | Dispositivo de detecção de lâmpada e método de detecção de lâmpada |
CA2991471A CA2991471C (en) | 2015-07-08 | 2015-07-08 | Lamp detection device and lamp detection method |
KR1020187000796A KR101920186B1 (ko) | 2015-07-08 | 2015-07-08 | 등화기 검출 장치 및 등화기 검출 방법 |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2015/069599 WO2017006451A1 (ja) | 2015-07-08 | 2015-07-08 | 灯火器検出装置及び灯火器検出方法 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2017006451A1 true WO2017006451A1 (ja) | 2017-01-12 |
Family
ID=57685055
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2015/069599 WO2017006451A1 (ja) | 2015-07-08 | 2015-07-08 | 灯火器検出装置及び灯火器検出方法 |
Country Status (10)
Country | Link |
---|---|
US (1) | US10074022B2 (ja) |
EP (1) | EP3321911B1 (ja) |
JP (1) | JP6455596B2 (ja) |
KR (1) | KR101920186B1 (ja) |
CN (1) | CN107851382B (ja) |
BR (1) | BR112018000133B1 (ja) |
CA (1) | CA2991471C (ja) |
MX (1) | MX364013B (ja) |
RU (1) | RU2668885C1 (ja) |
WO (1) | WO2017006451A1 (ja) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR102233391B1 (ko) * | 2014-06-16 | 2021-03-29 | 팅크웨어(주) | 전자 장치, 전자 장치의 제어 방법 및 컴퓨터 판독 가능한 기록 매체 |
CN112131414A (zh) * | 2020-09-23 | 2020-12-25 | 北京百度网讯科技有限公司 | 信号灯的图像的标注方法、装置、电子设备以及路侧设备 |
US11681780B2 (en) * | 2020-09-30 | 2023-06-20 | Nissan North America, Inc. | Annotation and mapping for vehicle operation in low-confidence object detection conditions |
CN112700410A (zh) * | 2020-12-28 | 2021-04-23 | 北京百度网讯科技有限公司 | 信号灯位置确定方法、装置、存储介质、程序、路侧设备 |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH05316361A (ja) * | 1992-04-24 | 1993-11-26 | Sony Corp | ブロック歪除去フィルタ |
JP2008134916A (ja) * | 2006-11-29 | 2008-06-12 | Denso Corp | 車載用車両前方認識装置 |
JP2013258685A (ja) * | 2012-05-16 | 2013-12-26 | Sharp Corp | 画像処理装置、動画像処理装置、映像処理装置、画像処理方法、映像処理方法、テレビジョン受像機、プログラム、及び記録媒体 |
Family Cites Families (18)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH03217807A (ja) * | 1990-01-23 | 1991-09-25 | Dainippon Screen Mfg Co Ltd | 合焦点条件検出方法 |
US6611610B1 (en) * | 1997-04-02 | 2003-08-26 | Gentex Corporation | Vehicle lamp control |
ES2391556T3 (es) * | 2002-05-03 | 2012-11-27 | Donnelly Corporation | Sistema de detección de objetos para vehículo |
JP4253271B2 (ja) * | 2003-08-11 | 2009-04-08 | 株式会社日立製作所 | 画像処理システム及び車両制御システム |
JP2005301518A (ja) * | 2004-04-08 | 2005-10-27 | Toyota Motor Corp | 信号機検出装置および信号機検出方法 |
WO2008038370A1 (en) * | 2006-09-28 | 2008-04-03 | Pioneer Corporation | Traffic information detector, traffic information detecting method, traffic information detecting program, and recording medium |
JP4746575B2 (ja) * | 2007-03-16 | 2011-08-10 | 株式会社東芝 | 画像処理装置及びその方法 |
CA2705375A1 (en) * | 2007-11-16 | 2009-05-22 | Tele Atlas B.V. | Method of and apparatus for producing lane information |
CN102834309B (zh) * | 2010-02-26 | 2016-12-21 | 金泰克斯公司 | 自动车辆设备监控、报警和控制系统 |
EP2439716B1 (en) * | 2010-09-16 | 2013-11-13 | Ricoh Company, Ltd. | Object identification device, moving object controlling apparatus having object identification device and information presenting apparatus having object identification device |
JP2013097676A (ja) * | 2011-11-02 | 2013-05-20 | Honda Elesys Co Ltd | 到達時間推定装置、到達時間推定方法、到達時間推定プログラム、及び情報提示装置 |
RU116258U1 (ru) * | 2011-11-15 | 2012-05-20 | Игорь Юрьевич Мацур | Система регулирования движения транспортных средств |
JP2014006885A (ja) * | 2012-05-31 | 2014-01-16 | Ricoh Co Ltd | 段差認識装置、段差認識方法及び段差認識用プログラム |
EP2871101B1 (en) * | 2012-07-03 | 2020-06-17 | Clarion Co., Ltd. | Vehicle surrounding monitoring device |
US9056395B1 (en) * | 2012-09-05 | 2015-06-16 | Google Inc. | Construction zone sign detection using light detection and ranging |
CN103345766B (zh) * | 2013-06-21 | 2016-03-30 | 东软集团股份有限公司 | 一种信号灯识别方法及装置 |
JP6233501B2 (ja) * | 2014-03-10 | 2017-11-22 | 日産自動車株式会社 | 信号機検出装置及び信号機検出方法 |
JP6222343B2 (ja) * | 2014-03-10 | 2017-11-01 | 日産自動車株式会社 | 信号機検出装置及び信号機検出方法 |
-
2015
- 2015-07-08 KR KR1020187000796A patent/KR101920186B1/ko active IP Right Grant
- 2015-07-08 WO PCT/JP2015/069599 patent/WO2017006451A1/ja active Application Filing
- 2015-07-08 MX MX2018000029A patent/MX364013B/es active IP Right Grant
- 2015-07-08 CN CN201580081528.9A patent/CN107851382B/zh active Active
- 2015-07-08 BR BR112018000133-3A patent/BR112018000133B1/pt active IP Right Grant
- 2015-07-08 RU RU2018104690A patent/RU2668885C1/ru active
- 2015-07-08 US US15/742,404 patent/US10074022B2/en active Active
- 2015-07-08 EP EP15897714.0A patent/EP3321911B1/en active Active
- 2015-07-08 JP JP2017527026A patent/JP6455596B2/ja active Active
- 2015-07-08 CA CA2991471A patent/CA2991471C/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH05316361A (ja) * | 1992-04-24 | 1993-11-26 | Sony Corp | ブロック歪除去フィルタ |
JP2008134916A (ja) * | 2006-11-29 | 2008-06-12 | Denso Corp | 車載用車両前方認識装置 |
JP2013258685A (ja) * | 2012-05-16 | 2013-12-26 | Sharp Corp | 画像処理装置、動画像処理装置、映像処理装置、画像処理方法、映像処理方法、テレビジョン受像機、プログラム、及び記録媒体 |
Non-Patent Citations (1)
Title |
---|
See also references of EP3321911A4 * |
Also Published As
Publication number | Publication date |
---|---|
BR112018000133A2 (ja) | 2018-09-04 |
KR101920186B1 (ko) | 2018-11-19 |
EP3321911A4 (en) | 2018-07-18 |
US10074022B2 (en) | 2018-09-11 |
CA2991471C (en) | 2018-08-07 |
JPWO2017006451A1 (ja) | 2018-04-19 |
EP3321911A1 (en) | 2018-05-16 |
CA2991471A1 (en) | 2017-01-12 |
BR112018000133B1 (pt) | 2023-01-17 |
US20180197026A1 (en) | 2018-07-12 |
MX364013B (es) | 2019-04-11 |
CN107851382B (zh) | 2019-04-09 |
EP3321911B1 (en) | 2019-09-11 |
MX2018000029A (es) | 2018-03-15 |
RU2668885C1 (ru) | 2018-10-04 |
JP6455596B2 (ja) | 2019-01-23 |
CN107851382A (zh) | 2018-03-27 |
KR20180016564A (ko) | 2018-02-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US9679207B2 (en) | Traffic light detecting device and traffic light detecting method | |
RU2655256C1 (ru) | Устройство обнаружения светофора и способ обнаружения светофора | |
JP6477876B2 (ja) | 信号機検出装置及び信号機検出方法 | |
US9679208B2 (en) | Traffic light detecting device and traffic light detecting method | |
JP6455596B2 (ja) | 灯火器検出装置及び灯火器検出方法 | |
US9811747B2 (en) | Traffic light detecting device and traffic light detecting method | |
US10942607B2 (en) | Manipulation detection device and video display system that are capable detecting an object on a video display surface | |
WO2016021037A1 (ja) | 信号機検出装置及び信号機検出方法 | |
JP2018005391A (ja) | 点滅物体検出方法及び点滅物体検出装置 | |
JP2017228131A (ja) | 反射物標検出方法及び反射物標検出装置 | |
JP7277205B2 (ja) | 信号機検出方法及び信号機検出装置 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 15897714 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2017527026 Country of ref document: JP Kind code of ref document: A |
|
ENP | Entry into the national phase |
Ref document number: 2991471 Country of ref document: CA |
|
WWE | Wipo information: entry into national phase |
Ref document number: 15742404 Country of ref document: US |
|
WWE | Wipo information: entry into national phase |
Ref document number: MX/A/2018/000029 Country of ref document: MX |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
ENP | Entry into the national phase |
Ref document number: 20187000796 Country of ref document: KR Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2018104690 Country of ref document: RU Ref document number: 2015897714 Country of ref document: EP |
|
REG | Reference to national code |
Ref country code: BR Ref legal event code: B01A Ref document number: 112018000133 Country of ref document: BR |
|
ENP | Entry into the national phase |
Ref document number: 112018000133 Country of ref document: BR Kind code of ref document: A2 Effective date: 20180103 |