JP6615691B2 - Speed measuring device - Google Patents

Description

  The present invention relates to a speed measuring device that measures the speed of a moving object.

  As an imaging device for imaging a subject, for example, Patent Document 1 describes an imaging device in which a plurality of charge accumulation periods having different lengths are set when reading out signal charges accumulated in a light receiving unit. .

  Patent Document 2 describes an imaging device that detects a luminance distribution of an image signal and sets an exposure time according to the detected luminance distribution.

Japanese Patent Laid-Open No. 2005-130054 JP 2005-65119 A

  By the way, a speed measuring device that converts reflected light from a road surface into an electric signal by photoelectric conversion and measures the speed of a moving body based on the electric signal is known. This speed measuring device has an advantage that it can measure the speed of the moving body with respect to the road surface with high accuracy without depending on the GPS radio wave environment as compared with the case of calculating the speed based on the position information of GPS (Global Positioning System). There is.

  When the invention of Patent Document 1 described above is applied to a speed measuring device, for example, when a moving body enters a tunnel, the luminance of each pixel decreases rapidly, but due to a response delay of a gain used for amplification of an image signal. There is a possibility that an image signal suitable for speed measurement for a certain time cannot be obtained.

  Further, since the invention of Patent Document 2 described above is a solution means for performing imaging multiple times, there is a high possibility that it is in principle not compatible with a request for high speed in machine vision in which the frame rate is important.

  Accordingly, an object of the present invention is to provide a speed measuring device that measures the speed of a moving object with high accuracy based on optical information even in an environment where the disturbance light changes drastically.

  In order to solve the above-described problems, the present invention is directed to a photoelectric conversion element that receives light reflected from a road surface and converts the received light into an electrical signal, and directs the light reflected from the road surface toward the photoelectric conversion element. A main body unit that is disposed on a moving body, and has a lens barrel that converges and forms an image on the photoelectric conversion element; a filter disposed between the photoelectric conversion element and the road surface; and the photoelectric conversion element A speed calculation unit that calculates the speed of the moving body by comparing two captured images having different imaging times based on the electrical signal input from the plane, and the plane including the filter has a light transmittance. The speed calculation unit selects a brightness distribution area suitable for the comparison among the plurality of areas, and based on the electrical signal from the photoelectric conversion element corresponding to the area. The speed of the moving body Characterized in that it out.

  According to the present invention, it is possible to provide a speed measurement device that measures the speed of a moving object with high accuracy based on optical information even in an environment in which disturbance light changes drastically.

It is a lineblock diagram of the speed measuring device concerning a 1st embodiment of the present invention. It is a disassembled perspective view of the main-body part with which a speed measuring device is provided. It is explanatory drawing regarding the measurement of the speed based on template matching, (a) is explanatory drawing of the last imaging result, (b) is explanatory drawing of this imaging result. It is a flowchart which shows the process of the speed calculation part with which a speed measurement apparatus is provided. It is a flowchart which shows the process of the speed calculation part with which a speed measurement apparatus is provided. (A) is explanatory drawing regarding the setting of a template area | region when a vehicle moves to a dark place, (b) is explanatory drawing regarding the setting of a template area | region when a vehicle moves to a bright place. It is a block diagram of the speed measurement apparatus which concerns on 2nd Embodiment of this invention. It is explanatory drawing which shows the positional relationship of the filter with which the main-body part of a speed measuring device is provided, and the laser beam which injects into a filter. (A) is explanatory drawing regarding calculation of the inclination of the vehicle with respect to a road surface, (b) is explanatory drawing of the imaging result in the state where the vehicle is not inclined in the left-right direction, (c) is the vehicle inclined in the left-right direction. It is explanatory drawing of the imaging result in a state. It is explanatory drawing of the image pick-up element with which the speed measuring device which concerns on 3rd Embodiment of this invention is provided. It is a flowchart which shows the process of the speed calculation part with which a speed measurement apparatus is provided. It is explanatory drawing which shows the relationship between each area | region and a gain in the speed measurement apparatus which concerns on the modification of this invention. It is explanatory drawing regarding the speed measurement apparatus which concerns on another modification of this invention, (a) is a brightness | luminance histogram of each pixel corresponding to 1st area | region, (b) is the brightness | luminance of each pixel corresponding to 2nd area | region. It is a histogram, (c) is a luminance histogram of each pixel corresponding to the third region, (d) is an explanatory diagram showing a relationship between a representative value of luminance and a gain when an electric signal is amplified. .

<< First Embodiment >>
<Configuration of speed measuring device>
FIG. 1 is a configuration diagram of a speed measuring device 100 according to the first embodiment.
The speed measuring device 100 is a device that measures the speed of the vehicle V (moving body). As shown in FIG. 1, the speed measurement device 100 includes a projector 10, a main body unit 20, a calculation device 30, and a display device 40.

  The projector 10 is a device that projects light toward the road surface F, and is installed in the vehicle V. The projector 10 is arranged so that the projected light is reflected by the road surface F, and a part of the light reflected by the road surface F is incident on the lens 21 b of the main body 20.

  The main body 20 is a camera that converts the reflected light from the road surface F into an electrical signal indicating a pattern of brightness and hue due to the unevenness of the road surface F, and further outputs the above-described electrical signal to the arithmetic device 30. V is installed. As shown in FIG. 1, the main body 20 includes a lens barrel 21, a filter 22, and an image sensor 23 (photoelectric conversion element).

  The lens barrel 21 converges the light reflected by the road surface F toward the image sensor 23 and forms an image on the image sensor 23. The lens barrel 21 includes a bottomed cylindrical container 21a and a lens 21b provided near the opening of the container 21a. As an example, FIG. 1 illustrates a configuration in which the lens barrel 21 includes one lens 21b. However, a plurality of lenses may be stacked in the axial direction of the container 21a.

  The filter 22 absorbs a part of light incident on itself, and has a thin film shape. For example, an ND filter (Neutral Density) can be used as the filter 22, but is not limited thereto. In the example illustrated in FIG. 1, the filter 22 is disposed between the lens 21 b and the image sensor 23. More specifically, the filter 22 is disposed with a predetermined gap between the image sensor 23 and the image sensor 23. The light transmitted through the filter 22 enters the image sensor 23.

FIG. 2 is an exploded perspective view of the main body 20 included in the speed measuring device 100. In FIG. 2, the container 21a (see FIG. 1) is not shown.
As shown in FIG. 2, the filter 22 includes a first region 22 a, a second region 22 b, and a third region 22 c as three “regions” having different light transmittances.

  The first region 22a is a rectangular region having the highest light transmittance, and is provided at a position including the optical axis K of the lens barrel 21 (see FIG. 1). The light transmittance in the first region 22a is approximately 100%, for example. Note that the vicinity of the optical axis K of the image pickup device 23 has a larger amount of light transmission than the peripheral region. Therefore, the first region 22a having the highest light transmittance is disposed in the region including the optical axis K. ing. By photoelectrically converting the light transmitted through the first region 22a, the unevenness of the road surface F can be obtained with high sensitivity without sacrificing the original performance of the camera (that is, the main body 20) even in a relatively dark environment. Can be imaged.

The second region 22b is a rectangular region whose light transmittance is lower than that of the first region 22a, and is adjacent to the first region 22a. The light transmittance in the second region 22b is, for example, 60%.
The third region 22c is a rectangular region whose light transmittance is lower than that of the first region 22a and the second region 22b, and is adjacent to the first region 22a. The light transmittance in the third region 22c is, for example, 20%.

As shown in FIG. 2, in the left-right direction perpendicular to the front-rear direction of the vehicle V, the second region 22b is disposed on one side of the first region 22a, and the third region 22c is disposed on the other side.
In the first region 22a, the second region 22b, and the third region 22c, the boundary line between two adjacent regions is parallel to the moving direction when the vehicle V (see FIG. 1) moves in the front-rear direction. It has become.

  The image sensor 23 shown in FIG. 2 is an element that receives light reflected by the road surface F (see FIG. 1) and transmitted through the filter 22 and converts the received light into an electrical signal. A large number of photodiodes are arranged on a plane in the left-right direction (row / column direction). Furthermore, although not shown, the image sensor 23 includes a plurality of amplifiers adjacent to the photodiode and connected to the photodiode on a one-to-one basis. Each photodiode of the image pickup device 23 generates a photocurrent by the light transmitted through the filter 22 and outputs the photocurrent as an electric signal to each amplifier. For example, a CMOS image sensor (Complementary Metal-Oxide-Semiconductor) that directly reads data in rows and columns can be used as a method of such an image sensor 23, but is not limited thereto.

  The arithmetic device 30 shown in FIG. 1 includes electronic circuits such as a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), and various interfaces, although not shown. Then, the program stored in the ROM is read out and expanded in the RAM, and the CPU executes various processes.

The arithmetic device 30 includes a speed calculation unit 31 and a storage unit 32.
The speed calculation unit 31 has a function of calculating the speed of the vehicle V by comparing two captured images with different imaging times based on the electrical signal input from the imaging element 23.
The storage unit 32 stores a program of the speed calculation unit 31, data of the captured image, and the like. The processing executed by the arithmetic device 30 will be described later.

  The display device 40 is a display, for example, and is connected to the arithmetic device 30. The display device 40 has a function of displaying the calculation result of the calculation device 30 and the like.

<About template matching>
FIGS. 3A and 3B are explanatory diagrams regarding speed measurement based on template matching. FIG. 3A is an explanatory diagram of the previous imaging result, and FIG. 3B is an explanatory diagram of the current imaging result.
While the vehicle V is traveling, the road surface F is imaged at a predetermined frame rate (for example, 50 imaging operations per second) by the main body unit 20 (see FIG. 1). As shown in FIG. 3A, a rectangular template region Q is set in the imaging range P of the imaging element 23. This template area Q is an area used as a reference when calculating the speed of the vehicle V. Then, by using two captured images having different imaging times, a pattern matching (comparison) is used to search where the image of the template region Q included in the previous captured image is located in the current captured image. ing. Such a process is repeated every time the road surface F is imaged.

  In the example shown in FIG. 3A, the template region Q is set in a part of the range corresponding to the second region 22b having a light transmittance of 60% in the imaging range P of the imaging device 23. More specifically, as the vehicle V moves forward, a rectangular template region Q is set on the upstream side in the direction in which the image moves in the imaging range P (see the arrow in FIG. 3A). Yes.

  Among the current imaging results shown in FIG. 3B, an area R corresponding to the image of the template area Q obtained by the previous imaging (see the partially enlarged view of FIG. 3A) is searched by pattern matching. Is done. Based on the number of pixels M between the region R and the template region Q, the actual length per pixel, and a predetermined frame rate, the speed calculation unit 31 (see FIG. 1) The speed is calculated.

  Such pattern matching based on the template region Q is referred to as “template matching”. For example, a normalized cross-correlation method can be used as a template matching method, but is not limited thereto.

  In the template matching, the template region Q is set in the imaging range P at a position corresponding to any of the first region 22a, the second region 22b, and the third region 22c (the second region 22b in FIG. 3). The upper limit of the speed that can be measured by template matching is defined by the total number of pixels N1 in the front-rear direction of the vehicle V. As described above, by setting the boundary lines of the first region 22a, the second region 22b, and the third region 22c in parallel to the front-rear direction instead of the left-right direction of the vehicle V, the measurable upper limit of the speed can be set. Enough can be secured.

<Processing in speed measuring device>
FIG. 4 is a flowchart showing the processing of the speed calculation unit 31 included in the speed measuring device 100 (see FIGS. 1 and 3 as appropriate). It is assumed that the vehicle V is traveling at “START” in FIG. 4 and the template region Q is set in the second region 22b where the light transmittance is 60%.

  In step S101, the speed calculation unit 31 determines whether or not the luminance distribution of the template region Q is random in the previous imaging result (one frame before) (see FIG. 3A). For example, when traveling on an asphalt road surface F, the luminance distribution of the template region Q is often random (that is, suitable for speed measurement). For example, when a white line appears in substantially the entire template region Q, it is often determined that the luminance distribution of the template region Q is not random (is uniform). Note that a statistical index such as a variance value and a preset threshold value can be used to determine whether or not the image is random.

If the luminance distribution of the template region Q is not random in step S101 (S101: No), the processing of the speed calculation unit 31 proceeds to step S102.
In step S102, the speed calculation unit 31 determines whether the average luminance of the template region Q is within a predetermined range. That is, the speed calculation unit 31 determines whether or not the average value (average luminance) of each pixel included in the template region Q has a size suitable for template matching (comparison). That is, the speed calculation unit 31 determines whether or not whiteout or blackout occurs (whether or not the magnitude of the gain used for imaging is appropriate). The “predetermined range” regarding the average luminance is set in advance based on a prior experiment or the like.

When the average brightness of the template region Q is within the predetermined range in step S102 (S102: Yes), the processing of the speed calculation unit 31 proceeds to step S103.
In step S103, the speed calculation unit 31 determines whether the position of the template region Q can be moved. That is, the speed calculation unit 31 determines whether or not the template region Q can be moved to a location where the luminance distribution is random in the second region 22b currently used as the template region Q. For example, when the luminance distribution of the template region Q is uniform and gray, it is estimated that the gain is appropriate, but there is no pattern such as concrete. In such a case, the speed calculation unit 31 attempts to move the template region Q within the second region 22b.
If it is determined in step S103 that the position of the template region Q can be moved (S103: Yes), the processing of the speed calculation unit 31 proceeds to step S104.

  In step S104, the speed calculation unit 31 moves the position of the template region Q. For example, the speed calculation unit 31 moves the template region Q to the upstream side in the direction in which the image moves in the imaging range P (see FIG. 3A). As a result, the template region Q can be moved to a location where the luminance distribution is random while sufficiently securing a speed upper limit measurable at a predetermined frame rate.

  After moving the position of the template region Q in step S104, the speed calculation unit 31 executes template matching in step S105. That is, the speed calculation unit 31 uses the luminance distribution of the template region Q (see FIG. 3A) as a reference, and determines the region R (see FIG. 3B) having the highest correlation value with the luminance distribution as the second region. Search within 22b.

If the average brightness of the template region Q is outside the predetermined range in step S102 (S102: No), the processing of the speed calculation unit 31 proceeds to step S106.
In step S <b> 106, the speed calculation unit 31 sets other regions that are suitable for template matching and have different light transmittances as regions for setting the template region Q (first region 22 a, second region 22 b, or third region 22 c). It is determined whether or not. When other regions with different light transmittances suitable for template matching have not been tried yet (S106: No), the processing of the speed calculation unit 31 proceeds to step S107.

  In step S107, the speed calculation unit 31 changes the area in which the template area Q is set (used). That is, the speed calculation unit 31 selects a region having a luminance distribution suitable for template matching (comparison) from the first region 22a, the second region 22b, and the third region 22c. Note that “suitable for comparison” means that the image is not overexposed or blackout, the average luminance is within a predetermined range, and the luminance distribution is moderately random. ing.

  After performing the process of step S107, the process of the speed calculation unit 31 returns to “START” (“RETURN” in FIG. 5). And the speed calculation part 31 sets the template area | region Q to the area | region suitable for template matching.

FIG. 6A is an explanatory diagram regarding the setting of the template region Q when the vehicle V moves to a dark place.
For example, it is assumed that the template region Q is set in a part of the range corresponding to the second region 22b in each imaging device 23 while traveling on the asphalt road surface F that is shining with sunlight (in FIG. 6A). (See “Previous imaging results”). When entering the tunnel from this state, even if auto gain is applied, it takes some time for it to function. As a result, the average luminance of the template region Q is too low and is not suitable for template matching (comparison). In such a case, in step S107, the speed calculation unit 31 sets a new template region Q in the first region 22a having a higher light transmittance than the second region 22b (see “Current Imaging” in FIG. 6A). (See Results).

  The first region 22a has a light transmittance of about 100%, and is provided in a region including the optical axis K (see FIG. 2) of the lens barrel 21, so that the light sensitivity is very high. As described above, immediately after entering the tunnel, the template area Q is set to a location corresponding to the first area 22a (that is, changed from the second area 22b to the first area 22a), thereby obtaining a predetermined frame rate. Template matching can be appropriately performed without omission.

FIG. 6B is an explanatory diagram regarding the setting of the template region Q when the vehicle V moves to a bright place.
For example, it is assumed that the template area Q is set in a part of the range corresponding to the second area 22b in each imaging element 23 while traveling on the asphalt road surface F ("previous imaging result" in FIG. 6B). See). After that, when the road surface F on which the vehicle V travels changes from asphalt to concrete, the average luminance of the template region Q is excessive and is not suitable for template matching (comparison). In such a case, in step S107, the speed calculation unit 31 sets a new template region Q in the third region 22c having a light transmittance lower than that of the second region 22b (see “Current Imaging” in FIG. 6B). (See Results).

  In this way, by setting the template region Q at a position corresponding to the third region 22c having a light transmittance of 20%, it is possible to appropriately capture the surface pattern even in concrete where the average luminance is likely to increase. Therefore, immediately after the road surface F changes from asphalt to concrete, template matching can be appropriately performed at a predetermined frame rate.

Returning again to FIG. 4, the description will be continued.
If another region having a different light transmittance has already been tried in step S106 (S106: Yes), the processing of the speed calculation unit 31 proceeds to step S108. That is, when there is no region suitable for template matching, the processing of the speed calculation unit 31 proceeds to step S108.
In step S108, the speed calculation unit 31 determines whether the position of the template region Q can be moved in a direction that sacrifices the measurable speed upper limit. That is, the speed calculation unit 31 sets the template region Q on the downstream side in the direction in which the image of the imaging result moves (see FIG. 3A) within the region currently used (for example, the second region 22b). It is determined whether or not it can be moved.

In step S108, when the position of the template region Q can be moved in the direction that sacrifices the measurable speed upper limit (S108: Yes), the processing of the speed calculation unit 31 proceeds to step S109.
In step S109, the speed calculation unit 31 moves the position of the template region Q in a direction that sacrifices the upper speed limit, and then performs template matching in step S105.

In step S103 or step S108, when the position of the template region Q cannot be moved any more (S103: No, S108: No), the processing of the speed calculation unit 31 proceeds to step S110 in FIG.
In step S110, the speed calculation unit 31 performs speed interpolation. That is, the speed calculation unit 31 performs simple extrapolation using the past history speed, or template matching using two images in which the most recent shot or a plurality of shots are vacant, and the vehicle V Calculate the speed. Thereby, for example, even when a white line or the like is temporarily captured in the template region Q, the speed of the vehicle V can be continuously calculated.

Further, after executing template matching in step S105 of FIG. 4, the processing of the speed calculation unit 31 proceeds to step S111 of FIG.
In step S <b> 111, the speed calculation unit 31 determines whether matching has been completed. That is, the speed calculation unit 31 uses the image of the template region Q (see FIG. 3A) included in the previous (one frame before) imaging result as a reference, the region R having the highest correlation value (FIG. 3B). It is determined whether or not the search is successful.

When matching is possible in step S111 (S111: Yes), the processing of the speed calculation unit 31 proceeds to step S112.
In step S112, the speed calculation unit 31 calculates the speed of the vehicle V. That is, the speed calculation unit 31 determines the number M of pixels between the template region Q and the region R searched based on the template matching (see FIG. 3B), the actual length per pixel, and the predetermined number. The speed of the vehicle V is calculated based on the frame rate.

Further, when the matching cannot be performed in step S111 (S111: No), the processing of the speed calculation unit 31 proceeds to step S113.
In step S113, the speed calculation unit 31 performs template matching again in consideration of rotation. For example, when the vehicle V is traveling while making a curve, the region R shown in FIG. 3B is in a positional relationship of rotation with respect to the template region Q by a predetermined angle. Therefore, when using a normalized correlation method or the like having a low response to rotation, it is necessary to try template matching in consideration of rotation.

  In step S <b> 114, the speed calculation unit 31 determines whether matching has been completed. In addition to whether or not a sufficiently high correlation value is obtained, for example, in the traveling direction of the vehicle V, a determination is made as to whether or not the matching is within a range that takes into account the acceleration of the vehicle V. Also good. When matching is possible in step S114 (S114: Yes), the processing of the speed calculation unit 31 proceeds to step S112. On the other hand, when the matching cannot be performed (S114: No), the processing of the speed calculation unit 31 proceeds to step S115. For example, when a grating (a net-like lid), a white line drawn on the road surface F, or the like is shown in the template area Q, template matching may not be performed at the position of the template area Q as it is.

  In step S115, the speed calculation unit 31 determines whether the position of the template region Q can be moved. In other words, the speed calculation unit 31 moves in the direction in which the image of the imaging result moves within the region currently used as the template region Q (first region 22a, second region 22b, or third region 22c) (FIG. 3 ( It is determined whether or not the template area Q can be moved upstream of (a). When the template area Q can be moved (S115: Yes), the processing of the speed calculation unit 31 proceeds to step S116.

In step S116, the speed calculation unit 31 moves the template region Q to a location where no grating or the like is shown.
In step S117, the speed calculation unit 31 performs template matching.
In step S <b> 118, the speed calculation unit 31 determines whether matching has been completed. If matching is possible (S118: Yes), in step S112, the speed calculation unit 31 calculates the speed based on the result of template matching. On the other hand, when the matching cannot be performed (S118: No), in step S110, the speed calculation unit 31 calculates the speed of the vehicle V based on the above-described speed interpolation.

If the template region Q cannot be moved in step S115 (S115: No), the processing of the speed calculation unit 31 proceeds to step S119.
In step S119, can the velocity calculation unit 31 move the position of the template region Q in the direction that sacrifices the measurable velocity upper limit (downstream side in the direction in which the image moves in the imaging range: see FIG. 3A)? Determine whether or not. When the position of the template region Q can be moved in a direction that sacrifices the upper limit of speed (S119: Yes), the processing of the speed calculation unit 31 proceeds to step S116.

On the other hand, when the position of the template region Q cannot be moved in the direction that sacrifices the upper speed limit (S119: No), in step S110, the speed calculation unit 31 calculates the speed of the vehicle V based on the speed interpolation described above. .
After performing the processing of step S110 or step S112, the processing of the speed calculation unit 31 acquires the next frame and returns to “START” in FIG. 4 (RETURN). In this way, the speed calculation unit 31 repeats a series of processes shown in FIGS. 4 and 5.

<Effect>
According to the first embodiment, the template region Q is set in a luminance distribution region suitable for template matching among the first region 22a, the second region 22b, and the third region 22c. Therefore, for example, immediately after the traveling vehicle V enters the tunnel, the template region Q is set (changed) in the first region 22a having the highest light transmittance, so that the vehicle V is set at a predetermined frame rate. You can continue to measure speed. Also, for example, immediately after exiting the tunnel, the template region Q is set (changed) in the second region 22b having a light transmittance of 60%, thereby continuously measuring the speed of the vehicle V at a predetermined frame rate. Can do. Thus, according to the present embodiment, the speed of the vehicle V can be measured with high accuracy in accordance with the state of the road surface F and the change in the travel environment of the vehicle V (change in brightness). That is, even in an environment where the disturbance light changes drastically, the speed of the vehicle V can be measured with high accuracy based on optical information.
In addition, since it is not so difficult to make an appropriate function after enabling the brightness auto-gain provided in many cameras, it is possible to further expand the corresponding brightness range.

  Further, the first region 22a having the highest light transmittance in the filter 22 is disposed at a position including the optical axis K of the lens barrel 21 (see FIG. 2). Therefore, even in a dark environment such as a tunnel, the uneven pattern of the road surface F can be appropriately imaged and the speed of the vehicle V can be measured with high accuracy.

  In the filter 22, the boundary line between the first region 22 a and the second region 22 b and the boundary line between the first region 22 a and the third region 22 c are relative to the moving direction when the vehicle V moves in the front-rear direction. Are parallel to each other (see FIG. 2). Therefore, compared to the case where the boundary line is perpendicular to the moving direction of the vehicle V, it is possible to sufficiently secure the upper limit of speed that can be measured at a predetermined frame rate.

<< Second Embodiment >>
In the second embodiment, the speed measurement device 100A (see FIG. 7) includes laser light irradiation units 51 and 52 that irradiate laser light toward the road surface F, and an inclination calculation unit 33 that calculates the inclination of the vehicle V. The point provided is different from the first embodiment. Others are the same as in the first embodiment. Therefore, a different part from 1st Embodiment is demonstrated and description is abbreviate | omitted about the overlapping part.

<Configuration of speed measuring device>
FIG. 7 is a configuration diagram of a speed measuring device 100A according to the second embodiment. In FIG. 7, the projector 10 and the vehicle V are not shown.
As illustrated in FIG. 7, the speed measurement device 100A includes a main body 20A, a calculation device 30A, a display device 40, and laser light irradiation units 51 and 52.

The laser light irradiation unit 51 is a laser pointer that irradiates laser light toward the road surface F, and is fixed to the right side of the main body 20 by a holder G.
The laser beam irradiation unit 52 is a laser pointer that irradiates a laser beam toward the road surface F, and is fixed to the left side of the main body unit 20 by a holder G.

  In the example shown in FIG. 7, the distance between the optical axis K of the lens barrel 21 and the laser light irradiation unit 51 and the distance between the optical axis K of the lens barrel 21 and the laser light irradiation unit 52 are equal in the left-right direction. ing. The main body unit 20 and the laser beam irradiation units 51 and 52 are installed on the back side of the vehicle V (not shown), for example.

  The laser beam irradiation units 51 and 52 are arranged so that each laser beam and the optical axis K of the lens barrel 21 are parallel to each other. Furthermore, the laser beam irradiation units 51 and 52 are provided at positions where the laser beam irradiated from the laser beam irradiation units 51 and 52 is imaged by the main body unit 20A.

As illustrated in FIG. 7, the arithmetic device 30 </ b> A includes a speed calculation unit 31, a storage unit 32, and an inclination calculation unit 33.
The speed calculation unit 31 has a function of calculating the speed of the vehicle V by the same method as in the first embodiment.
The inclination calculation unit 33 has a function of calculating the inclination of the vehicle V with respect to the road surface F based on the position of the laser beam on the image in the imaging range of the main body 20A. The processing of the inclination calculation unit 33 will be described later.

FIG. 8 is an explanatory diagram showing the positional relationship between the filter 22A included in the main body 20A of the speed measuring device 100A and the laser light incident on the filter 22A.
As shown in FIG. 8, the filter 22 </ b> A has first regions 22 a, second regions 22 b, third regions 22 c, fourth regions 22 d, and fifth regions 22 e as “regions” having different light transmittances. And. The first region 22a including the optical axis K (see FIG. 7) of the lens barrel 21 has the highest light transmittance (for example, approximately 100%), then the second region 22b (for example, 60%), and the third region. The light transmittance decreases in the order of 22c (for example, 20%), the fourth region 22d (for example, several percent), and the fifth region 22e (for example, several percent).

  Incidentally, the light transmittance of the fourth region 22d and the light transmittance of the fifth region 22e may be substantially the same, and one transmittance may be higher than the other.

  In the example shown in FIG. 8, the fourth region 22d, the third region 22c, the first region 22a, the second region 22b, and the fifth region 22e are arranged adjacent to each other in order toward the left side of the vehicle V. In the image sensor 23, the ranges corresponding to the first region 22a, the second region 22b, and the third region 22c are used for speed calculation based on template matching, as in the first embodiment.

The fourth region 22d is a region where the laser beam irradiated from the laser beam irradiation unit 51 (see FIG. 7) on the right side of the lens barrel 21 and reflected by the road surface F is incident.
The fifth region 22e is a region where the laser beam irradiated from the laser beam irradiation unit 51 (see FIG. 7) on the left side of the lens barrel 21 and reflected by the road surface F is incident.
As described above, the laser light is incident on the fourth region 22d and the fifth region 22e, which have lower light transmittance than the other regions. As a result, the brightness around the laser light in the imaging result is lowered, and thus the position of the laser light having high brightness can be easily specified.

<Processing in speed measuring device>
FIG. 9A is an explanatory diagram regarding calculation of the inclination of the vehicle V with respect to the road surface F. FIG.
As shown in FIG. 9A, ½ of the angle of view of the main body 20A is θ. The lower end surface of the lens barrel 21 (the center of the lens 21b on the road surface F side) is set as the origin O, and the x, y, and z axes are set with the origin O as a reference. The positive direction of the x axis coincides with the right direction with respect to the front-rear direction of the vehicle V (not shown). The positive direction of the z axis coincides with the direction in which the laser beam is irradiated from the laser beam irradiation units 51 and 52 (that is, the z axis and each laser beam are parallel).

Further, on the road surface F, the position where the laser light emitted from the laser light irradiation unit 51 is reflected is D1, and the value of the z coordinate of the position D1 is D1 _z . Similarly, in the road surface F, the position where the laser beam emitted from the laser light irradiating unit 52 is reflected and D2, the value of the z-coordinate of the position D2 and D2 _z.
Further, N is the total number of pixels in the x direction (the left-right direction of the vehicle V) in the imaging range P (see FIG. 9B). Then, the actual length δ in the x direction that appears in one pixel in the imaging range P is expressed by the following equation (1).

Further, in the imaging range P, the position where the laser beam of the laser beam irradiation unit 51 is imaged is E1 (see FIG. 9B), and the number of pixels from the center line W in the x direction to the position E1 is n1 (FIG. 9). (See (b)). Similarly, in the imaging range P, the position where the laser beam of the laser beam irradiation unit 52 is imaged is E2, and the number of pixels from the center line W in the x direction to the position E2 is n2.
Further, the distance between the optical axis K of the lens barrel 21 and the central axis of the laser light irradiation unit 51 and the distance between the optical axis K of the lens barrel 21 and the central axis of the laser light irradiation unit 52 are set to L (FIG. 9). (See (a)). Then, the number of pixels n1 described above is expressed by the following equation (2).

Therefore, the above-described value D1 _z is expressed by the following formula (3). Note that the angle θ (1/2 of the field angle of the main body 20A), the total number N of pixels in the x direction, and the distance L between the optical axis K and the central axis of the laser light irradiation unit 51 are known. The number of pixels n1 is obtained based on the imaging result of the main body 20A.

Similarly, the value D2 _z is also determined. Furthermore, the angle at which the vehicle V is inclined in the left-right direction with respect to the road surface F is φ (that is, the angle formed by the plane F perpendicular to the optical axis K of the lens barrel 21 and the road surface F is φ). And the angle φ is obtained by the following equation (4) using the values D1 _z and D2 _z described above. Incidentally, the angle φ is a positive value when the right portion of the vehicle V is inclined with respect to the road surface F below the left portion, and the angle φ is a negative value in the opposite state.

FIG. 9B is an explanatory diagram of an imaging result in a state where the vehicle V is not inclined in the left-right direction. When the vehicle V is not tilted in the left-right direction (that is, angle φ = 0 °: see FIG. 9A), the number of pixels n1, n2 from the center line W in the left-right direction in the imaging range P is equal. Incidentally, as the values D1 _z and D2 _z become smaller (that is, the distance between the laser light irradiation parts 51 and 52 and the road surface F becomes shorter), the values of the number of pixels n1 and n2 become larger.

FIG. 9C is an explanatory diagram of an imaging result in a state where the vehicle V is tilted in the left-right direction.
For example, when the vehicle V is inclined in the left-right direction with respect to the road surface F, the distance between the laser light irradiation unit 52 and the road surface F becomes longer than the distance between the laser light irradiation unit 51 and the road surface F. Consider (D1 _z <D2 _z ). In this case, the positions E1 and E2 of the laser light in the imaging range P have the positional relationship shown in FIG. That is, the number of pixels n1 is larger than the number of pixels n2, and the angle φ indicating the inclination of the vehicle V becomes a positive value. In this way, the inclination calculating unit 33 calculates the angle φ indicating the inclination of the vehicle V with respect to the road surface F based on the position of the laser beam on the image in the imaging range P. Note that the above-described method for calculating the angle φ is an example, and the present invention is not limited to this.

<Effect>
According to the second embodiment, the speed calculation unit 31 can calculate the speed of the vehicle V, and the inclination calculation unit 33 can also calculate the inclination of the vehicle V with respect to the road surface F. That is, based on the position of the laser beam in the imaging range P, the inclination (posture) of the vehicle V with respect to the road surface F can be calculated.

<< Third Embodiment >>
In the third embodiment, the filter 22 (see FIGS. 1 and 2) is omitted from the configuration of the first embodiment. Instead, the photodiode (not shown) of the image sensor 23 has a different length in each region. The exposure time is set. Other configurations are the same as those in the first embodiment. Therefore, a different part from 1st Embodiment is demonstrated and description is abbreviate | omitted about the overlapping part.

<Configuration of speed measuring device>
FIG. 10 is an explanatory diagram of the image sensor 23 provided in the speed measurement device according to the third embodiment. In FIG. 10, illustration of individual photodiodes arranged in the front-rear and left-right directions is omitted.
As shown in FIG. 10, the “region” in which the photodiodes of the image sensor 23 are arranged has a first region 23a, a second region 23b, and a third region 23c.
The first area 23 a is a rectangular area having the longest exposure time when imaging the road surface F, and is provided at a position including the optical axis K of the lens barrel 21. The second area 23b is a rectangular area whose exposure time is shorter than that of the first area 23a, and is adjacent to the first area 23a. The third area 23c is a rectangular area whose exposure time is shorter than that of the first area 23a and the second area 23b, and is adjacent to the first area 23a.

  As shown in FIG. 10, in the first region 23a, the second region 23b, and the third region 23c, the boundary line between two adjacent regions is parallel to the moving direction when the vehicle V moves in the front-rear direction. It has become. A template region Q for performing template matching is set in any one of the first region 23a, the second region 23b, and the third region 23c (for example, the second region 23b).

<Processing in speed measuring device>
FIG. 11 is a flowchart showing the processing of the speed calculation unit 31 (see FIG. 10 as appropriate). In addition, the same step number is attached | subjected to the process similar to 1st Embodiment (refer FIG. 4). In addition, as described with reference to FIG. 10, it is assumed that the road surface F is imaged with different lengths of exposure time in the first region 23 a, the second region 23 b, and the third region 23 c.

If the average brightness of the template region Q is outside the predetermined range in step S102 (S102: No), the processing of the speed calculation unit 31 proceeds to step S201.
In step S201, the speed calculation unit 31 sets another template area Q (first area 23a, second area 23b, or third area 23c) that is suitable for template matching and has a different exposure time length. It is determined whether or not the area has been tried. If another region with a different exposure time length suitable for template matching has not yet been tried (S201: No), the processing of the speed calculation unit 31 proceeds to step S202.

  In step S202, the speed calculation unit 31 changes an area in which the template area Q is set (used). That is, the speed calculation unit 31 selects a luminance distribution region suitable for template matching (comparison) from the first region 23a, the second region 23b, and the third region 23c.

For example, when the vehicle V enters the tunnel in a state where the template area Q is set in the second area 23b where the exposure time is medium, the average brightness of the template area Q becomes too low, so template matching (comparison) Not suitable for. Accordingly, in step S202, the speed calculation unit 31 sets the template region Q in the first region 23a having an exposure time longer than that of the second region 23b.
Further, for example, it is assumed that the template area Q is set in the second area 23b in which the exposure time is medium while traveling on the asphalt road surface F. After that, when the road surface F on which the vehicle V travels changes to concrete, the average brightness of the template region Q is excessive in the second region 23b, so that it is not suitable for template matching (comparison). Accordingly, in step S202, the speed calculation unit 31 sets the template region Q in the third region 23c having an exposure time shorter than that of the second region 23b.

  After performing the process of step S202, the process of the speed calculation unit 31 returns to “START” (“RETURN”). Then, the speed calculation unit 31 sets the template area Q in an area suitable for template matching, and calculates the speed of the vehicle V based on the electrical signal from each image sensor corresponding to the template area Q. As a result, the speed of the vehicle V can be continuously measured with high accuracy even when the traveling environment changes. In addition, since the process after step S103 is the same as that of 1st Embodiment (refer FIG. 4, FIG. 5), description is abbreviate | omitted.

<Effect>
According to the third embodiment, the template region Q is set in a luminance distribution region suitable for template matching among the first region 23a, the second region 23b, and the third region 23c having different exposure times. . As a result, even when the traveling environment changes, the speed of the vehicle V can be continuously measured with high accuracy.

≪Modification≫
As mentioned above, although each embodiment demonstrated the speed measurement apparatus 100 etc. which concern on this invention, this invention is not limited to these description, A various change can be made.
For example, in the first embodiment, the case where the electric signal from the image sensor 23 is amplified with a predetermined gain in the amplifier has been described, but the present invention is not limited to this. That is, for each image sensor corresponding to the first region 22a (see FIG. 2), the electrical signal from each image sensor may be amplified with a gain that corrects the light amount distribution in the first region 22a. . The same applies to the other second region 22b and third region 22c. An example of such control will be described with reference to FIG.

FIG. 12 is an explanatory diagram illustrating a relationship between each region and a gain in the speed measurement device according to the modification.
In the example illustrated in FIG. 12, in the filter 22, a small gain is set for each image sensor corresponding to the first region 22 a having the highest light transmittance. On the other hand, a large gain is set for each imaging element corresponding to the third region 22c having the lowest light transmittance. Further, a medium gain is set for each image sensor corresponding to the second region 22b having a medium light transmittance.

  In this way, by setting the gain so as to cancel the difference in transmittance height in the filter 22, the position of the pattern suitable for matching is set in the front-rear direction of the vehicle V in step S <b> 106 of FIG. 4. You can search from the vertical direction. That is, it is possible to obtain the effect of the filter 22 that is resistant to changes in the amount of light in the surroundings (that can be flexibly adapted) without substantially changing the processing required in the speed measurement device using an image that does not use the filter 22.

  Further, as will be described below, in the configuration of the first embodiment, luminance histograms of the first region 22a, the second region 22b, and the third region 22c are respectively generated, and the gain is changed based on the luminance histogram. May be.

FIG. 13A is a luminance histogram of each pixel corresponding to the first region 22a in the velocity measuring device according to another modified example. It is assumed that the data shown in FIGS. 13A, 13B, and 13C are obtained as a result of amplifying the electrical signal from each imaging element with a predetermined gain (that is, the same in each region). .
As described above, since the first region 22a has the highest light transmittance (for example, approximately 100%), pixels with relatively high luminance occupy the majority. Note that the luminance J1 shown in FIG. 13A is a representative value of the luminance of each pixel corresponding to the first region 22a. As such a representative value, for example, a value in which the number of pixels equal to or higher than the luminance J1 and the number of pixels lower than the luminance J1 can be used (the luminance J2 shown in FIG. 13B and FIG. 13C). The same applies to the luminance J3 shown in FIG.

FIG. 13B is a luminance histogram of each pixel corresponding to the second region 22b.
In the filter 22, since the second region 22b has a medium light transmittance (for example, 60%), pixels having a medium luminance occupy the majority.
FIG. 13C is a luminance histogram of each pixel corresponding to the third region 22c.
In the filter 22, the third region 22c has the lowest light transmittance (for example, 20%), and therefore, pixels with relatively low luminance occupy the majority.

FIG. 13D is an explanatory diagram illustrating a relationship between a representative value of luminance and a gain when an electric signal is amplified.
As shown in FIG. 13D, the gain at the time of amplifying the electric signal of the image sensor 23 may be set to a smaller value as the representative value of luminance is higher. That is, the speed calculation unit 31 amplifies the electrical signal with a gain corresponding to the luminance distribution of the region corresponding to the installation range of the filter 22 with respect to the electrical signal from the image sensor 23 that receives the light transmitted through the filter 22. For example, with respect to the first region 22a, the luminance J1 that is a representative value is relatively high, so that the gain G1 is set (changed) to a smaller value corresponding to the luminance J1. In this way, the difference in transmittance in each region of the filter 22 may be canceled by the gain. This makes it possible to obtain the effect of the filter 22 that can withstand changes in the amount of light in the surroundings (that can flexibly respond) without substantially changing the processing necessary for the speed measurement device using an image that does not use the filter 22.

  Moreover, although 1st Embodiment demonstrated the structure which arrange | positions the 1st area | region 22a (refer FIG. 2) with the highest light transmittance in the position containing the optical axis K of the lens-barrel 21, it does not restrict to this. For example, the most frequently used region (for example, the second region 22b having a medium transmittance) may be disposed at a position including the optical axis K of the lens barrel 21. This makes it possible to use a region with the highest resolution in a general lens for a long period of time, which is advantageous for speed measurement.

  Further, the processing of steps S108 and S109 of FIG. 4 described in the first embodiment may be omitted. The same applies to steps S108 and S109 of FIG. 11 described in the second embodiment.

  In the first and second embodiments, the case where a CMOS image sensor is used as the image sensor 23 included in the main body 20 has been described. However, the present invention is not limited to this. For example, a CCD (Charge Coupled Device) image sensor may be used as the image sensor 23.

  Moreover, although 1st Embodiment (refer FIG. 1) demonstrated the structure which provides the filter 22 between the lens 21b and the image pick-up element 23, it is not restricted to this. That is, the filter 22 may be provided at an arbitrary position in the optical axis direction between the image sensor 23 and the road surface F. For example, in the case where a plurality of lenses are arranged in the optical axis direction, the filter 22 may be provided between the lens and another lens. Further, the filter 22 may be provided on the road surface F side of the lens closest to the road surface F (that is, the front lens). As a result, the transmittance of light incident on the image sensor 23 changes continuously, so that the template region Q can be set at a location with an appropriate luminance distribution.

In each embodiment, the configuration in which the lens barrel 21 includes the lens 21b has been described. However, in addition to the lens 21b, a prism or a mirror may be used.
In the first embodiment, the configuration in which the filter 22 includes three regions (first region 22a, second region 22b, and third region 22c: see FIG. 2) having different light transmittances has been described. Not limited to. That is, the number of the above-described regions may be two, or may be four or more. The same applies to the second and third embodiments.

  Further, the filter may be configured such that the light transmittance continuously changes (in a gradation) in a direction perpendicular to the moving direction when the vehicle V moves in the front-rear direction. Such a configuration is also included in the matter that “a plane including a filter has a plurality of regions having different light transmittances”. Further, in the filter having the above-described configuration, the laser beams of the laser beam irradiation units 51 and 52 are disposed in a region having a smaller light transmittance than other regions (for example, a region having a light transmittance of 30% or less). May be incident. Thereby, in the imaging result, since the brightness around the laser light from the laser light irradiation unit is low, it is easy to specify the position of the laser light having high brightness.

  In the second embodiment, the laser beam irradiation unit 51 is provided on the right side of the main body 20A (see FIG. 7), and the laser beam irradiation unit 52 (see FIG. 7) is provided on the left side of the main body 20A. Not limited to this. For example, one laser beam irradiation unit may be provided on each side of the main body 20A in the front-rear direction. Thereby, the inclination (posture) of the vehicle V in the front-rear direction can be calculated by the same method as in the second embodiment. Further, three or more laser beam irradiation units may be provided, and the front / rear and left / right inclinations of the vehicle V may be calculated. In such a configuration, the laser beam from the laser beam irradiation unit is preferably irradiated so as to form an image in a region having the lowest light transmittance. As a result, the brightness around the laser beam in the imaging result is lowered, so that the position of the laser beam with high brightness can be easily specified.

  Moreover, although 2nd Embodiment demonstrated the case where the area | regions 22a-22e were arranged so that a laser beam may inject into the area | region where the light transmittance is the smallest in the filter 22A, it is not restricted to this. In other words, the laser light may be incident not only on the region where the light transmittance is the smallest, but also on the region where the light transmittance is 50% or less, more preferably, the light transmittance is 30% or less.

  In the second embodiment, the configuration in which the laser beam irradiation units 51 and 52 are arranged so that the optical axis K of the lens barrel 21 (see FIG. 7) and each laser beam are parallel to each other has been described. Not limited to. That is, the laser beam may be irradiated obliquely from the laser beam irradiation units 51 and 52.

  Moreover, although 1st Embodiment demonstrated the structure in which the main-body part 20 (refer FIG. 1) is provided with the filter 22 of 1 sheet, it does not restrict to this. For example, no filter is provided in a region including the optical axis of the lens barrel 21 (a region corresponding to the first region 22a: see FIG. 2), and a filter having a predetermined transmittance (for example, 60%) is provided in the second region 22b. In addition to the arrangement, another filter having a predetermined transmittance (for example, 20%) may be arranged in the third region 22c. That is, the plane including the filter may have a configuration having a plurality of regions having different light transmittances.

  Moreover, although each embodiment demonstrated the case where the main-body parts 20 and 20A were cameras, it is not restricted to this. For example, the light reflected from the road surface is imaged on a comb-shaped photoelectric conversion element through a lens and a slit, a change in photocurrent generated in the photoelectric conversion element is converted into a pulse train, and the pulse train is counted. Each embodiment can also be applied to a so-called spatial filter method for calculating the speed of the vehicle V.

  Moreover, each embodiment can be combined suitably. For example, the second embodiment and the third embodiment are combined, and the exposure times of the image sensors in the first region 23a, the second region 23b, and the third region 23c (see FIG. 10) are set to different lengths. (3rd Embodiment) You may make it calculate the inclination of the vehicle V based on the position of the laser beam in an imaging range (2nd Embodiment).

Moreover, although each embodiment demonstrated the structure which installs the projector 10 (refer FIG. 1) in the vehicle V, you may abbreviate | omit the projector 10. FIG.
Moreover, although each embodiment demonstrated the structure which measures the speed of the vehicle V by the speed measurement apparatus 100 grade | etc., It is not restricted to this. For example, the speed of the “moving body” such as a two-wheeled vehicle or a railway vehicle may be measured by the speed measuring device 100 or the like.

DESCRIPTION OF SYMBOLS 100,100A Speed measuring apparatus 10 Floodlight 20,20A Main part 21 Lens barrel 22,22A Filter 22a 1st area | region (area | region)
22b Second region (region)
22c 3rd area | region (area | region)
22d Fourth region (region)
22e Fifth region (region)
23 Imaging device (photoelectric conversion device)
23a First region (region)
23b Second region (region)
23c Third region (region)
30, 30A Arithmetic unit 31 Speed calculation unit 32 Storage unit 33 Inclination calculation unit 40 Display unit 51, 52 Laser light irradiation unit F Road surface K Optical axis P Imaging range Q Template region V Vehicle (moving body)

Claims (11)

A photoelectric conversion element that receives light reflected by the road surface and converts the received light into an electrical signal;
A lens barrel that converges light reflected by the road surface toward the photoelectric conversion element and forms an image on the photoelectric conversion element;
A filter disposed between the photoelectric conversion element and the road surface;
A main body unit installed on the mobile body,
A speed calculator that calculates the speed of the moving body by comparing two captured images with different imaging times based on the electrical signal input from the photoelectric conversion element;
The plane including the filter has a plurality of regions having different light transmittances,
The speed calculation unit selects a region having a luminance distribution suitable for the comparison among a plurality of the regions, and determines the speed of the moving body based on the electrical signal from the photoelectric conversion element corresponding to the region. A speed measuring device characterized by calculating. The speed measuring device according to claim 1, wherein, in the plurality of regions, a boundary line between two adjacent regions is parallel to a moving direction when the moving body moves in the front-rear direction. The velocity measuring device according to claim 1 or 2, wherein a region including the optical axis of the barrel among the plurality of regions has higher light transmittance than other regions. The speed calculation unit sets a template area used as a reference when calculating the speed of the moving body to any of the plurality of areas, and when the average luminance of the template area is excessive and not suitable for the comparison, The speed measurement device according to any one of claims 1 to 3, wherein a new template region is set in another region having a light transmittance lower than that of the template region at the present time. . The speed calculation unit sets a template area used as a reference when calculating the speed of the moving body to any one of the plurality of areas, and when the average luminance of the template area is too low to be suitable for the comparison, The speed measurement device according to any one of claims 1 to 3, wherein a new template region is set in another region having a light transmittance higher than that of the template region at the present time. . The speed calculation unit amplifies the electric signal with a gain corresponding to a luminance distribution of the region corresponding to an installation range of the filter with respect to the electric signal from the photoelectric conversion element that receives light transmitted through the filter. The speed measuring device according to any one of claims 1 to 3, wherein: A laser beam irradiation unit configured to irradiate a laser beam toward the road surface;
An inclination calculating unit that calculates an inclination of the moving body with respect to the road surface;
The laser beam irradiation unit is arranged so that the laser beam is incident on at least two places in the imaging range of the main body unit,
The inclination calculation unit calculates an inclination of the moving body with respect to the road surface based on a position on the image of the laser beam in an imaging range of the main body. The speed measuring device according to claim 1. A photoelectric conversion element that receives light reflected by the road surface and converts the received light into an electrical signal;
A lens barrel that converges the light reflected on the road surface toward the photoelectric conversion element and forms an image on the photoelectric conversion element, and a main body portion installed on the moving body,
A speed calculator that calculates the speed of the moving body by comparing two captured images with different imaging times based on the electrical signal input from the photoelectric conversion element;
The photoelectric conversion element has a plurality of regions with different exposure times set by the speed calculation unit,
The speed calculation unit selects a region having a luminance distribution suitable for the comparison among a plurality of the regions, and determines the speed of the moving body based on the electrical signal from the photoelectric conversion element corresponding to the region. A speed measuring device characterized by calculating. The speed measurement apparatus according to claim 8, wherein, in the plurality of regions, a boundary line between two adjacent regions is parallel to a moving direction when the moving body moves in the front-rear direction. The speed calculation unit sets a template area used as a reference when calculating the speed of the moving body to any of the plurality of areas, and when the average luminance of the template area is excessive and not suitable for the comparison, The speed measurement apparatus according to claim 8 or 9, wherein a new template area is set in another area having an exposure time shorter than that of the current template area. The speed calculation unit sets a template area used as a reference when calculating the speed of the moving body to any one of the plurality of areas, and when the average luminance of the template area is too low to be suitable for the comparison, The speed measurement apparatus according to claim 8 or 9, wherein a new template area is set in another area having an exposure time longer than that of the current template area.