JP2008135856A - Body recognizing device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a body recognizing device capable of recognizing an object by a single imaging means while taking the distance to the object into consideration. <P>SOLUTION: A near-infrared night vision device 1 to which the body recognizing device according to the present invention is applied includes a near-infrared-ray projector 10 which projects near infrared rays, a single-lens near-infrared camera 12, a control unit 21 which controls the lamp power of the near-infrared-ray projector 10 according to a vehicle speed, and an ECU 20 having a recognizing unit 22 which recognizes a walker from an image picked up by the near-infrared camera 12. The control unit 21 controls the lamp power of the near-infrared-ray projector 10 so that the luminance of a walker at a distance of a TTC (prediction time to collision) of <4 seconds gets saturated. Further, the recognizing unit 22 decides that a walker having saturated luminance in the picked-up image is present at a distance of a TTC of <4 seconds (within a predetermined distance) and excludes the walker from a target to be warned. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an object recognition apparatus.

While imaging the driver's face with the main camera while irradiating the near infrared rays with the main projector, imaging the driver's face with the sub camera while irradiating the near infrared rays with the sub projector, and using the images of the main camera and the sub camera, A technique for obtaining a distance from a driver's face based on the principle of triangulation is described in Patent Document 1.
JP-A-2005-323180

  As described above, in order to obtain the distance to the object using the principle of triangulation, a plurality of cameras are required. On the other hand, for example, in an automobile, in recent years, a plurality of cameras have been mounted on a single vehicle for driving assistance, danger prediction, danger avoidance assistance, etc., resulting in problems such as cost and mountability. It is coming.

  The present invention has been made to solve the above problems, and provides an object recognition apparatus capable of recognizing an object in consideration of the distance to the object by a single imaging means. With the goal.

  An object recognition apparatus according to the present invention includes: a light irradiating unit that irradiates light; a control unit that adjusts an output of the light irradiating unit; a single imaging unit that images a region irradiated with light by the light irradiating unit; A recognition means for recognizing an object from an image picked up by the image pickup means, and the control means adjusts the output of the light irradiation means so that the luminance in the picked-up image of the object at a predetermined distance becomes a predetermined luminance. The recognition means recognizes the object in consideration of the luminance in the captured image of the object.

  According to the object recognition device of the present invention, the brightness of the captured image of the target object is adjusted by adjusting the output of the light irradiation means so that the brightness of the captured image of the target object at a predetermined distance becomes the predetermined brightness. And the distance to the object. Therefore, by recognizing the luminance in the captured image of the object when recognizing the object, the object can be recognized in consideration of the distance from the object.

  In the object recognition apparatus according to the present invention, it is preferable that the control unit adjusts the output of the light irradiation unit so that the luminance in the captured image of the target existing within a predetermined distance is saturated. At that time, for the object whose luminance in the captured image is saturated, the distance to the object is directly measured by recognizing the object within the predetermined distance by the recognition means. Instead, the object can be recognized in consideration of the distance to the object.

  In the object recognition apparatus according to the present invention, it is preferable that the recognition means excludes an object recognized as being present within a predetermined distance from the recognition object.

  For example, in a situation that calls attention to a detection object approaching the driver of the vehicle, it is preferable to recognize the object at a short distance within a predetermined distance with the naked eye of the driver There is. According to the object recognition device of the present invention, the object recognized as existing within the predetermined distance is excluded from the recognition object. It is possible to call attention.

  In the object recognition apparatus according to the present invention, it is preferable that the control unit adjusts the output of the light irradiation unit according to the sensitivity of the imaging unit and / or the exposure time.

  The luminance in the captured image of the object also changes depending on the sensitivity of the imaging means and / or the exposure time. According to the object recognition device of the present invention, since the output of the light irradiation unit is adjusted according to the sensitivity and / or the exposure time of the imaging unit, it is possible to more appropriately adjust the luminance in the captured image of the object. It becomes possible.

  Moreover, it is preferable that a control means adjusts the output of a light irradiation means according to the moving speed of an object recognition apparatus.

  In this case, since the distance to the object having a predetermined luminance is adjusted according to the moving speed, it is possible to recognize the object in consideration of the time until it reaches the object. Become.

  An object recognition apparatus according to the present invention includes a light irradiation unit that irradiates light, a single imaging unit that images a region irradiated with light by the light irradiation unit, an output of the light irradiation unit, and an imaging unit A control unit that adjusts sensitivity and exposure time; and a recognition unit that recognizes an object from an image captured by the imaging unit. The control unit is configured such that the luminance in the captured image of the object separated by a predetermined distance is a predetermined luminance. As described above, at least one of the output of the light irradiation unit, the sensitivity of the imaging unit, and the exposure time is adjusted, and the recognition unit recognizes the target in consideration of the luminance in the captured image of the target. And

  The luminance in the captured image of the object changes depending on the output of the light irradiation means, the sensitivity of the imaging means, and the exposure time. According to the object recognition apparatus of the present invention, at least one of the output of the light irradiation unit, the sensitivity of the imaging unit, and the exposure time is adjusted so that the luminance in the captured image of the object that is separated by a predetermined distance becomes the predetermined luminance. As a result, the luminance in the captured image of the target object is related to the distance to the target object. Therefore, when the object is recognized, the object can be recognized in consideration of the distance from the object by considering the luminance in the captured image of the object. In addition, it is possible to increase the degree of freedom in controlling brightness adjustment.

  An object recognition apparatus according to the present invention includes a light irradiating unit that irradiates light, a single imaging unit that captures an area irradiated with light by the light irradiating unit, and a luminance correction of an image captured by the imaging unit. Correction means for recognizing the object from the image corrected by the correction means, and the correction means is imaged so that the luminance in the captured image of the object separated by a predetermined distance becomes the predetermined luminance. The brightness of the obtained image is corrected, and the recognition means recognizes the object in consideration of the brightness in the captured image of the corrected object.

  According to the object recognition device of the present invention, the brightness of an image captured so that the brightness in the captured image of the object separated by a predetermined distance becomes the predetermined brightness, thereby correcting the brightness in the captured image of the object. The brightness and the distance to the object are related. Therefore, by considering the corrected luminance of the object when recognizing the object, the object can be recognized in consideration of the distance from the object.

  According to the present invention, the output of the light irradiation means is adjusted so that the luminance in the captured image of the object separated by a predetermined distance becomes the predetermined luminance, and the target is determined in consideration of the luminance in the captured image of the object. Since the recognition configuration is adopted, the object can be recognized in consideration of the distance from the object by a single imaging unit.

  DESCRIPTION OF EMBODIMENTS Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the drawings. In the figure, the same reference numerals are used for the same or corresponding parts.

[First Embodiment]
First, the configuration of the object recognition apparatus according to the first embodiment will be described with reference to FIGS. Here, the case where the object recognition device according to the first embodiment is applied to the near-infrared night vision device 1 mounted on a vehicle will be described as an example. FIG. 1 is a block diagram showing a configuration of the near-infrared night vision device 1. FIG. 2 is a diagram for explaining the near infrared irradiation distance of the near infrared night vision apparatus 1. FIG. 3 is a quadrant graph showing the relationship between the vehicle speed and the distance to the object, the distance to the object and the object brightness, the object brightness and the lamp power, and the relationship between the vehicle speed and the lamp power. FIG. 4 is a diagram showing the luminance of a person who is 50 meters ahead, and FIG. 5 is a diagram showing the luminance of a person who is 100 meters ahead. FIG. 6 is a graph showing the relationship between the distance to a person and the pixel luminance value.

  The near-infrared night-vision device 1 displays a near-infrared image in front of the vehicle imaged by a near-infrared camera and assists a driver during night driving, and recognizes a pedestrian as a target object from the captured image. The recognition result (pedestrian information) is displayed superimposed on the near-infrared image. The near-infrared night vision device 1 mainly includes two near-infrared projectors 10, 10, a single near-infrared camera 12, a vehicle speed sensor 14, an electronic control device (hereinafter referred to as “ECU”) 20, and a warning device 30. I have. The near-infrared projector 10 functions as a light irradiating unit described in the claims, and the near-infrared camera 12 functions as an imaging unit.

  The two near-infrared projectors 10, 10 are arranged at the left and right front ends of the vehicle and are attached toward the front of the vehicle. The near-infrared projector 10 irradiates near-infrared rays in front of the vehicle. The near-infrared projector 10 is driven by a driver circuit 11 that is controlled by a control signal from the ECU 20. Here, the output of the near-infrared projector 10, that is, the lamp power, is adjusted by the ECU 20 according to the vehicle speed. Details of the lamp power adjustment method will be described later. The near-infrared projector 10 is turned on when the near-infrared night vision device 1 is activated, and is turned off when the near-infrared projector 1 is stopped. In addition, it is good also as a structure which turns ON / OFF the near-infrared light projector 10 according to a vehicle speed.

  The near-infrared camera 12 is a monocular near-infrared camera, and is disposed on the front side of the vehicle (for example, the back side of the room mirror) and attached toward the front of the vehicle. The near-infrared camera 12 takes in near-infrared light (such as reflected light of near-infrared light from the near-infrared projector 10), and generates a near-infrared image based on light and shade according to the strength of the near-infrared light. This near-infrared image is composed of a near-infrared image of a frame every certain time (for example, 1/30 second). The near-infrared camera 12 transmits near-infrared image information of each frame as an image signal to the ECU 20 at regular intervals.

  The vehicle speed sensor 14 is a sensor that detects the wheel speed of the vehicle on which the near infrared night vision device 1 is mounted. The vehicle speed sensor 14 is also connected to the ECU 20 and outputs a detection result to the ECU 20. The ECU 20 calculates the vehicle speed from the detection result of the vehicle speed sensor 14.

  The ECU 20 includes a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), and the like, and a control unit 21 and a recognition unit 22 are configured by the CPU executing a program stored in the ROM. Is done. The ECU 20 receives an image signal from the near-infrared camera 12 at regular intervals, and recognizes a pedestrian that needs to call the driver's attention from the near-infrared image in consideration of the luminance of the pedestrian in the image. . Then, the ECU 20 displays the pedestrian recognition result on the alerting device 30. In the present embodiment, the control unit 21 corresponds to the control unit described in the claims, and the recognition unit 22 corresponds to the recognition unit.

  The control unit 21 adjusts the lamp power of the near-infrared projector 10 so that the luminance in the captured image of the pedestrian away from the predetermined distance becomes the predetermined luminance. In the present embodiment, the predetermined luminance is determined as the luminance at which the pixel luminance value of the captured image is saturated. Therefore, the lamp power of the near-infrared projector 10 is adjusted so that the luminance in the captured image of the pedestrian existing within a predetermined distance is saturated.

  Further, in the present embodiment, as shown in FIG. 2, the predetermined distance L (m) is the collision prediction time (TTC (Time To Collision) = distance L with respect to the object / vehicle speed v, hereinafter referred to as “TTC”). The distance was determined to be 4 seconds. That is, the distance L at which the brightness of the pedestrian is saturated changes according to the vehicle speed v (m / sec). The predetermined distance L is far when the vehicle speed v is high, and the predetermined distance L is close when the vehicle speed v is low. As a result, the control unit 21 adjusts the lamp power of the near-infrared projector 10 so that the brightness of a pedestrian whose TTC is present at a distance of less than 4 seconds is saturated. The reason why TTC is set to 4 seconds is that, in general, TTC is preferably 4 seconds or more for alerting the driver with sufficient margin, but the predetermined distance is TTC 4 seconds. It goes without saying that the distance is not limited to.

  Here, the relationship between the vehicle speed and the lamp power of the near-infrared projector 10 (lamp power characteristics) will be described with reference to the four-quadrant graph of FIG. The first quadrant of the four-quadrant graph shows the relationship between the vehicle speed v (m / sec) and the distance L (m) to the pedestrian (object) when TTC is 4 seconds. As shown in the first quadrant, the distance L increases as the vehicle speed v increases.

  In the second quadrant, the relationship between the distance L to the pedestrian and the luminance (%) of the pedestrian is shown. As shown in the second quadrant, the luminance of the pedestrian decreases as the distance L to the pedestrian increases. FIG. 4 and FIG. 5 show the brightness of a person 50 meters ahead and the brightness of a person 100 meters ahead, respectively. Here, when the lamp power of the near-infrared projector 10, the sensitivity of the near-infrared camera 12, and the shutter speed are determined in advance, the luminance in the captured image of the pedestrian is unique according to the distance to the pedestrian. Is determined. Therefore, by adjusting the lamp power of the near-infrared projector 10, the sensitivity of the near-infrared camera 12, and the shutter speed, the saturation luminance and the distance at that time can be adjusted. Hereinafter, a case where the lamp power of the near-infrared projector 10 is mainly adjusted will be described as an example. However, in addition to or instead of the lamp power of the near-infrared projector 10, the sensitivity and shutter speed (exposure time) of the near-infrared camera 12 are set. It is good also as a structure to adjust.

  In the third quadrant, the relationship between the brightness of the pedestrian and the lamp power P (%) of the near-infrared projector 10 is shown. As shown in the third quadrant, the lamp power P is adjusted to increase as the pedestrian brightness decreases. In the region where the lamp power P exceeds 100 (%), the brightness of the captured image is adjusted by increasing the sensitivity of the near-infrared camera 12 or decreasing the shutter speed.

  The fourth quadrant shows the relationship between the lamp power P and the vehicle speed v (lamp power characteristics). As shown in the fourth quadrant, the lamp power P is adjusted to increase as the vehicle speed v increases. That is, the control unit 21 drives the driver circuit 11 based on the lamp power characteristic shown in the fourth quadrant, whereby the luminance in the captured image of the pedestrian existing within a distance (within a predetermined distance) of less than TTC 4 seconds. The lamp power of the near infrared projector 10 is adjusted so as to be saturated.

  The recognizing unit 22 extracts a pedestrian from the near-infrared image and recognizes whether or not the extracted pedestrian is a pedestrian to be alerted in consideration of the luminance of the pedestrian. The method for extracting pedestrians is not particularly limited, and various methods can be applied. As an extraction method, for example, there is a method of preparing a pedestrian template and pattern matching using the template. Specifically, a rectangular area of a predetermined size is sequentially cut out from the near-infrared image, and the degree of matching between the template image and each cut out rectangular area image is obtained. It is determined that the person is a pedestrian.

  The recognizing unit 22 determines that a pedestrian whose luminance in the captured image is saturated exists within a distance (within a predetermined distance) of less than TTC 4 seconds, and detects the pedestrian from an alert (warning) target. exclude. On the other hand, a pedestrian whose luminance in the captured image is not saturated is determined to be present at a distance of TTC 4 seconds or longer, and the pedestrian is recognized as an alerting target. The recognition result is transmitted to the alerting device 30 as a display signal.

  Here, a method for setting a pixel luminance value, that is, a saturation luminance that is determined to be saturated, will be described with reference to FIG. FIG. 6 is a graph showing the relationship between the distance from the pedestrian and the pixel luminance value. The horizontal axis in FIG. 6 is the distance (m) from the pedestrian, and the vertical axis is the luminance value (%) of the pixel. is there. As indicated by a thin line in FIG. 6, in a region where the distance to the pedestrian is 0 to 30 (m), the pixel luminance value increases as the distance increases, and the distance to the pedestrian is 30 to 200 (m). In the area), the pixel luminance value decreases as the distance increases. When the lamp power of the near-infrared projector 10 is increased, the graph moves in a direction in which the pixel luminance value increases. Conversely, when the lamp power is decreased, the pixel luminance value decreases. The graph moves in parallel.

  In FIG. 6, the saturation luminance (80%) at TTC 4 seconds and vehicle speed 60 (km / h) is shown superimposed on the graph (see the thick line in FIG. 6). That is, when the vehicle speed is 60 (km / h), the pixel luminance value (80%) at the distance 66.7 (m) where the TTC is 4 seconds is set as the saturation luminance. In this case, since all the pedestrians within the distance of 66.7 (m) are saturated in luminance, by determining whether the luminance is saturated, the distance is less than TTC 4 seconds (66.7 m). Existing pedestrians can be excluded from the alerting target. Therefore, it becomes possible to detect only a pedestrian whose TTC is present at a distance of 4 seconds or more and call attention. In this way, if a pedestrian who is present at a distance of less than 4 seconds is excluded from the alerting target, the possibility that the pedestrian is already within a short distance may fall within the irradiation range of the low beam of the headlight. This is because it is preferable to look directly at the pedestrian rather than directing the driver's line of sight toward the alerting device 30.

  When the TTC is set to 4 seconds or more, the saturation luminance is set to a value lower than 80 (%). Conversely, when the TTC is set to less than 4 seconds, the saturation luminance is from 80 (%). Is also set to a high value. As described above, the lamp power of the near-infrared projector 10 is adjusted according to the vehicle speed, but the saturation luminance may be changed according to the vehicle speed.

  The alerting device 30 presents a near-infrared image and pedestrian information recognized from the near-infrared image to the driver, and is configured using, for example, a liquid crystal display or a head-up display. The alerting device 30 receives a display signal from the ECU 20 and displays an image indicated by the display signal. Examples of the display image of the pedestrian information include an image in which a pedestrian recognized in the captured near-infrared image is highlighted by being surrounded by a rectangular frame.

  Next, the operation of the object recognition device applied to the near infrared night vision device 1 will be described with reference to FIGS. FIG. 7 is a flowchart showing a processing procedure of object recognition processing by the object recognition apparatus according to the first embodiment. FIG. 8 and FIG. 9 are flowcharts showing the processing procedure of the lamp power changing process and the saturated object excluding process executed in the object recognition process. This object recognition process is performed by the ECU 20, and is repeatedly executed at a predetermined timing from when the near-infrared night vision apparatus 1 is turned on until it is turned off.

  In step S100, a determination is made as to whether or not the initial setting of the lamp power of the near-infrared projector 10 has been completed. If the lamp power of the near-infrared projector 10 has already been initialized, the process proceeds to step S104. On the other hand, when the initial setting of the lamp power of the near-infrared projector 10 has not been completed yet, 100 (%) is set as the initial value of the lamp power in step S102. That is, the near-infrared projector 10 irradiates near-infrared rays in front of the vehicle with an output of 100 (%). Thereafter, the process proceeds to step S104.

  In step S104, the vehicle speed obtained based on the wheel speed detected by the vehicle speed sensor 14 is read. In the subsequent step S106, the lamp power changing process of the near infrared projector 10 is executed. Here, the lamp power changing process will be described with reference to FIG.

  In step S200, the previous value Pn-1 of the lamp power of the near-infrared projector 10 is temporarily stored in the memory. In the subsequent step S202, the lamp power correction amount Pp is obtained based on the vehicle speed v and the saturation luminance Const set in accordance with TTC (4 seconds in the present embodiment).

  Subsequently, in step S204, the lamp power correction value Pp obtained in step S202 is added to the lamp power previous value Pn−1 stored in step S200, and the lamp power current value Pn is obtained. Then, the driver circuit 11 is controlled by the control signal generated based on the obtained lamp power current value Pn, and the near-infrared projector 10 is driven. In this embodiment, since the TTC is set to 4 seconds, the relationship between the vehicle speed v and the current lamp power value Pn is the lamp power characteristic shown in the fourth quadrant of the above-described four-quadrant graph. After the process in step S204 is completed, the process proceeds to step S108 in FIG.

  In step S106, a captured image is read from the near-infrared camera 12, and a pedestrian is extracted from the captured image by pattern matching using a template as described above. In the subsequent step S110, a saturation target exclusion process is performed in which a pedestrian whose distance is close and whose luminance is saturated is excluded from the alert target. Here, the saturation target exclusion process will be described with reference to FIG.

  In step S300, the pedestrian extracted in step S106 is read. In subsequent step S302, the luminance of the pedestrian read in step S300 is obtained. Here, it is preferable to use the average value of the luminance of the pedestrian pixels as the luminance of the pedestrian. However, a representative point may be selected and its luminance may be set as the luminance of the pedestrian.

  Subsequently, in step S304, a determination is made as to whether or not the luminance of the pedestrian obtained in step S302 is higher than the saturation luminance Const. Since the saturation luminance Const is as described above, the description thereof is omitted here. Here, when the luminance of the pedestrian is higher than the saturated luminance Const, it is determined in step S306 that the pedestrian is at a distance of less than TTC 4 seconds, and is excluded from the alerting target. Thereafter, the process proceeds to step S308. On the other hand, when the luminance of the pedestrian is equal to or lower than the saturation luminance Const, it is determined that the pedestrian is at a distance of TTC 4 seconds or longer and is recognized as a pedestrian to be alerted. Therefore, the process proceeds to step S308 without performing the exclusion process in step S306.

  In step S308, a determination is made as to whether or not all pedestrians have been checked for whether or not they are to be alerted. Here, when the check about all the pedestrians is complete | finished, a process transfers to step S112 shown by FIG. On the other hand, when there are still pedestrians that have not been checked, the process proceeds to step S300, and the processes of steps S300 to S308 are repeatedly executed until the check is completed for all pedestrians.

  After determining whether or not all pedestrians are to be alerted, in step S112, it is determined whether or not the pedestrian to be alerted has been recognized. Here, when the pedestrian to be alerted is not recognized, the process temporarily exits without alerting the driver. On the other hand, when a pedestrian to be alerted is recognized, the process proceeds to step S114.

  In step S114, the alert device 30 alerts the driver, for example, by highlighting the recognized pedestrian in a rectangular frame.

  According to the present embodiment, the lamp power of the near-infrared projector 10 is set so that the luminance is saturated (predetermined luminance) at a distance (predetermined distance) where TTC is 4 seconds. In this way, since all the pedestrians existing within a distance of TTC 4 seconds are saturated in luminance, the luminance in the captured image of the pedestrian is taken into consideration, and the distance to the pedestrian is not directly measured. It is possible to determine whether the pedestrian is present at a distance of less than TTC 4 seconds or at a distance of TTC 4 seconds or more.

  Moreover, according to this embodiment, since the lamp power of the near-infrared projector 10 is adjusted according to the vehicle speed, the distance from the pedestrian whose brightness is saturated can be adjusted according to the vehicle speed. As a result, it is possible to select and recognize only a pedestrian whose time to reach a pedestrian to be alerted, that is, a TTC of 4 seconds or more.

  For example, when the pedestrian is already at a short distance, the possibility of entering the low beam irradiation range of the headlight is high, and the pedestrian is directly observed rather than directing the driver's line of sight toward the alerting device 30. It is preferable to do so. According to this embodiment, since the pedestrian recognized that the TTC is present at a distance of less than 4 seconds is excluded from the alerting target, the TTC is 4 without directly measuring the distance to the pedestrian. It is possible to recognize and call only pedestrians that exist at a distance of more than a second, that is, pedestrians that really need attention.

  The brightness in the captured image of the pedestrian varies depending on the sensitivity of the near infrared camera 12 and the shutter speed (exposure time). According to this embodiment, since the lamp power of the near-infrared projector 10 can be adjusted according to the sensitivity of the near-infrared camera 12 and the exposure time, it is possible to more appropriately adjust the luminance in the captured image of the pedestrian. It becomes possible.

  According to the embodiment, the lamp power of the near-infrared projector 10, the sensitivity of the near-infrared camera 12, and the shutter speed are combined so that the luminance in the captured image of the pedestrian existing at a distance of less than 4 seconds is a saturated luminance. Can be adjusted. Therefore, it is possible to increase the degree of freedom in controlling brightness adjustment.

[Second Embodiment]
Next, the configuration of the object recognition apparatus according to the second embodiment will be described with reference to FIG. FIG. 10 is a block diagram showing a configuration of the near-infrared night-vision device 2 to which the object recognition device according to the second embodiment is applied. In FIG. 2, the same or equivalent components as those of the near infrared night vision apparatus 1 are denoted by the same reference numerals.

  The near-infrared night vision device 2 differs from the near-infrared night vision device 1 in that an ECU 24 is provided instead of the ECU 20. This ECU 24 replaces the control unit 21 that adjusts the lamp power of the near-infrared projector 10 so that the luminance in the captured image of the pedestrian away from the predetermined distance becomes the predetermined luminance, in the captured image of the pedestrian away from the predetermined distance. The ECU 20 differs from the ECU 20 described above in that it includes a correction unit 25 that corrects the luminance in the captured image so that the luminance of the image becomes a predetermined luminance. The correcting unit 25 functions as correcting means described in the claims.

  In the present embodiment, the predetermined luminance is determined as the luminance at which the pixel luminance value of the captured image is saturated. Therefore, the brightness of the captured image is adjusted so that the brightness in the captured image of the pedestrian existing within a predetermined distance is saturated. In the present embodiment, the predetermined distance is determined as a distance at which TTC is 4 seconds. That is, the distance at which the luminance of the pedestrian is saturated changes according to the vehicle speed, and the predetermined distance becomes far when the vehicle speed is high, and the predetermined distance becomes close when the vehicle speed is low. As a result, the correction unit 25 adjusts the brightness of the captured image so that the brightness of a pedestrian who has a TTC of less than 4 seconds is saturated.

  Moreover, in the near-infrared night vision apparatus 1 mentioned above, the near-infrared projector 10 was driven by the driver circuit 11 controlled by the control signal from ECU20 (control part 21), but in the near-infrared night vision apparatus 2, Without such a configuration, the near-infrared projector 10 is driven with a predetermined constant output (for example, 100%).

  Other configurations are the same as or similar to the above-described near-infrared night-vision device 1, and the description thereof is omitted here.

  Next, the operation of the object recognition device applied to the near infrared night vision device 2 will be described with reference to FIGS. FIG. 11 is a flowchart illustrating a processing procedure of object recognition processing by the object recognition apparatus according to the second embodiment. FIG. 12 is a flowchart showing the processing procedure of the video brightness change process executed in the object recognition process. This object recognition process is performed by the ECU 20, and is repeatedly executed at a predetermined timing from when the near-infrared night vision apparatus 2 is turned on until it is turned off.

  The operation of the object recognition device applied to the near-infrared night vision device 2 is that a video brightness change process (step S406 in FIG. 11) is executed instead of the lamp power change process (step S106 in FIG. 7) described above. Unlike the first embodiment, the other processing contents (steps S400 to S404, S408 to S414) are the same as or similar to the above-described processing contents (steps S100 to S104, S108 to S114), and thus the description thereof is omitted here. To do.

  A specific processing procedure of the video brightness changing process is shown in the flowchart of FIG. Next, the video luminance change process will be described with reference to FIG.

  In step S500, the video brightness B before correction is temporarily stored in the memory. In subsequent step S502, the luminance correction amount Bp is obtained based on the saturation luminance Const set according to the vehicle speed v and TTC (4 seconds in the present embodiment).

  Subsequently, in step S504, the luminance correction amount Bp obtained in step S502 is added to the video luminance B before correction stored in step S500, and the corrected video luminance B is obtained. After the process in step S504 is completed, the process proceeds to step S408 in FIG. Note that, as described above, the processing contents in steps S408 to S414 are the same as or similar to the processing contents in steps S108 to S114, and thus the description thereof is omitted here.

  According to the present embodiment, the brightness of the captured image is corrected so that the brightness is saturated (becomes predetermined brightness) at a distance (predetermined distance) at which TTC is 4 seconds. In this way, all captured images of pedestrians within a distance within 4 seconds of TTC are saturated in luminance, so the distance from the pedestrian can be directly measured by taking into account the corrected pedestrian luminance. Therefore, it is possible to determine whether the pedestrian is present at a distance of less than TTC 4 seconds or at a distance of TTC 4 seconds or more.

  Moreover, according to this embodiment, since the brightness | luminance correction amount is adjusted according to a vehicle speed, the distance with the pedestrian where brightness | luminance is saturated can be adjusted according to a vehicle speed. As a result, it is possible to select and recognize only a pedestrian whose time to reach a pedestrian to be alerted, that is, a TTC of 4 seconds or more. In addition, the same effects as those of the first embodiment described above can be obtained.

  Although the embodiment of the present invention has been described above, the present invention is not limited to the above embodiment, and various modifications can be made. For example, in the above embodiment, the present invention is applied to object recognition based on a near-infrared image captured by a near-infrared camera, but also for object recognition based on an invisible image captured by an invisible camera other than near-infrared or a visible image captured by a visible camera. Applicable.

  Moreover, in the said embodiment, although applied to the near-infrared night vision apparatus mounted in a vehicle, it is applicable to various uses other than a near-infrared night vision apparatus, without being limited to a vehicle-mounted apparatus.

  Moreover, in the said embodiment, although the pedestrian was made into the recognition target object, it cannot be overemphasized that a recognition target object is not restricted to a pedestrian.

It is a block diagram which shows the structure of the near-infrared night vision apparatus to which the object recognition apparatus which concerns on 1st Embodiment was applied. It is a figure for demonstrating the near-infrared irradiation distance of a near-infrared night vision apparatus. It is a four-quadrant graph showing the relationship between the vehicle speed and the distance to the object, the distance to the object and the object brightness, the object brightness and the lamp power, and the vehicle speed and the lamp power. It is a figure which shows the brightness | luminance of the person who is 50m ahead. It is a figure which shows the brightness | luminance of the person who is 100m ahead. It is a graph which shows the relationship between the distance with a person, and a pixel luminance value. It is a flowchart which shows the process sequence of the object recognition process by the object recognition apparatus which concerns on 1st Embodiment. It is a flowchart which shows the process sequence of the lamp power change process performed in the object recognition process shown by the flowchart of FIG. It is a flowchart which shows the process sequence of the saturation target exclusion process performed in the object recognition process shown by the flowchart of FIG. It is a block diagram which shows the structure of the near-infrared night vision apparatus to which the object recognition apparatus which concerns on 2nd Embodiment was applied. It is a flowchart which shows the process sequence of the object recognition process by the object recognition apparatus which concerns on 2nd Embodiment. It is a flowchart which shows the process sequence of the image | video brightness | luminance change process performed in the object recognition process shown by the flowchart of FIG.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1, 2 ... Near-infrared night vision apparatus, 10 ... Near-infrared projector, 11 ... Driver circuit, 12 ... Near-infrared camera, 14 ... Vehicle speed sensor, 20, 24 ... ECU, 21 ... Control part, 22 ... Recognition part, 25 ... Correction unit, 30...

Claims (8)

Light irradiation means for irradiating light;
Control means for adjusting the output of the light irradiation means;
A single imaging means for imaging a region irradiated with light by the light irradiation means;
Recognizing means for recognizing an object from an image picked up by the image pickup means,
The control means adjusts the output of the light irradiation means so that the brightness in the captured image of the object separated by a predetermined distance becomes a predetermined brightness,
The object recognition apparatus characterized in that the recognition means recognizes an object in consideration of luminance in a captured image of the object.   The object recognition apparatus according to claim 1, wherein the control unit adjusts the output of the light irradiation unit so that luminance in a captured image of an object existing within the predetermined distance is saturated.   The object recognition apparatus according to claim 2, wherein the recognition unit recognizes that an object whose luminance in the captured image is saturated exists within the predetermined distance.   The object recognition apparatus according to claim 3, wherein the recognition unit excludes an object recognized as existing within the predetermined distance from a recognition object.   5. The object recognition apparatus according to claim 1, wherein the control unit adjusts an output of the light irradiation unit according to a sensitivity and / or an exposure time of the imaging unit.   The object recognition apparatus according to claim 1, wherein the control unit adjusts an output of the light irradiation unit in accordance with a moving speed of the object recognition apparatus. Light irradiation means for irradiating light;
A single imaging means for imaging a region irradiated with light by the light irradiation means;
Control means for adjusting the output of the light irradiation means and the sensitivity and exposure time of the imaging means;
Recognizing means for recognizing an object from an image picked up by the image pickup means,
The control unit adjusts at least one of the output of the light irradiation unit, the sensitivity of the imaging unit, and the exposure time so that the luminance in the captured image of the object separated by a predetermined distance becomes a predetermined luminance,
The object recognizing apparatus, wherein the recognizing unit recognizes an object in consideration of luminance in a captured image of the object. Light irradiation means for irradiating light;
A single imaging means for imaging a region irradiated with light by the light irradiation means;
Correction means for correcting the brightness of the image captured by the imaging means;
Recognizing means for recognizing an object from the image corrected by the correcting means,
The correction unit corrects the luminance of the captured image so that the luminance in the captured image of the object that is separated by a predetermined distance becomes the predetermined luminance,
The object recognition apparatus characterized in that the recognition means recognizes an object in consideration of luminance in a captured image of the object after correction.