JP2007265016A - Vehicle detection device and vehicle detection method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a vehicle detection device for accurately detecting the head and tail of a vehicle even when road surface reflection due to a head light becomes one so as to be spread. <P>SOLUTION: This vehicle detection device is provided with a current image input means 11 for inputting a current image, a background image generation means 2 for generating a background image; a difference image generation means 3 for generating a difference image; a road surface reflection recognition means 8 for detecting a road surface reflection region; a circumscribed rectangle detection means 5 for detecting a vehicle candidate region having a featured region from a new difference image excluding the road surface reflection region from the difference image; and a vehicle recognition means 6 for recognizing the featured region as a vehicle. The road surface reflection recognizing means 8 calculates a luminance for every luminance calculation region, and calculates a luminance threshold based on the maximum luminance, and when the area of a road surface reflection candidate region configured so that the luminance calculation regions having luminance which exceeds the luminance threshold are continued in location is larger than a predetermined value, the road surface reflection recognition means 8 detects the road surface reflection candidate region as a road surface reflection area. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a vehicle detection device and a vehicle detection method for detecting a vehicle from an image obtained by photographing with a camera.

  Conventionally, for vehicle candidates obtained by performing background difference, frame difference processing, light detection processing, etc. on an image captured by a video camera, vehicle detection processing is performed by combining time-series tracking processing. It is known to measure the traffic volume by detecting the trajectory or to detect sudden abnormal events (stop, low speed, traffic jam, runaway, existence of falling objects, etc.).

  For example, video information captured using a video camera is transmitted to the vehicle movement measuring device main body, this information is converted into digital data, and digital data for two screens captured at a certain interval and a vehicle exist. Image processing is performed using data in the image memory that stores information on the absence of information, and when it rains at night, whether it is headlight light or road surface reflection light using a threshold value based on the average brightness for each scanning line There is known a vehicle dynamics measurement method for determining the above and writing the processing result in an image memory and outputting the vehicle tracking, traveling speed, etc. (see, for example, Patent Document 1).

As for the light detection process, for example, the video of a TV camera is binarized after morphology, each light emitter is labeled, the position and shape of each light emitter is calculated, stored in a table, There is known a method of extracting a pair of lights and separating and judging road surface reflections and parallel traveling vehicles from horizontal pair lights to judge a headlight of a vehicle for each lane (see, for example, Patent Document 2). .
JP-A-4-314199 JP 11-353580 A

However, in the conventional vehicle detection device, when the monitoring place is a tunnel environment or the like and the camera is installed at a low position, the vehicle candidate is detected by detecting two headlights or two road surface reflections. The road surface reflection by the headlights is greatly increased to one, and the vehicle head and the vehicle tail cannot be accurately detected, resulting in an error.
In addition, there is a situation in which headlight light of a vehicle traveling far away enters a monitoring area that is a vehicle detection target area, so that the road surface reflection by the headlight or the headlight is detected as a vehicle candidate. .

  The present invention has been made to solve the above-described circumstances, and can accurately detect the head and tail of a vehicle even when the road surface reflection by the headlight becomes one and spreads. An object of the present invention is to provide a vehicle detection device and a vehicle detection method. Moreover, it aims at providing the vehicle detection apparatus and vehicle detection method which can prevent the detection of the vehicle before approaching into the monitoring range.

  In order to achieve the above object, a first vehicle detection device of the present invention is a vehicle detection device that detects a vehicle that has turned on a light, and sequentially captures an image and inputs it as a current image; A background image creation unit that creates a background image based on a plurality of current images sequentially input by the image input unit, a difference image creation unit that creates a difference image based on the current image and the background image, and the current image A road surface reflection detection unit for detecting a road surface reflection area by the light of the vehicle from within the image, and a vehicle candidate area for detecting a vehicle candidate area having a characteristic area from the new difference image obtained by removing the road surface reflection area from the difference image A detection unit, and a vehicle recognition unit that recognizes the characteristic region of the vehicle candidate region as a vehicle, and the road surface reflection detection unit has a luminance calculation area having a predetermined size in the current image. The luminance calculation area having a luminance that exceeds the luminance threshold for calculating a luminance threshold based on the maximum luminance indicating the maximum luminance among the luminances and performing a comparison with the luminance is positional. When the area of the road surface reflection candidate area formed by being continuous to the road surface is larger than a predetermined value, the road surface reflection candidate area is detected as the road surface reflection area.

  With this configuration, even when the road surface reflection by the headlight becomes one and spreads, it is possible to accurately detect the front and the rear of the vehicle. In addition, by measuring the brightness and detecting the road surface reflection area of the headlight, the vehicle head and the vehicle tail can be accurately detected even when the camera is installed at a low position.

  Moreover, the 2nd vehicle detection apparatus of this invention is set as the structure which has a vehicle tracking part which tracks the said vehicle.

  With this configuration, it is possible to accurately detect a traveling state such as a vehicle speed by tracking the vehicle while accurately grasping the vehicle.

  In addition, the third vehicle detection device of the present invention includes an image input unit that sequentially captures images and inputs them as a current image, and a background that creates a background image based on a plurality of current images that are sequentially input by the image input unit. An image creation unit; a difference image creation unit that creates a difference image based on the current image and the background image; a vehicle candidate region detection unit that detects a vehicle candidate region having a characteristic region from the difference image; A vehicle recognition unit that recognizes the characteristic region of the vehicle candidate region as a vehicle, and a vehicle detection determination unit that determines presence or absence of vehicle detection based on a position where the vehicle is located, and the vehicle detection determination The unit calculates an average luminance value for each horizontal line in the vehicle candidate area, extracts a maximum value line indicating a horizontal line having the largest one of the average luminance values, and the maximum value line is Are configured to detect said recognition vehicle only when located within the area as a vehicle.

  With this configuration, detection of the vehicle before entering the monitoring range can be prevented. Further, it is possible to prevent erroneous detection at the time of vehicle detection without recognizing the headlight light of the vehicle traveling far away as the vehicle in the monitoring area.

  In addition, the fourth vehicle detection device of the present invention has a vehicle tracking unit that tracks the vehicle when the vehicle is detected by the vehicle detection determination unit.

  With this configuration, it is possible to prevent erroneous recognition of a traveling state such as the speed of a vehicle traveling far away and erroneous detection of sudden abnormal events such as vehicle congestion.

  The first vehicle detection method of the present invention is a vehicle detection method in which a vehicle with a light turned on is detected. In the image input step in which images are sequentially captured and input as a current image; A step of creating a background image based on a plurality of current images sequentially input, a step of creating a difference image based on the current image and the background image, and a vehicle light from within the current image A road surface reflection detection step in which a road surface reflection region is detected, a step in which a vehicle candidate region having a characteristic region is detected from a new difference image obtained by removing the road surface reflection region from the difference image, and the vehicle candidate region An area having a characteristic is recognized as a vehicle, and the road surface reflection detecting step is performed for each luminance calculation area of a predetermined size in the current image. The luminance is calculated, a luminance threshold is calculated based on the maximum luminance indicating the maximum luminance among the luminances, and the luminance calculation region having a luminance that exceeds the luminance threshold for comparison with the luminance is continuously located In this case, when the area of the road surface reflection candidate area configured is larger than a predetermined value, the road surface reflection candidate area is detected as the road surface reflection area.

  With this configuration, even when the road surface reflection by the headlight becomes one and spreads, it is possible to accurately detect the front and the rear of the vehicle. In addition, by measuring the brightness and detecting the road surface reflection area of the headlight, the vehicle head and the vehicle tail can be accurately detected even when the camera is installed at a low position.

  In the second vehicle detection method of the present invention, sequential images are captured and input as current images, and an image input step and a background image is created based on a plurality of current images sequentially input in the image input step. A step of creating a difference image based on the current image and the background image, a step of detecting a vehicle candidate region having a characteristic region from the difference image, and the feature of the vehicle candidate region A vehicle recognition unit that recognizes a certain area as a vehicle, and a vehicle detection determination step that determines whether or not a vehicle is detected based on a position where the vehicle is located. The vehicle detection determination step includes: An average luminance value is calculated for each horizontal line in the candidate area, and a maximum value line indicating a horizontal line having the largest one among the average luminance values is extracted. Maximum line is a method for the recognition vehicle only when located within a predetermined area is detected as the vehicle.

  With this configuration, detection of the vehicle before entering the monitoring range can be prevented. Further, it is possible to prevent erroneous detection at the time of vehicle detection without recognizing the headlight light of the vehicle traveling far away as the vehicle in the monitoring area.

  The present invention can accurately detect the head and tail of a vehicle even when the road surface reflection by the headlight becomes one and spreads. Moreover, the detection of the vehicle before entering the monitoring range can be prevented.

  Even if the camera is installed at a low position by measuring the brightness of the headlight road surface by measuring the brightness, the vehicle head and tail can be detected accurately and the vehicle is traveling far away. Without recognizing the headlight light of the vehicle as a vehicle in the monitoring area, it is possible to prevent errors and erroneous detections during vehicle detection.

  Hereinafter, a vehicle detection apparatus according to an embodiment of the present invention will be described with reference to the drawings.

(First embodiment)
FIG. 1 is a block diagram of a vehicle detection apparatus 100 according to the first embodiment of the present invention. The vehicle detection apparatus 100 includes a current image input unit 1, a background image generation unit 2, a difference image generation unit 3, a binary image generation unit 4, a circumscribed rectangle detection unit 5, a vehicle recognition unit 6, a vehicle detection unit 7, a road surface. It has reflection recognition means 8 and image storage means 10. The image storage unit 10 includes a current image storage unit 11, a background image storage unit 12, a difference image storage unit 13, and a binary image storage unit 14.

  The current image input means 1 sequentially inputs the current image within the monitoring range using a video camera or the like. The background image creation means 2 creates a background image by averaging or accumulating based on the current image stored by the current image storage means 11 and the original background image stored by the background image storage means 11. The difference image creation means 3 takes the difference between the current image stored by the current image storage means 11 and the background image stored by the background image storage means 12 and creates a difference image using background difference processing, frame difference processing, or the like. To do. The binarized image creating unit 4 creates a binarized image of the difference image stored by the difference image storing unit 13.

  The road surface reflection recognizing means 8 detects a high luminance area having a luminance of a predetermined value or higher continuously in the current image stored by the current image storage means 11, and uses the high luminance area as a road surface by a vehicle headlight. Recognized as a reflection area. The road surface reflection recognition means 8 will be described in detail later. The circumscribed rectangle detecting means 5 detects a circumscribed rectangle that becomes a characteristic area of the vehicle in consideration of the binarized image stored by the binarized image storing means and the influence of the road surface reflection recognized by the road surface reflection recognizing means 8. To do. The vehicle recognition means 6 recognizes an image in the circumscribed rectangle as a vehicle by processing the image data included in the circumscribed rectangle area detected by the circumscribed rectangle detecting means 5 using a technique such as pattern recognition. To do. The vehicle tracking unit 7 performs a tracking process of the vehicle recognized by the vehicle recognition unit 6.

  The current image storage unit 11 stores the current image input by the current image input unit 1. The background image storage unit 12 stores the background image created by the background image creation unit 2. The difference image storage means 13 stores the difference image created by the difference image creation means 3. The binarized image storage unit 14 stores the binarized image created by the binarized image creation unit 4.

  Here, the current image input means 1 is an example of an image input unit. The background image creation means 2 is an example of a background image creation unit. Moreover, the difference image creation means 3 is an example of a difference image creation unit. The road surface reflection recognition unit 8 is an example of a road surface reflection detection unit. The circumscribed rectangle detecting means 5 is an example of a vehicle candidate area detecting unit. The vehicle recognition means 6 is an example of a vehicle recognition unit. The vehicle tracking means 7 is an example of a vehicle tracking unit.

Next, the road surface reflection recognition means 8 will be described in detail.
The road surface reflection recognizing means 8 measures the luminance for each block in a predetermined range divided into a plurality in the current image sequentially input, and recognizes it as a high luminance region when the measured luminance is higher than a threshold value.

  As shown in FIG. 2 (a), the headlight road surface reflection state is generally two left and right, but when the camera installation height is low, such as in a tunnel, as shown in FIG. 2 (b). In some cases, two road surface reflection areas are expanded into one. FIG. 2 is a diagram showing a road surface reflection state of the headlight according to the first embodiment of the present invention. In the present embodiment, this characteristic is used, and if the road surface reflection area exceeds a certain range (M × N (cm)) where the area of continuous predetermined luminance or higher is exceeded, the road surface reflection area is recognized and removed from the vehicle area. . Even when there are two road surface reflections, they can be removed if each exceeds a predetermined range.

  When removing the road surface reflection area, first, as shown in FIG. 3, for example, 4 × 4 pixels of the current image are taken as one block, and average luminance C_ave (i, j) (i: horizontal, j: vertical) for each block. Is calculated. FIG. 3 is a diagram showing a current image divided into a plurality of blocks in the first embodiment of the present invention. Subsequently, the maximum luminance C_max of the current image is measured. And in order to recognize as a road surface reflection candidate, the threshold value C_th used for comparison with the average luminance is a value obtained by multiplying the maximum luminance by a predetermined value. That is, C_th = C_max × α. This α indicates a magnification and is generally 0.7 to 0.8. If the threshold C_th is less than or equal to the minimum luminance C_th_min of the current image, the minimum luminance is set as the threshold. That is, C_th = C_th_min.

  Next, the validity / invalidity of the road surface reflection candidate is determined for each block. Here, if the average luminance C_ave (i, j) is equal to or greater than the threshold C_th, it is considered valid as a road surface reflection candidate. That is, if the frequency of the average luminance is higher than the threshold set based on the maximum luminance, the block is validated as a road surface reflection candidate. This processing for each block can reduce the load of image processing compared to the labeling processing for detecting the area for each pixel × 1 pixel.

  Next, the vertical and horizontal continuity of the high luminance area is checked. Here, the continuity of the effective blocks is checked as road surface reflection in the vertical and horizontal directions for the above road surface reflection candidates. Then, as shown in FIG. 4, if the vertical M and horizontal N (cm) or more, it is recognized as a road surface reflection region. FIG. 4 is a diagram showing a road surface reflection area in the first embodiment of the present invention. For example, the length is 100 cm or more and the width is 100 cm or more. These analysis processes can also be performed by a labeling process, but the processing speed is slowed down due to the heavy load.

  Next, the road surface of FIG. 5B is obtained from the vehicle candidate area obtained by the difference processing between the current image and the background image shown by the broken line in FIG. 5A, that is, the difference image which is a vehicle characteristic area including road surface reflection. The reflection area is removed, and the actual vehicle area in FIG. 5C, that is, a new difference image is extracted. FIG. 5 is a diagram showing a process of creating a new difference image in the first embodiment of the present invention.

  According to the vehicle detection device 100 of the first embodiment of the present invention, even when the camera is installed at a low position by measuring the luminance and detecting the road surface reflection area of the headlight, The vehicle head and the vehicle tail can be accurately detected.

Next, the operation at the time of vehicle detection of the vehicle detection device 100 will be described.
FIG. 6 is an operation flowchart of the vehicle detection apparatus 100 according to the first embodiment of the present invention.

  First, the current image input means 1 sequentially inputs the current image, and the background image creation means 2 creates a background image based on the current image input in time series and the original background image. (Step S1) Subsequently, the difference image creating means 3 creates a difference image between the current image and the background image (Step S2), and the binarized image creating means 4 creates a binarized image from the difference image (Step S1). S3).

  Next, the road surface reflection recognizing means 4 calculates the average luminance for every 4 × 4 pixels (blocks) of the current image (step S11), and calculates the maximum luminance of the current image (step S12). Subsequently, the road surface reflection recognizing means 4 compares the average brightness with the average brightness and calculates a threshold value based on the maximum brightness of the current image. If the average brightness in block units is equal to or greater than the threshold value, the road surface reflection recognition means 4 is effective. If it is less than the threshold value, it is determined that the road surface reflection candidate is invalid (step S13). Then, the road surface reflection recognition unit 4 checks the vertical and horizontal continuity of the effective portion of the road surface reflection candidate (step S14), and detects the road surface reflection area based on the result (step S15). And a road surface reflection area is excluded from an image with a difference (vehicle characteristic area including road surface reflection), and an image with a new difference (vehicle area) is specified (Step S16).

  Next, the circumscribed rectangle detecting means 5 detects the circumscribed rectangle of the pixel having a new difference (step S17), and the vehicle recognizing means 6 recognizes a characteristic part from the circumscribed rectangle as a vehicle (step S18). The tracking means 7 enters a vehicle tracking process (S6). In this embodiment, circumscribed rectangles are detected, but shapes other than rectangles may be used.

  According to the vehicle detection device 100 in the first embodiment of the present invention as described above, the vehicle detection device detects a vehicle with a light turned on, and sequentially captures images and inputs them as the current image. A background image creating unit 2 that creates a background image based on a plurality of current images sequentially input by the current image input unit 11, and a difference image creating unit 3 that creates a difference image based on the current image and the background image. A road surface reflection recognition means 8 for detecting a road surface reflection area by a vehicle light from the current image, and a circumscribed rectangle for detecting a vehicle candidate area having a characteristic area from the new difference image obtained by removing the road surface reflection area from the difference image The vehicle has a detection means 5 and a vehicle recognition means 6 for recognizing a characteristic area of the vehicle candidate area as a vehicle. The road surface reflection recognition means 8 has a luminance calculation of a predetermined size in the current image. The brightness is calculated for each area, the brightness threshold is calculated based on the maximum brightness indicating the maximum brightness among the brightness, and the brightness calculation areas having brightness exceeding the brightness threshold for comparison with the brightness are consecutively located When the area of the road surface reflection candidate area configured is larger than a predetermined value, the road surface reflection candidate area is detected as the road surface reflection area, and the road surface reflection by the headlight becomes one and spreads However, it is possible to accurately detect the head and tail of the vehicle.

(Second Embodiment)
Next, the vehicle detection apparatus 200 in the 2nd Embodiment of this invention is demonstrated.
FIG. 7 is a block diagram showing the configuration of the vehicle detection apparatus according to the second embodiment of the present invention. In the figure, the same reference numerals are given to the same parts as those in FIG. 1 described in the first embodiment, and the description thereof is omitted. The vehicle detection device 200 includes vehicle determination means 9 instead of the road surface reflection recognition means 8.
The vehicle determination means 9 is an example of a vehicle detection determination unit.

  When the circumscribed rectangular feature area recognized as including the vehicle by the vehicle recognizing means 6 is out of the monitoring range with a brightness of a predetermined value or more, the vehicle determination means 9 recognizes the road surface reflection area by the headlight of the vehicle. The vehicle area is determined by excluding the part.

  If the road surface reflection area cannot be recognized, as shown in FIG. 8, when the road surface reflection by the headlight of a vehicle traveling far away enters the monitoring range, the headlight light gradually enters the monitoring range. The apparent travel distance will be reduced, the speed of the vehicle (headlight light) being measured will be reduced, that is, it will be recognized that the vehicle speed is slow, and false alarms may occur as sudden events such as vehicle stoppage or low speed. However, the vehicle determination means 9 can measure the average luminance for each horizontal line of the circumscribed rectangle, and recognize the road reflection region when the horizontal line whose measured average luminance is the maximum value is outside the monitoring range. . FIG. 8 is a diagram showing a vehicle traveling far away in the second embodiment of the present invention.

  Next, the calculation of the average luminance for each horizontal line by the vehicle determination means 9 will be specifically described. FIG. 9 is a diagram showing a horizontal line and a maximum value line in a vehicle candidate rectangle according to the second embodiment of the present invention.

  First, as shown in FIG. 9A, the average luminance value for each horizontal line of the vehicle candidate rectangle detected by the circumscribed rectangle detecting means 5 is measured, and then, as shown in FIG. 9B, the average luminance value is measured. The horizontal line with the maximum value (maximum value line) is detected. Then, it is determined whether or not the detected maximum value line is within the monitoring range. If it is outside the monitoring range as shown in FIG. 10, registration as a vehicle candidate is not performed. FIG. 10 is a diagram showing that the rectangle of the vehicle candidate in the second embodiment of the present invention is outside the monitoring range. The reason for determining whether or not to be a vehicle candidate based on the maximum value line is that the maximum value line having the maximum average luminance value is considered to be the road surface reflection portion closest to the actual vehicle.

Next, the operation at the time of vehicle detection of the vehicle detection device 200 will be described.
FIG. 11 is an operation flowchart of the vehicle detection apparatus 200 according to the second embodiment of the present invention.

  First, the current image input means 1 sequentially inputs the current image, and the background image creation means 2 creates a background image based on the current image input in time series and the original background image (step S1), and the difference image The creation means 3 creates a difference image between the current image and the background image by difference processing (step S2). Subsequently, the binarized image creating means 4 creates a binarized image from the difference image (step S3), and the circumscribed rectangle detecting means 5 detects the circumscribed rectangle of the pixel having the difference (step S4). Then, the vehicle recognition means 6 recognizes the detected circumscribed rectangular feature as a vehicle (step S5). In this embodiment, circumscribed rectangles are detected, but shapes other than rectangles may be used.

  Next, the vehicle determination means 9 calculates the average brightness for each horizontal line of the detected vehicle candidate rectangle, which is a circumscribed rectangle (step S21), and then the maximum value of the average brightness (the maximum average brightness is Cmax, the maximum value). The line is defined as I line) (step S22). Then, the vehicle determination unit 9 determines whether or not the maximum average luminance Cmax of the maximum value line is equal to or greater than a luminance threshold Cmax_th indicating a predetermined luminance (step S23). Is determined (step S24). If the maximum value line is within the monitoring range, the vehicle determination means 9 determines and registers the vehicle candidate rectangle as a vehicle (step S26), and performs tracking processing of the registered vehicle (step S6). When the maximum average brightness Cmax of the maximum value line is larger than the brightness threshold value Cmax_th, and when the maximum value line is outside the monitoring range, the vehicle determination unit 9 performs the process without determining and registering the vehicle. The process ends (step S25).

  Such a vehicle detection device 200 according to the second embodiment of the present invention includes a current image input unit 1 that sequentially captures images and inputs them as a current image, and a plurality of current images that are sequentially input by the current image input unit 1. A background image creating means 2 for creating a background image based on the difference image creating means 3 for creating a difference image based on the current image and the background image; and a vehicle candidate area having a characteristic area from the difference image. There are circumscribed rectangle detecting means 5, vehicle recognition means 6 for recognizing the characteristic area of the vehicle candidate area as a vehicle, and vehicle determining means 9 for determining the presence or absence of vehicle detection based on the location of the vehicle. Then, the vehicle determination means 9 calculates an average luminance value for each horizontal line in the vehicle candidate area, extracts a maximum value line indicating a horizontal line having the largest one among the average luminance values, and determines the maximum value. Ins In the structure for detecting a vehicle recognition vehicle only when located within a predetermined region, it is possible to prevent the detection of the vehicle before entering the monitoring range.

  INDUSTRIAL APPLICABILITY The present invention is useful for a vehicle detection device or the like that can accurately detect the head and tail of a vehicle even when the road surface reflection by the headlight becomes one and spreads. Further, the present invention is useful for a vehicle detection device that can prevent detection of a vehicle before entering the monitoring range.

The block diagram which shows the structure of the vehicle detection apparatus in the 1st Embodiment of this invention. The figure which shows the road surface reflection state of the headlight in the 1st Embodiment of this invention. The figure which shows the present image divided | segmented into the some block in the 1st Embodiment of this invention The figure which shows the road surface reflective area | region in the 1st Embodiment of this invention. The figure which shows the creation process of the new difference image in the 1st Embodiment of this invention. The flowchart which shows operation | movement of the vehicle detection apparatus in the 1st Embodiment of this invention. The block diagram which shows the structure of the vehicle detection apparatus in the 2nd Embodiment of this invention. The figure which shows the vehicle which drive | works the distant place in the 2nd Embodiment of this invention. The figure which shows the rectangle of the vehicle candidate in the 2nd Embodiment of this invention. The figure which shows that the rectangle of the vehicle candidate in the 2nd Embodiment of this invention is outside a monitoring range. Operation | movement flowchart of the vehicle detection apparatus in the 2nd Embodiment of this invention

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Vehicle detection apparatus 1 Current image input means 2 Background image creation means 3 Difference image creation means 4 Binary image creation means 5 circumscribed rectangle detection means 6 Vehicle recognition means 7 Vehicle tracking means 8 Road surface reflection recognition means 9 Vehicle determination means 10 Image Storage means 11 Current image storage means 12 Background image storage means 13 Difference image storage means 14 Binary image storage means 200 Vehicle detection device

Claims (6)

A vehicle detection device for detecting a vehicle that has turned on a light,
An image input unit that sequentially captures images and inputs them as current images;
A background image creation unit that creates a background image based on a plurality of current images sequentially input by the image input unit;
A difference image creation unit for creating a difference image based on the current image and the background image;
A road surface reflection detection unit for detecting a road surface reflection area by the light of the vehicle from within the current image;
A vehicle candidate area detection unit for detecting a vehicle candidate area having a characteristic area from the new difference image obtained by removing the road surface reflection area from the difference image;
A vehicle recognition unit that recognizes the characteristic area of the vehicle candidate area as a vehicle;
Have
The road surface reflection detector is
Calculating the brightness for each brightness calculation area of a predetermined size in the current image;
A luminance threshold is calculated based on the maximum luminance indicating the maximum luminance among the luminances;
When the area of the road surface reflection candidate area constituted by the position of the luminance calculation area having a luminance that exceeds the luminance threshold for comparison with the luminance is consecutively greater than a predetermined value, the road surface reflection candidate area is A vehicle detection device that detects the road surface reflection area. The vehicle detection device according to claim 1,
The vehicle detection apparatus which has a vehicle tracking part which tracks the said vehicle. An image input unit that sequentially captures images and inputs them as current images;
A background image creation unit that creates a background image based on a plurality of current images sequentially input by the image input unit;
A difference image creation unit for creating a difference image based on the current image and the background image;
A vehicle candidate area detection unit for detecting a vehicle candidate area having a characteristic area from the difference image;
A vehicle recognition unit that recognizes the characteristic area of the vehicle candidate area as a vehicle;
A vehicle detection determination unit that determines presence or absence of vehicle detection based on a position where the vehicle is located;
The vehicle detection determination unit
Calculating an average luminance value for each horizontal line in the vehicle candidate region;
Extracting a maximum value line indicating a horizontal line having a maximum one of the average luminance values;
A vehicle detection device that detects the recognized vehicle as a vehicle only when the maximum value line is located within a predetermined region. The vehicle detection device according to claim 3,
The vehicle detection apparatus which has a vehicle tracking part which tracks the said vehicle, when a vehicle is detected by the said vehicle detection determination part. A vehicle detection method in which a vehicle with a light turned on is detected,
An image input step in which sequential images are captured and input as the current image;
A step of creating a background image based on a plurality of current images sequentially input in the image input step;
A difference image is created based on the current image and the background image;
A road surface reflection detection step in which a road surface reflection area by the light of the vehicle is detected from within the current image;
A vehicle candidate area having a characteristic area is detected from a new difference image obtained by removing the road surface reflection area from the difference image;
Recognizing the characteristic area of the vehicle candidate area as a vehicle;
Have
The road surface reflection detection step includes:
Luminance is calculated for each luminance calculation area of a predetermined size in the current image,
A luminance threshold is calculated based on the maximum luminance indicating the maximum luminance among the luminances,
When the area of the road surface reflection candidate region formed by the position of the luminance calculation region having a luminance that exceeds the luminance threshold value for comparison with the luminance is consecutively greater than a predetermined value, the road surface reflection candidate region is A vehicle detection method detected as the road surface reflection area. An image input step in which sequential images are captured and input as current images;
A step of creating a background image based on a plurality of current images sequentially input in the image input step;
A difference image is created based on the current image and the background image;
A vehicle candidate area having a characteristic area is detected from the difference image;
A vehicle recognition unit for recognizing the characteristic area of the vehicle candidate area as a vehicle;
A vehicle detection determination step in which presence / absence of vehicle detection is determined based on a position where the vehicle is located;
Have
The vehicle detection determination step includes
An average luminance value is calculated for each horizontal line in the vehicle candidate area,
A maximum value line indicating a horizontal line having a maximum one of the average luminance values is extracted,
A vehicle detection method in which the recognized vehicle is detected as a vehicle only when the maximum value line is located within a predetermined region.

Images