JP2014174101A - Object detection device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an object detection device capable of acquiring a distance in a wide range while suppressing arithmetic quantity.SOLUTION: The object detection device includes: a projection part 100 for projecting light to a target region; an imaging part 200 for imaging the target region by a CMOS image sensor 240; an imaging signal processing circuit 23 for acquiring the pixel value of a predetermined pixel on the CMOS image sensor 240; and an object detection part 21c for detecting an object to be detected existing in the target region on the basis of the luminance value acquired by the imaging signal processing circuit 23. The object detection part 21c acquires a measured area Sb of a measured shape J of the object to be detected in measurement, and acquires the distance information of the object to be detected on the basis of the measured area Sb of the measured shape J and a reference area Sa of a reference shape H of the object to be detected when the object to be detected is within a reference distance D.

Description

  The present invention relates to an object detection device that detects an object in a target area.

  Conventionally, object detection devices using light have been developed in various fields. An object detection apparatus using a so-called distance image sensor can detect not only a planar image on a two-dimensional plane but also the shape and movement of the detection target object in the depth direction.

  As a distance image sensor, a distance image sensor of a type that irradiates a target region with laser light having a predetermined dot pattern is known (for example, Non-Patent Document 1). In such a distance image sensor, a dot pattern when the reference surface is irradiated with laser light is picked up by the image pickup device, and the picked-up dot pattern is held as a reference dot pattern. Then, the reference dot pattern is compared with the actually measured dot pattern captured at the time of actual measurement, and distance information is acquired. Specifically, distance information with respect to the reference region is acquired by a triangulation method based on the position of the reference region set on the standard dot pattern on the measured dot pattern.

19th Annual Conference of the Robotics Society of Japan (September 18-20, 2001) Proceedings, P1279-1280

  However, since the distance image sensor needs to generate light of a dot pattern, there is a problem that the configuration of a projection unit that projects light onto a target area becomes complicated. In addition, since it is necessary to search for dots included in the reference area on the measured dot pattern, there is a problem that complicated calculation processing is required for obtaining distance information.

  Furthermore, when an object is brought close to the image sensor so that the output signal of the image sensor is saturated, a high brightness value is output uniformly in the captured image, and thus the distance cannot be acquired properly. Problems also arise.

  In view of the above problems, an object of the present invention is to provide an object detection device capable of acquiring a distance over a wide range while suppressing a calculation amount.

  A main aspect of the present invention relates to an object detection apparatus. The object detection apparatus according to this aspect includes a projection unit that projects light onto a target area, an imaging unit that captures the target area with an image sensor, and a luminance acquisition unit that acquires a luminance value of a predetermined pixel on the image sensor. And an object detection unit that detects a detection target object existing in the target area based on the luminance value acquired by the luminance acquisition unit. The object detection unit obtains the size of the actually measured shape of the detection target object during actual measurement, and the size of the actually measured shape and the size of the reference shape of the detection target object when the detection target object is at a reference distance Based on the information, information on the distance of the detection target object is acquired.

  ADVANTAGE OF THE INVENTION According to this invention, the object detection apparatus which can acquire distance over a wide range can be provided, suppressing the amount of calculations.

  The effects and significance of the present invention will become more apparent from the following description of embodiments. However, the embodiment described below is merely an example when the present invention is implemented, and the present invention is not limited to the following embodiment.

It is a figure which shows the external appearance structure of the personal computer which incorporated the object detection apparatus which concerns on embodiment. It is a figure which shows the structure of the object detection apparatus which concerns on embodiment. It is a figure which shows the structure of the projection part and imaging part which concern on embodiment. It is a figure which shows the sensitivity of the image sensor which concerns on embodiment. It is a figure which shows the relationship between the waveform of the distance conversion function which prescribes | regulates the relationship between the luminance value which concerns on embodiment, and the magnitude | size of the detection target object reflected in the captured image. It is a figure explaining the setting method of the reference | standard shape registration table which concerns on embodiment. It is a flowchart which shows the reference | standard shape registration process which concerns on embodiment. It is a flowchart which shows the object detection process which concerns on embodiment. It is a figure explaining the setting method of the reference | standard shape registration table which concerns on the example of a change. It is a figure explaining the setting method of the reference | standard shape registration table which concerns on the example of a change.

  Embodiments of the present invention will be described below with reference to the drawings.

  In this embodiment, the object detection apparatus according to the present invention is applied to a notebook personal computer. In addition, the object detection apparatus according to the present invention can be appropriately applied to other devices such as a desktop personal computer and a television. Note that the object detection device according to the present invention does not necessarily have to be integrally mounted on other devices, and may constitute a single device alone.

  In the embodiment described below, the imaging signal processing circuit 23 corresponds to a “luminance acquisition unit” recited in the claims. The CMOS image sensor 240 corresponds to an “image sensor” recited in the claims. The distance conversion function shown in FIG. 5A corresponds to “regulation information” described in the claims. The distance information acquired by the object detection unit 21c corresponds to “information regarding the distance of the object at the time of actual measurement” described in the claims. The average luminance Ba corresponds to a “luminance value” recited in the claims. The reference area Sa corresponds to “a size of the reference shape” recited in the claims. The actual measurement area Sb corresponds to “the size of the actual measurement shape” recited in the claims. However, the description of the correspondence between the above claims and the present embodiment is merely an example, and the invention according to the claims is not limited to the present embodiment.

  FIG. 1 is a diagram showing a schematic configuration of a personal computer 1 according to the present embodiment. As shown in FIG. 1, the personal computer 1 includes an object detection device 2 and an information processing unit 3. In addition, the personal computer 1 includes a keyboard 4, an operation pad 5, and a monitor 6.

The object detection device 2 projects light (infrared light) in an infrared wavelength band longer than the visible light wavelength band over the entire target region, and receives the reflected light with a CMOS image sensor, thereby achieving the target. A distance to an object existing in the region (hereinafter referred to as “distance information”) is acquired. Based on the distance information acquired by the object detection device 2, the information processing unit 3 detects a predetermined object existing in the target area, and further detects the movement of the object. Then, the information processing unit 3 controls the function of the personal computer 1 according to the movement of the object.

  For example, when a user performs a predetermined gesture using a hand, distance information corresponding to the gesture is transmitted from the object detection device 2 to the information processing unit 3. Based on this information, the information processing unit 3 detects the user's hand as a detection target object, and functions associated with the movement of the hand (screen enlargement / reduction, screen brightness adjustment, page turning, etc.) Execute.

  FIG. 2 is a diagram illustrating configurations of the object detection device 2 and the information processing unit 3.

  The object detection device 2 includes a projection unit 100 and an imaging unit 200 as the configuration of the optical unit. The projection unit 100 and the imaging unit 200 are arranged so as to be aligned in the X-axis direction.

  The projection unit 100 includes a light source 110 that emits light in an infrared wavelength band.

  The imaging unit 200 includes an aperture 210, an imaging lens 220, a filter 230, and a CMOS image sensor 240. In addition, the object detection apparatus 2 includes a CPU (Central Processing Unit) 21, an infrared light source driving circuit 22, an imaging signal processing circuit 23, an input / output circuit 24, and a memory 25 as a circuit unit. Yes.

  The configuration in which the projection unit 100 includes the light source 110 that emits light in the infrared wavelength band and the imaging unit 200 includes the imaging lens 220 and the filter 230 is an example of the configuration according to claim 7. It is.

  The light projected from the light source 110 onto the target area is reflected by an object existing in the target area and enters the imaging lens 220 via the aperture 210.

  The aperture 210 limits light from the outside so as to match the F number of the imaging lens 220. The imaging lens 220 collects the light incident through the aperture 210 on the CMOS image sensor 240. The filter 230 is a bandpass filter that transmits light in the wavelength band including the emission wavelength of the light source 110 and cuts light in the visible wavelength band.

  As will be described later, the CMOS image sensor 240 has sensitivity to the wavelength band of infrared light emitted from the light source 110. The CMOS image sensor 240 receives the light collected by the imaging lens 220 and outputs a signal corresponding to the amount of received light to the imaging signal processing circuit 23 for each pixel. Here, in the CMOS image sensor 240, the output speed of the signal is increased so that the signal of the pixel can be output to the imaging signal processing circuit 23 with high response from the light reception in each pixel.

  In the present embodiment, the effective imaging area of the CMOS image sensor 240 (area where signals are output as a sensor) is, for example, the size of VGA (640 horizontal pixels × 480 vertical pixels). The imaging effective area of the CMOS image sensor 240 may have other sizes such as an XGA (horizontal 1024 pixels × vertical 768 pixels) size or an SXGA (horizontal 1280 pixels × vertical 1024 pixels) size.

  The CPU 21 controls each unit according to a control program stored in the memory 25. With this control program, the CPU 21 is provided with functions of the light source control unit 21a, the reference shape acquisition unit 21b, and the object detection unit 21c.

The light source control unit 21 a controls the infrared light source driving circuit 22. The reference shape acquisition unit 21b acquires a reference shape and a reference area for acquiring distance information of the detection target object based on a signal output from the CMOS image sensor 240 at the time of presetting, and a reference shape registration table in the memory 25 Register with T. The reference shape registration process performed by the reference shape acquisition unit 21b will be described later with reference to FIG.

  The infrared light source driving circuit 22 drives the light source 110 according to a control signal from the CPU 21.

  The imaging signal processing circuit 23 drives the CMOS image sensor 240 under the control of the CPU 21, acquires the luminance signal of each pixel from the signal output from the CMOS image sensor 240, and outputs the acquired luminance signal to the CPU 21. To do. As will be described later, the imaging signal processing circuit 23 applies the exposure time set by the CPU 21 to each pixel of the CMOS image sensor 240, and further sets the CPU 21 for the signal output from the CMOS image sensor 240. A gain signal is applied to obtain a luminance signal for each pixel.

  The CPU 21 detects the detection target object of the reference shape registered in the reference shape registration table T based on the luminance signal supplied from the imaging signal processing circuit 23, and determines the distance from the object detection device 2 to the detection target object. The calculation is performed by the processing by the object detection unit 21c. The distance information is acquired for each pixel of the CMOS image sensor 240. The object detection unit 21 c detects the movement of the detection target object based on the acquired distance information, and outputs a function control signal corresponding to a predetermined movement pattern to the CPU 31 of the information processing unit 3. The object detection processing by the object detection unit 21c will be described in detail later with reference to FIG.

  The input / output circuit 24 controls data communication with the information processing unit 3.

  In addition to the control program executed by the CPU 21, the memory 25 holds a distance conversion function used for acquiring distance information. In addition, the memory 25 is also used as a work area during processing in the CPU 21. The distance conversion function will be described later with reference to FIG.

  The information processing unit 3 includes a CPU 31, an input / output circuit 32, and a memory 33. In addition to the configuration shown in FIG. 2, the information processing unit 3 is provided with a configuration for driving and controlling each unit of the personal computer 1. For convenience, the configuration of these peripheral circuits is not shown.

  The CPU 31 controls each unit according to a control program stored in the memory 33. With this control program, the function of the function control unit 31a for controlling the function of the personal computer 1 is given to the CPU 31 in accordance with a signal from the object detection unit 21c.

  The function control unit 31a controls the function of the personal computer 1 according to the function control signal from the object detection unit 21c.

  FIG. 3A and FIG. 3B are diagrams illustrating configurations of the projection unit 100 and the imaging unit 200. 3A is a plan view showing the projection unit 100 and the imaging unit 200 installed on the circuit board 300, and FIG. 3B is a cross-sectional view taken along line AA ′ of FIG. is there.

  Referring to FIGS. 3A and 3B, the light source 110 includes a light emitting diode (LED), a semiconductor laser, and the like. The light source 110 is adjusted by a lens or the like so that infrared light gathers in the vicinity of the imaging range in the target area in order to increase the use efficiency of infrared light. The light source 110 is mounted on the circuit board 300.

The imaging unit 200 includes a lens barrel 250 and an imaging lens holder 260 in addition to the aperture 210, the imaging lens 220, the filter 230, and the CMOS image sensor 240 described above. The CMOS image sensor 240 is mounted on the circuit board 300. The imaging lens 220 is attached to the lens barrel 250, and the lens barrel 250 is attached to the imaging lens holder 260 while holding the imaging lens 220. The imaging lens holder 260 has a recess on the lower surface, and the filter 230 is attached to the recess. In this way, the imaging lens holder 260 is installed on the circuit board 300 so as to cover the CMOS image sensor 240 while holding the imaging lens 220 and the filter 230.

  In the present embodiment, the imaging lens 220 is composed of four lenses. However, the number of lenses constituting the imaging lens 220 is not limited to this, and the imaging lens 220 may be configured from other numbers of lenses.

  In addition to the projection unit 100 and the imaging unit 200, the circuit unit 400 constituting the object detection device 2 is mounted on the circuit board 300. The CPU 21, the infrared light source driving circuit 22, the imaging signal processing circuit 23, the input / output circuit 24 and the memory 25 shown in FIG. 2 are included in the circuit unit 400.

  FIG. 3C is a diagram schematically illustrating a projection state of infrared light on the target region and an imaging state of the target region by the imaging unit 200.

  In the lower part of FIG. 3C, an infrared light projection state by the projection unit 100 and a target region imaging state by the imaging unit 200 are shown. 3C schematically shows the infrared light projection range in the target area and the imaging range of the imaging unit 200 for the target area. Furthermore, an area corresponding to the effective imaging area of the CMOS image sensor 240 is shown in the upper part of FIG. In FIG. 3C, Lmin indicates the minimum distance that can be acquired by the object detection device 2. In the upper part of FIG. 3C, a projection range, an imaging range, and an effective imaging area when the target area is at the maximum distance are shown.

  As shown in FIG. 3C, the imaging range and the projection range overlap each other in the target area, and the imaging effective region is positioned in a range where the imaging range and the projection range overlap.

  In order to image the target area with the CMOS image sensor 240, an area corresponding to the effective imaging area of the CMOS image sensor 240 needs to be included in the projection range. On the other hand, as the distance becomes shorter, the projection range becomes narrower and eventually the projection range does not cover the area corresponding to the imaging effective area. Therefore, the minimum distance Lmin is set to be longer than at least the minimum limit distance that the projection range can cover the entire imaging effective area.

  FIG. 4 is a diagram schematically showing the sensitivity of each pixel on the CMOS image sensor 240.

  In the present embodiment, a color sensor is used as the CMOS image sensor 240. Therefore, the CMOS image sensor 240 includes three types of pixels that detect red, green, and blue, respectively.

In FIG. 4, R, G, and B indicate the sensitivity of red, green, and blue pixels included in the CMOS image sensor 240, respectively. As shown in FIG. 4, the sensitivity of the red, green, and blue pixels is substantially the same in the infrared wavelength band of 800 nm or more (see the hatched portion in FIG. 4). Therefore, when the wavelength band of visible light is removed by the filter 230 shown in FIG. 3B, the sensitivity of the red, green, and blue pixels of the CMOS image sensor 240 becomes substantially equal to each other. For this reason, when the same amount of infrared light is incident on the red, green, and blue pixels, the values of the signals output from the pixels of the respective colors are substantially equal. Therefore, there is no need to adjust the signal from each pixel between the pixels, and the signal from each pixel can be used as it is for obtaining distance information.

  A diagram schematically showing the waveform of the distance conversion function held in the memory 25 is shown in the upper part of FIG. A diagram schematically showing the relationship between the distance from the object detection device 2 to the detection target object and the luminance is shown in the lower part of FIG. In FIG. 5A, a hand is shown as the detection target object, and the darker the hatching, the greater the amount of received light. The vertical axis of the graph shown in FIG. 5A is the luminance value (tone value), and the horizontal axis is the distance (cm). It should be noted that hatching is added to a range of luminance values higher than the maximum luminance value 255 of the CMOS image sensor 240.

  As shown in FIG. 5A, the distance conversion function defines the relationship between the luminance value acquired via the CMOS image sensor 240 and the distance corresponding to the luminance value. In general, the amount of light traveling straight is attenuated in inverse proportion to the square of the distance. Accordingly, the infrared light emitted from the projection unit 100 is attenuated to one-quarter of the distance obtained by adding the distance from the projection unit 100 to the target area and the distance from the target area to the imaging unit 200. The light is received by the imaging unit 200. For this reason, as shown in FIG. 5A, the luminance value acquired via the CMOS image sensor 240 decreases as the distance to the object increases, and increases as the distance to the object decreases. Therefore, the distance conversion function that defines the relationship between the distance and the luminance has a curved waveform as shown in FIG.

  In the object detection device 2, the exposure time of the CMOS image sensor 240 and the gain applied to the acquisition of the brightness value are adjusted so that the relationship between the distance and the brightness value substantially matches the distance conversion function. The luminance value is acquired with 256 gradations.

  Such adjustment is performed when the object detection apparatus 2 is manufactured. The adjusted exposure time and gain are stored in the memory 25 and used when acquiring distance information.

  As described above, when a detection target object (for example, a hand) is brought closer to the object detection device 2, the amount of reflected light with respect to the CMOS image sensor 240 increases in inverse proportion to the square of the distance, and the luminance increases. Become. When the detection target object is brought close to the predetermined distance L0, the output of the CMOS image sensor 240 is saturated and the luminance becomes 255 which is the maximum value. When the object to be detected is brought closer to the object detection device 2 side than the distance L0, the output of the CMOS image sensor 240 is saturated, so that the luminance value increases even if the amount of received light increases according to the distance. Disappear.

  Further, when the detection target object is moved away from the object detection device 2, the amount of reflected light with respect to the CMOS image sensor 240 is decreased in inverse proportion to the square of the distance, and the luminance is decreased. When the detection target object is moved away to a predetermined distance Le, the amount of light received by the CMOS image sensor 240 decreases to the detection limit, and the luminance becomes 0, which is the minimum. When the object to be detected is further away from the object detection device 2 than the distance Le, the amount of received light is smaller than the detection limit of the CMOS image sensor 240. Therefore, even if the amount of received light decreases according to the distance, the luminance decreases. Disappear.

  As described above, when the object is positioned at a shorter distance than the distance L0, the luminance does not change, and therefore the distance cannot be measured from the change in luminance.

  FIG. 5B is a diagram illustrating the distance from the object detection device 2 to the detection target object and the size of the detection target object reflected in the captured image. In FIG. 5B, a hand is shown as the detection target object, and the brightness is higher as the hatching is thinner.

  FIG. 5B shows the size of the hand at a predetermined distance when the hand is gradually moved away from the short distance to the long distance in order from the left. As shown in the figure, when the hand is gradually moved away from the short distance, the size of the object shape reflected in the captured image is gradually reduced. The brightness of the hand shown on the leftmost side in the figure and the position of the hand on the right is the maximum value (255).

  In general, the size of the shape of the detection target object reflected in the captured image becomes smaller in inverse proportion to the square of the distance, like the luminance. Therefore, if the shape of the detection target object and the size of the detection target object at the predetermined reference distance D are held in advance, the size of the detection target object held based on the same distance conversion function as described above. The distance can be calculated from the ratio of the size of the detection target object at the time of actual measurement. In the present embodiment, the reference distance D is set to a distance L0 at which the luminance is approximately the maximum value.

  For example, as shown in FIG. 5A, parameters such as the gain and exposure time of the CMOS image sensor 240 are adjusted in advance so that the brightness reaches a maximum value of 255 at a position 30 cm from the object detection device 2. If the area of the detection target object when the luminance reaches the maximum value (255) is S, the distance is calculated to be approximately 25 cm when the area of the detection target object at the time of actual measurement is 1.5S. When the area of the detection target object at the time of actual measurement is 0.5S, it is calculated as approximately 45 cm.

  In this way, the distance at the time of actual measurement can be calculated from the ratio of the area of the detection target object at the reference distance D at which the luminance is substantially the maximum value and the area of the detection target object at the time of actual measurement. Since the detection target object reflected in the captured image changes in size even when the detection target object is brought closer to the object detection device 2 as the output value of the CMOS image sensor 240 is saturated, the distance is acquired using only the luminance value. Compared to the case, distance acquisition can be performed in a wider range.

  As described above, the configuration in which the distance of the detection target object is acquired based on the area of the detection target object at the reference distance D that is held in advance and the area of the detection target object at the time of actual measurement is claimed. It is an example of the structure of claim | item 1. Further, the configuration in which the reference distance D is set to a distance at which the luminance is approximately the maximum value is an example of a configuration according to claim 4.

  As described above, in order to calculate the distance from the area of the detection target object at the time of actual measurement and the distance conversion function, before the object detection, the reference shape H of the detection target object at the reference distance D at which the luminance is approximately the maximum value. The reference area Sa of the detection target object is set in the reference shape registration table T of the memory 25.

  FIG. 6A is a schematic diagram illustrating an example of a captured image output from the CMOS image sensor 240. FIG. 6B to FIG. 6E are diagrams illustrating examples of captured images around a person's hand. FIG. 6F shows the reference shape registration table T.

  In the captured image of FIG. 6A, a person with his hand protruding forward is reflected. The captured images in FIGS. 6A to 6E indicate that the darker the hatching, the lower the luminance, and the thinner the hatching, the higher the luminance.

At the time of setting the reference shape registration table T, first, as shown in FIGS. 6B to 6E, the imaging unit 200 captures an image of a person gradually approaching the object detection device 2 with his / her hand. . Then, the acquired captured image is divided into determination areas W1 to Wn having a predetermined size. The size of the determination areas W1 to Wn is set to be smaller than the size of the detection target object, as shown in FIG. When it is determined that any of the determination areas W1 to Wn includes a detection target object having a luminance that is approximately the maximum value, it is determined that the detection target object is positioned at the reference distance D. Then, the shape of the detection target object at that position is registered in the reference shape registration table T as the reference shape H, and the number of pixels included in the reference shape H is registered as the reference area Sa.

  For example, as shown in FIGS. 6B to 6E, when a person is picked up in a state where the hand gradually protrudes forward, the determination is made when the picked-up image shown in FIG. 6D is picked up. It is determined that the detection target object whose luminance is approximately the maximum value is included in the region Wi. Then, the shape of the detection target object including the determination area Wi is extracted by a predetermined contour extraction engine. Then, the number of pixels included in the extracted contour is counted, and the reference shape H and the reference area Sa are set in the reference shape registration table T of the memory 25 as shown in FIG.

  Thus, when the reference shape H and the reference area Sa are set in the reference shape registration table T, the shape of the detection target object in the captured image at the time of actual measurement is extracted, and the actual measurement area of the actual measurement shape J of the detection target object at the actual measurement time. From the ratio of Sb and the reference area Sa of the reference shape H, distance acquisition is performed based on the distance conversion function.

  The configuration in which the reference shape H and the reference area Sa of the detection target object are set as described above is an example of a configuration according to claim 2. Further, a configuration including a distance conversion function that defines a relationship between distances corresponding to luminance values, and in which the reference shape H and the reference area Sa of the detection target object when the luminance is substantially the maximum value is set in claim 3. It is an example of the structure of description. Furthermore, the configuration in which the distance of the detection target object at the time of actual measurement is acquired based on the ratio between the actual measurement area Sb of the shape of the detection target object at the time of actual measurement and the reference area Sa of the reference shape H and the distance conversion function. 6 is an example of a configuration described in 6.

  As shown in FIG. 6D, the sizes of the determination areas W1 to Wn are set to be smaller than the size of the detection target object at the position of the reference distance D where the luminance of the detection target object is approximately the maximum value. Is desirable. This is because the way the light strikes depends on the direction and shape of the finger or hand. For example, when the size of the determination areas W1 to Wn is substantially the same as the shape of the hand, it is necessary to adjust the direction and shape of the finger or hand on the user side so that the brightness of the entire hand becomes a substantially maximum value. There is. On the other hand, in the present embodiment, the sizes of the determination regions W1 to Wn are set smaller than the size of the detection target object, and therefore the brightness of the detection target object is appropriately set regardless of the shape and orientation of the detection target object. It is possible to determine a position at which becomes a substantially maximum value. Thereby, the reference shape H and the reference area Sa can be appropriately set in the reference shape registration table T.

  The configuration in which the determination areas W1 to Wn smaller than the reference area Sa of the reference shape H are set at the reference distance D where the luminance of the detection target object is substantially maximum as described above is described in claim 5. It is an example of a structure.

  FIG. 7 is a flowchart showing the reference shape registration process. The processing of FIG. 7 is executed by the function of the reference shape acquisition unit 21b among the functions of the CPU 21 shown in FIG.

  When the reference shape registration process is started and the imaging timing arrives (S101: YES), the CPU 21 reads the exposure time and gain set as described above from the memory 25 and sets them in the imaging signal processing circuit 23. Thereby, the imaging signal processing circuit 23 acquires a captured image from the CMOS image sensor 240 with the set exposure time and gain (S102), and acquires a luminance value for each pixel from the acquired captured image (S103). . The acquired luminance value is transmitted to the CPU 21.

  The CPU 21 classifies areas of a predetermined size on the captured image as determination areas W1 to Wn (S104). The determination areas W1 to Wn are set smaller than the size of the detection target object. For example, it is set to an area of 5 vertical pixels × 5 horizontal pixels.

  The CPU 21 sets 1 to the variable i (S105), and calculates the average luminance Ba in the determination area Wi (S106). Then, the CPU 21 determines whether or not the average luminance Ba of the determination area Wi is equal to or higher than the luminance threshold Bsh1 (S107).

  When the average brightness Ba of the determination area Wi is not equal to or higher than the brightness threshold Bsh1 (S107: NO), the CPU 21 adds 1 to the variable i (S108), and determines whether the variable i exceeds the variable n (S109). . When the variable i does not exceed the variable n (S109: NO), the CPU 21 returns the process to S106, calculates the average luminance Ba of the next determination area Wi (S106), and the average luminance Ba is equal to or higher than the luminance threshold Bsh1. It is determined whether or not there is (S107). As long as the average brightness Ba of the determination area Wi is not equal to or higher than the brightness threshold Bsh1 (S107: NO), the CPU 21 repeats the process (S106 to S109) until the final determination area Wn is reached (S109: NO).

  The luminance threshold value Bsh1 is set to a value close to the maximum luminance value (255) (for example, 250) in order to determine whether the output of the CMOS image sensor 240 is saturated.

  When the average brightness Ba of the determination areas W1 to Wn is not equal to or higher than the brightness threshold Bsh1 (S107: NO) and the process is completed up to the last determination area Wn (S109: YES), the CPU 21 determines whether or not the reference shape registration process is finished. Is determined (S110). When the reference shape registration process has not ended (S110: NO), the CPU 21 returns the process to S101 and executes a reference shape determination process for the next captured image (S101 to S113).

  When the average luminance Ba of the determination area Wi is equal to or higher than the luminance threshold Bsh1 (S107: YES), the CPU 21 extracts the outline of the detection target object including the determination area Wi by a predetermined outline extraction engine (S111). The contour extracted in this way is set as a reference shape H. Then, the number of pixels included in the contour of the reference shape H is counted (S112). The number of pixels counted in this way is set as a reference area Sa. Then, the CPU 21 sets the extracted reference shape H and reference area Sa in the reference shape registration table T (S113). As a result, the size (area) of the detection target object when the shape and brightness of the detection target object are approximately the maximum value (255) is set in the reference shape registration table T of the memory 25. .

  Thus, when the reference shape H and the reference area Sa are registered in the reference shape registration table T, the CPU 21 determines whether or not the reference shape registration process is completed (S110). If the reference shape registration process has not ended (S110: NO), the CPU 21 returns the process to S101 and executes the process for the next captured image (S101 to S113). When the reference shape registration process ends (S110: YES), the CPU 21 completes the process.

  FIG. 8 is a flowchart showing the object detection process. 8 is executed by the function of the object detection unit 21c among the functions of the CPU 21 shown in FIG.

  When the object detection process is started and the imaging timing comes (S201: YES), the CPU 21 reads the exposure time and gain set as described above from the memory 25 and sets them in the imaging signal processing circuit 23. Thereby, the imaging signal processing circuit 23 acquires a captured image from the CMOS image sensor 240 with the set exposure time and gain (S202), and acquires a luminance value for each pixel from the acquired captured image (S203). . The acquired luminance value is transmitted to the CPU 21.

The CPU 21 creates a luminance image by setting the luminance value received from the imaging signal processing circuit 23 at each pixel position (S204). Then, the CPU 21 sets a value obtained by subtracting the predetermined value ΔB from the luminance value of the highest gradation in the luminance image (the luminance value indicating the closest approach to the object detection device 2) as the luminance threshold Bsh2 (S205).

  Next, the CPU 21 classifies an area having a luminance value (gradation value) higher than the luminance threshold Bsh2 as a target area on the luminance image (S206). Then, the CPU 21 executes a predetermined contour extraction engine, compares the contour of the segmented target region with the contour of the reference shape H of the reference shape registration table T, and compares the contour target corresponding to the contour of the reference shape H. The region is extracted as a region corresponding to the detection target object (S207). This extracted area is taken as a measured shape J. The CPU 21 counts the number of pixels included in the contour of the extracted actual measurement shape J, and calculates the actual measurement area Sb of the detection target object (S208).

  The CPU 21 acquires distance information by calculation based on the distance conversion function held in the memory 25 according to the ratio between the calculated actual area Sb and the reference area Sa held in the reference shape registration table T (S209). Further, the CPU 21 acquires the pixel position at the center of the detection target object (S210). The acquired distance information and pixel position correspond to the three-dimensional distance information of the detection target object. The acquired three-dimensional distance information is held in the memory 25.

  The CPU 21 acquires the movement of the detection target object based on the acquired three-dimensional distance information and the three-dimensional distance information held in the memory 25 acquired so far (S211). Then, the CPU 21 determines whether or not the acquired movement of the detection target object matches a predetermined movement pattern set in advance (S212).

  When the movement of the detection target object does not match the predetermined movement pattern (S212: NO), the CPU 21 advances the process to S214. When the movement of the detection target object matches a predetermined movement pattern (S212: YES), the CPU 21 outputs a function control signal corresponding to the predetermined movement pattern to the information processing unit 3 of the personal computer 1 shown in FIG. When receiving the function control signal, the information processing unit 3 of the personal computer 1 executes the functions of the personal computer 1 (screen enlargement / reduction, screen brightness adjustment, page feed, etc.) according to the signal.

  Thus, when the distance acquisition and motion detection processing of the detection target object is completed, the CPU 21 determines whether or not the object detection operation is completed (S214). If the object detection operation has not ended (S214: NO), the CPU 21 returns to S201 and waits for the next captured image to be acquired. When the next captured image is acquired (S201: YES), the CPU 21 executes the processing from S202 onward, and executes the distance acquisition of the detection target object and the motion detection processing from the captured image (S202 to S213). . When the object detection process ends (S214: YES), the CPU 21 completes the process.

<Effect of Embodiment>
As described above, according to the present embodiment, distance acquisition is performed based on the reference area Sa of the reference shape H registered by the reference shape registration process and the actual area Sb of the actual shape J acquired at the time of actual measurement. Even at a short distance where the output of the image sensor 240 is saturated, the distance can be appropriately acquired. Therefore, distance acquisition can be performed over a wide range.

  In addition, according to the present embodiment, distance information is acquired from the calculation based on the distance conversion function held in the memory 25 according to the ratio between the reference area Sa and the actual measurement area Sb. Distance information can be acquired.

In addition, according to the present embodiment, the determination regions W1 to Wn are set to be smaller than the size of the detection target object at the reference distance D where the luminance of the detection target object is approximately the maximum value. Regardless of the orientation, the position where the luminance of the detection target object is approximately the maximum value can be determined appropriately. Thereby, the reference shape H and the reference area Sa can be appropriately set in the reference shape registration table T.

<Example of change>
While the embodiments of the present invention have been described above, the present invention is not limited to the above-described embodiments, and various modifications can be made to the configuration example of the present invention.

  For example, in the above embodiment, as shown in FIG. 6 (f), one reference shape H and one reference area Sa are set in the reference shape registration table T, but as shown in FIG. A plurality of different reference shapes H and reference areas Sa may be set. Further, reference shapes having different angles with the same reference shape H may be set.

  According to this modified example, even if the shape of the detection target object changes during actual measurement, the detection target object can be detected properly, and distance acquisition can be performed appropriately.

  Further, in the above embodiment, the distance of the detection target object at the time of actual measurement is acquired using the position at which the CMOS image sensor 240 is saturated by the detection target object as the reference distance D. However, the present invention is not limited to this. For example, as illustrated in FIG. 10, the distance of the detection target object at the time of actual measurement may be acquired with the distance Lm at which the luminance value is approximately half (128) as the reference distance D. In this case, the reference shape registration table T registers the reference shape H and the reference area Sa at a distance Lm corresponding to a luminance value of 128.

  According to this modified example, the distance can be acquired in accordance with the change in the size of the detection target object as in the above-described embodiment, and thus the distance can be acquired in a wide range.

  In the case of this modified example, since the reference distance D is set to a slightly long distance, an object other than the detection target object is likely to be positioned ahead of the reference distance D when the reference shape is registered. When the reference shape is registered, if an object other than the detection target object is positioned in front of the reference distance D, the reference shape H of the detection target object cannot be properly acquired. Thus, as in the above embodiment, the reference distance D is It is desirable to set the luminance value at the shortest distance to the distance L0 at which the luminance value is approximately the maximum value (255).

  Further, in the above embodiment, the average luminance Ba included in the determination areas W1 to Wn is calculated and the position (reference distance D) where the CMOS image sensor 240 is saturated is determined, but is included in the determination areas W1 to Wn. The position at which the CMOS image sensor 240 is saturated may be determined based on the total luminance value. In addition, the position where the CMOS image sensor 240 is saturated may be determined based on the maximum luminance among the luminances included in the determination regions W1 to Wn.

  In the above embodiment, the determination areas W1 to Wn having a predetermined size are set in order to determine the position where the CMOS image sensor 240 is saturated by the detection target object. The position at which the CMOS image sensor 240 is saturated by the object to be detected may be determined based on the brightness of. However, in order to suppress erroneous detection due to noise or the like, it is desirable that the position where the CMOS image sensor 240 is saturated is determined based on a plurality of luminance values included in the determination area of a predetermined size.

  In the above embodiment, the distance acquisition is performed in accordance with the reference area Sa of the reference shape H and the actual measurement area Sb of the actual measurement J in all distance ranges, but the luminance acquisition shown in FIG. If it is within the possible range, distance acquisition may be performed according to the luminance value.

In the above embodiment, the reference area Sa and the actual measurement area Sb are obtained by counting the number of pixels included in the reference shape H and the actual measurement shape J. The reference area Sa and the actually measured area Sb may be acquired by calculation based on the value.

  In the above embodiment, the reference shape acquisition unit 21b registers the reference shape H and the reference area Sa in the reference shape registration table T in the memory 25 before the object detection. At the time of manufacture, an assumed reference shape H and reference area Sa of the detection target object may be registered. In this case, the function of the reference shape acquisition unit 21b shown in FIG. 2 is omitted. In the case of this modification, it is desirable that a large number of reference shapes H and reference areas Sa are registered in the reference shape registration table T in order to cope with changes in the shape of the detection target object.

  Also in this modified example, since the distance can be acquired according to the change in the size of the detection target object as in the above embodiment, the distance can be acquired in a wide range.

  Note that the reference shape H and the reference area Sa of the detection target object are assumed to change according to the detection situation (for example, the detection target person is switched). Therefore, in order to cope with changes in the detection situation, the reference shape H and the reference area Sa are set in the reference shape registration table T by the reference shape acquisition unit 21b before object detection as in the above embodiment. Is preferable.

  Further, in the above embodiment and the modified example, the luminance value is acquired for all the pixels on the CMOS image sensor 240. However, the luminance value is not necessarily acquired for all the pixels. A value may be obtained.

  In the above embodiment, the filter 230 of the imaging unit 200 is arranged between the imaging lens 220 and the CMOS image sensor 240. However, the arrangement position of the filter 230 is not limited to this, and the imaging lens 220 is not limited thereto. It may be closer to the target area.

  In the above embodiment, the distance information is acquired by software processing by the function of the CPU 21, but acquisition of the distance information may be realized by hardware processing by a circuit.

  Furthermore, in the above-described embodiment, the CMOS image sensor 240 is used as the light receiving element, but a CCD image sensor can be used instead. Also, light in a wavelength band other than infrared can be used for distance acquisition.

  In addition, the embodiment of the present invention can be variously modified as appropriate within the scope of the technical idea shown in the claims.

DESCRIPTION OF SYMBOLS 1 ... Personal computer 2 ... Object detection apparatus 3 ... Information processing part 21b ... Reference | standard shape acquisition part 21c ... Object detection part 23 ... Imaging signal processing circuit (luminance acquisition part)
DESCRIPTION OF SYMBOLS 100 ... Projection part 110 ... Light source 200 ... Imaging part 220 ... Imaging lens 230 ... Filter 240 ... CMOS image sensor (image sensor)
Ba: Average luminance (luminance value)
D: Reference distance H: Reference shape J: Measured shape Sa: Reference area (size of reference shape)
Sb ... Measured area (size of measured shape)
W1 to Wn ... Determination area

Claims (7)

A projection unit that projects light onto the target area;
An imaging unit for imaging the target area by an image sensor;
A luminance acquisition unit for acquiring a luminance value of a predetermined pixel on the image sensor;
An object detection unit that detects a detection target object existing in the target area based on the luminance value acquired by the luminance acquisition unit;
The object detection unit obtains the size of the actually measured shape of the detection target object during actual measurement, and the size of the actually measured shape and the size of the reference shape of the detection target object when the detection target object is at a reference distance And obtaining information on the distance of the detection target object based on
An object detection apparatus characterized by that. The object detection apparatus according to claim 1,
A reference shape acquisition unit that acquires the reference shape of the detection target object at the reference distance and the size of the reference shape based on the luminance value acquired by the luminance acquisition unit;
An object detection apparatus characterized by that. The object detection device according to claim 2,
The reference shape acquisition unit includes definition information that defines a relationship between a luminance value and a distance, and acquires the reference shape and the size of the reference shape at the reference distance based on the definition information. An object detection device. The object detection device according to claim 3,
The reference distance is set to a distance at which a luminance value acquired from the image sensor is saturated from the distance when the detection target object is brought close to the imaging unit.
An object detection apparatus characterized by that. The object detection device according to claim 4,
The reference shape acquisition unit sets a determination region smaller than the size of the reference shape at the reference distance, and the luminance value acquired from the image sensor is saturated based on the luminance value included in the determination region. To determine whether
An object detection apparatus characterized by that. In the object detection device according to any one of claims 3 to 5,
The object detection unit acquires information on the distance of the detection target object at the time of actual measurement based on the amount of change in the actual measurement shape with respect to the size of the reference shape and the specified information.
An object detection apparatus characterized by that. In the object detection device according to any one of claims 1 to 6,
The projection unit includes a light source that emits light in an infrared wavelength band,
The imaging unit includes a filter that transmits light in the infrared wavelength band, and an imaging lens that focuses the light irradiated on the target region onto the image sensor.
An object detection apparatus characterized by that.

Images