Distance measuring device

Abstract

A distance measuring device (1) comprises: a first imaging unit (10) and a second imaging unit (20) arranged side by side so that visual fields (10a, 20a) thereof overlap each other; a projection unit (30) that projects pattern light (30a) having a plurality of types of light regions different from one another in wavelength band distributed by prescribed patterns to the overlapping range of the visual fields (10a, 20a); and a measuring unit (45) that carries out a stereo corresponding point search process for images respectively acquired by the first imaging unit (10) and second imaging unit (20), and measures a distance to an object surface to which the pattern light (30a) has been projected.

Description

Technical Field

[0001] The present invention relates to a distance measurement device that processes an image acquired by a stereo camera to measure the distance to an object.

Background Art

[0002] Conventionally, a distance measurement device that processes an image acquired by a stereo camera to measure the distance to an object is known. In this device, parallax is detected from the images captured by each camera. A pixel block having the highest correlation with a target pixel block on one image (reference image) is searched for on the other image (reference image). The search range is set in the separation direction of the cameras with the same position as the target pixel block as the reference position. The amount of pixel shift of the pixel block extracted by the search with respect to the reference position is detected as parallax. From this parallax, the distance to the object is calculated by the triangulation method.

[0003] In such a distance measurement device, furthermore, light of a specific pattern can be projected onto an object. Thereby, even when the surface of the object is plain, the above search can be performed accurately.

[0004] The following Patent Document 1 describes a configuration for generating dot pattern light from laser light emitted from a semiconductor laser by a diffractive optical element. In this configuration, the diffractive optical element has a multi-stage diffraction efficiency difference, and a dot pattern having multi-stage luminance gradations is formed by this diffraction efficiency difference.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0006] However, the surface of an object may have low reflectivity or high light absorption in a given wavelength range. In the configuration of Patent Document 1 described above, if the wavelength of the laser light falls within this wavelength range, the camera cannot properly acquire the brightness gradation of the dot pattern. As a result, it becomes difficult to properly perform the stereo correspondence point search process for the pixel blocks, and consequently, the distance to the object surface cannot be measured properly.

[0007] In view of these problems, the present invention aims to provide a distance measuring device that can accurately measure the distance to an object surface regardless of the reflectance and light absorption rate of the object surface. [Means for solving the problem]

[0008] The distance measuring device according to the main aspect of the present invention is arranged side by side so that their fields of view overlap. Each is equipped with a monochrome image sensor. The first imaging unit and the second imaging unit, and multiple types of optical regions with different wavelength bands. and the realm of light A projection unit projects pattern light, which is distributed in a predetermined pattern, onto the overlapping area of ​​the field of view, and performs stereo correspondence point search processing on the images acquired by the first imaging unit and the second imaging unit, respectively. Identify the second pixel block in the other image that best matches the pixel values ​​of each first pixel block in one of the aforementioned images, and then, based on the pixel shift between the first and second pixel blocks, determine the results using triangulation. The system includes a measuring unit that measures the distance to the surface of an object onto which the patterned light is projected.

[0009] According to the distance measuring device of this embodiment, patterned light in which multiple types of light regions with different wavelength bands are distributed in a predetermined pattern is projected onto the object surface. Therefore, even if the object surface has a low reflectivity or a high light absorption rate for any of these wavelength bands, patterns from light in the other wavelength bands are included in the images captured by the first and second imaging units. As a result, the uniqueness of each pixel block is maintained by the distribution pattern of light in the other wavelength bands, and stereo correspondence point search can be performed with high accuracy. Thus, the distance to the object surface can be measured with high accuracy. [Effects of the Invention]

[0010] As described above, the present invention provides a distance measuring device that can accurately measure the distance to an object surface regardless of the reflectance and light absorption rate of the object surface.

[0011] The effects and significance of the present invention will become even clearer from the description of the embodiments shown below. However, the embodiments shown below are merely examples of how to implement the present invention, and the present invention is not limited in any way to those described in the embodiments below. [Brief explanation of the drawing]

[0012] [Figure 1] Figure 1 shows the basic configuration of a distance measuring device according to an embodiment. [Figure 2] Figure 2 shows the configuration of a distance measuring device according to an embodiment. [Figure 3] Figures 3(a) and 3(b) schematically illustrate a method for setting pixel blocks for a first image according to an embodiment, respectively. [Figure 4] Figure 4(a) is a schematic diagram showing a state in which a target pixel block is set on a first image according to the embodiment. Figure 4(b) is a schematic diagram showing a search range set on a second image to search for the target pixel block in Figure 4(a) according to the embodiment. [Figure 5] Figure 5(a) is a schematic diagram showing the configuration of a filter according to the embodiment. Figure 5(b) is a magnified view of a part of the filter according to the embodiment. [Figure 6] Figures 6(a) and 6(b) schematically show the optical regions of light that have passed through different types of filter regions according to the respective embodiments. [Figure 7] Figures 7(a) and 7(b) schematically show the optical regions of light that have passed through different types of filter regions according to the respective embodiments. [Figure 8] Figures 8(a) to 8(d) are graphs showing various spectral characteristics according to the respective embodiments. Figure 8(e) is a graph showing the maximum brightness of each dot light according to the embodiment. [Figure 9] Figures 9(a) to 9(d) are graphs showing various spectral characteristics according to the embodiment. Figure 9(e) is a graph showing the maximum luminance of each dot light according to the embodiment. [Figure 10] Figures 10(a) to 10(d) are graphs showing various spectral characteristics according to the embodiment. Figure 10(e) is a graph showing the maximum luminance of each dot light according to the embodiment. [Figure 11] Figure 11 is a flowchart showing the setting process of the light emission amount (driving current) of each light source according to the embodiment. [Figure 12] Figure 12 is a diagram showing the configuration of the distance measuring device according to Modification 1. [Figure 13] Figures 13(a) to 13(d) are graphs showing various spectral characteristics according to Modification 1. Figure 13(e) is a graph showing the maximum luminance of each dot light according to Modification 1. [Figure 14] Figure 14(a) is a diagram schematically showing the configuration of the filter according to Modification 2. Figure 14(b) is a diagram showing an enlarged view of a part of the filter according to Modification 2.

[0013] However, the drawings are for illustrative purposes only and do not limit the scope of this invention.

Embodiments for Carrying Out the Invention

[0014] Hereinafter, embodiments of the present invention will be described with reference to the drawings. For convenience, X, Y, and Z axes orthogonal to each other are added to each figure. The X-axis direction is the arrangement direction of the first imaging unit and the second imaging unit, and the positive Z-axis direction is the imaging direction of each imaging unit.

[0015] Figure 1 is a diagram showing the basic configuration of the distance measuring device 1.

[0016] The distance measuring device 1 includes a first imaging unit 10, a second imaging unit 20, and a projection unit 30.

[0017] The first imaging unit 10 captures an area of ​​field of view 10a oriented in the positive Z-axis direction. The second imaging unit 20 captures an area of ​​field of view 20a oriented in the positive Z-axis direction. The first imaging unit 10 and the second imaging unit 20 are arranged side by side in the X-axis direction such that their respective fields of view 10a and 20a overlap. The imaging direction of the first imaging unit 10 may be slightly tilted from the positive Z-axis direction towards the second imaging unit 20, and the imaging direction of the second imaging unit 20 may be slightly tilted from the positive Z-axis direction towards the first imaging unit 10. The positions of the first imaging unit 10 and the second imaging unit 20 in the Z-axis direction and in the Y-axis direction are the same.

[0018] The projection unit 30 has a field of view 10a of the first imaging unit 10 and a field of view of the second imaging unit 20. 20a A patterned light 30a, in which light is distributed in a predetermined pattern, is projected onto the overlapping area. The projection direction of the patterned light 30a by the projection unit 30 is the positive Z-axis direction. The patterned light 30a is projected onto the surface of object A1 that exists in the overlapping area of ​​fields 10a and 20a.

[0019] The distance measuring device 1 measures the distance D0 to object A1 by stereo correspondence point search using the captured images captured by the first imaging unit 10 and the second imaging unit 20, respectively. At this time, pattern light 30a is projected onto the surface of object A1 from the projection unit 30. As a result, the pattern of the pattern light 30a is projected onto the captured images of the first imaging unit 10 and the second imaging unit 20. Therefore, even if the surface of object A1 is plain, stereo correspondence point search can be performed with high accuracy, and the distance D0 to the surface of object A1 can be accurately measured.

[0020] Here, the surface of object A1 may have high light absorption and low reflectivity in a predetermined wavelength range. In this case, if the wavelength range of pattern light 30a is included in this wavelength range, the first imaging unit 10 and the second imaging unit 20 may not be able to properly image the pattern of pattern light 30a. As a result, the stereo correspondence point search described above may not be performed properly, and consequently, the distance D0 to the surface of object A1 may not be measured accurately.

[0021] Therefore, in this embodiment, the pattern light 30a is configured such that multiple types of light regions with different wavelength bands are distributed in a predetermined pattern. Even if the surface of object A1 has a low reflectivity or a high light absorption coefficient for any of these wavelength bands, patterns from light in the other wavelength bands are captured by the first imaging unit 10 and the second imaging unit 20. Thus, stereo correspondence point search can be properly performed based on the patterns of light in the other wavelength bands, and the distance to the surface of object A1 can be measured with high accuracy.

[0022] Figure 2 shows the configuration of the distance measuring device 1.

[0023] The first imaging unit 10 includes an imaging lens 11 and an image sensor 12. The imaging lens 11 focuses light from the field of view 10a onto the imaging surface 12a of the image sensor 12. The imaging lens 11 does not have to be a single lens, but may be composed of a combination of multiple lenses. The image sensor 12 is a monochrome image sensor. The image sensor 12 is, for example, a CMOS image sensor. The image sensor 12 may also be a CCD.

[0024] The second imaging unit 20 has the same configuration as the first imaging unit 10. The second imaging unit 20 includes an imaging lens 21 and an image sensor 22. The imaging lens 21 focuses light from the field of view 20a onto the imaging surface 22a of the image sensor 22. The imaging lens 21 does not have to be a single lens, but may be composed of a combination of multiple lenses. The image sensor 22 is a monochrome image sensor. The image sensor 22 is, for example, a CMOS image sensor. The image sensor 22 may also be a CCD.

[0025] The projection unit 30 comprises light sources 31-33, an optical system 34, a filter 35, and a projection lens 36.

[0026] Light sources 31-33 emit light in different wavelength bands. For example, light source 31 emits light in the orange wavelength band, light source 32 emits light in the green wavelength band, and light source 33 emits light in the blue wavelength band. Light sources 31-33 are light-emitting diodes. Light sources 31-33 may be other types of light sources, such as semiconductor lasers.

[0027] The optical system 34 comprises collimator lenses 341-343 and dichroic mirrors 344 and 345.

[0028] Collimator lenses 341-343 convert the light emitted from light sources 31-33 into approximately parallel light, respectively. Dichroic mirror 344 transmits the light incident from collimator lens 341 and reflects the light incident from collimator lens 342. Dichroic mirror 345 transmits the light incident from dichroic mirror 344 and reflects the light incident from collimator lens 343. In this way, the light emitted from light sources 31-33 is integrated and guided to filter 35.

[0029] The filter 35 generates patterned light 30a from light of each wavelength band derived from the optical system 34, in which multiple types of light regions with different wavelength bands are distributed in a predetermined pattern. The configuration and operation of the filter 35 will be explained later with reference to Figures 5(a) and (b).

[0030] The projection lens 36 projects the pattern light 30a generated by the filter 35. The projection lens 36 does not have to be a single lens, but may be composed of a combination of multiple lenses.

[0031] The distance measuring device 1 comprises, as a circuit configuration, a first image processing unit 41, a second image processing unit 42, a light source drive unit 43, a brightness adjustment unit 44, a measurement unit 45, a control unit 46, and a communication interface 47.

[0032] The first image processing unit 41 and the second image processing unit 42 control the image sensors 12 and 22, and perform processing such as brightness correction and camera calibration on the pixel signals of the first and second images output from the image sensors 12 and 22, respectively.

[0033] The light source drive unit 43 drives the light sources 31 to 33 respectively with the drive current value set by the brightness adjustment unit 44.

[0034] The brightness adjustment unit 44 sets the drive current values ​​of the light sources 31 to 33 in the light source drive unit 43 based on the pixel signals (brightness) of the second image input from the second imaging processing unit 42. More specifically, the brightness adjustment unit 44 sets the drive current values ​​(amount of light emitted) of the light sources 31 to 33 so that the maximum brightness based on the light from the light sources 31 to 33 acquired based on the pixel signals from the second imaging unit 20 differs from that of the light sources 31 to 33. The processing of the brightness adjustment unit 44 will be explained later with reference to Figure 11.

[0035] The measurement unit 45 compares the first image and the second image input from the first image processing unit 41 and the second image processing unit 42, respectively, to perform stereo correspondence point search and obtains the distance to the surface of object A1 for each pixel block on the first image. The measurement unit 45 transmits the distance information for all acquired pixel blocks to an external device via the communication interface 47.

[0036] Specifically, the measurement unit 45 sets a pixel block to be used for distance acquisition (hereinafter referred to as the "target pixel block") on the first image, and searches for a pixel block corresponding to this target pixel block, that is, a pixel block that best matches the target pixel block (hereinafter referred to as the "matching pixel block"), within a defined search range on the second image. The measurement unit 45 then acquires the amount of pixel displacement between a pixel block at the same position as the target pixel block on the second image (hereinafter referred to as the "reference pixel block") and the matching pixel block extracted from the second image through the above search, and calculates the distance from the acquired amount of pixel displacement to the surface of object A1 at the position of the target pixel block.

[0037] The measurement unit 45 and the communication interface 47 may be composed of semiconductor integrated circuits consisting of an FPGA (Field Programmable Gate Array). Alternatively, each of these units may be composed of other semiconductor integrated circuits such as a DSP (Digital Signal Processor), a GPU (Graphics Processing Unit), and an ASIC (Application Specific Integrated Circuit).

[0038] The control unit 46 is composed of a microcomputer or the like and controls each part according to a predetermined program stored in its internal memory.

[0039] Figures 3(a) and 3(b) schematically illustrate the method for setting up the pixel block 102 for the first image 100. Figure 3(a) shows the method for setting up the pixel block 102 for the entire first image 100, and Figure 3(b) shows a magnified view of a portion of the first image 100.

[0040] As shown in Figures 3(a) and (b), the first image 100 is divided into a plurality of pixel blocks 102, each containing a predetermined number of pixel regions 101. A pixel region 101 is an area corresponding to one pixel on the image sensor 12. In other words, a pixel region 101 is the smallest unit of the first image 100. In the example of Figures 3(a) and (b), one pixel block 102 is composed of nine pixel regions 101 arranged in three rows and three columns. However, the number of pixel regions 101 included in one pixel block 102 is not limited to this.

[0041] Figure 4(a) schematically shows the state in which the target pixel block TB1 is set on the first image 100, and Figure 4(b) schematically shows the search range R0 set on the second image 200 in order to search for the target pixel block in Figure 4(a).

[0042] In Figure 4(b), for convenience, the second image 200 acquired from the second imaging unit 20 is divided into multiple pixel blocks 202, similar to the first image 100. Each pixel block 202 contains the same number of pixel regions as the pixel block 102 described above.

[0043] In Figure 4(a), the target pixel block TB1 is the pixel block 102 on the first image 100 that is to be processed. Also, in Figure 4(b), the reference pixel block TB2 is the pixel block 202 on the second image 200 that is in the same position as the target pixel block TB1.

[0044] The measurement unit 45 in Figure 2 identifies a reference pixel block TB2 located at the same position as the target pixel block TB1 on the second image 200. The measurement unit 45 then sets the position of the identified reference pixel block TB2 as the reference position P0 of the search range R0, and sets the search range R0 to be the range extending from this reference position P0 in the direction of separation between the first imaging unit 10 and the second imaging unit 20.

[0045] The direction in which the search range R0 extends is set in the direction in which the pixel block corresponding to the target pixel block TB1 (matching pixel block MB2) is shifted from the reference position P0 due to parallax on the second image 200. Here, the search range R0 is set to the range of 12 pixel blocks 202 lined up to the right from the reference position P0 (corresponding to the X-axis direction in Figure 1). However, the number of pixel blocks 202 included in the search range R0 is not limited to this. Also, the starting point of the search range R0 is not limited to the reference pixel block TB2; for example, the starting point of the search range R0 may be set to a position shifted a few blocks to the right from the reference pixel block TB2.

[0046] The measurement unit 45 searches for a pixel block (matching pixel block MB2) corresponding to the target pixel block TB1 within the search range R0 thus set. Specifically, the measurement unit 45 calculates a correlation value between the target pixel block TB1 and each search position while shifting the search position one pixel at a time to the right from the reference pixel block TB2. For example, SSD or SAD may be used for the correlation value. The measurement unit 45 then identifies the pixel block at the search position with the highest correlation within the search range R0 as the matching pixel block MB2.

[0047] Furthermore, the measurement unit 45 acquires the amount of pixel displacement of the fitted pixel block MB2 relative to the reference pixel block TB2. Then, the measurement unit 45 calculates the distance to the surface of object A1 using triangulation based on the acquired amount of pixel displacement and the distance between the first imaging unit 10 and the second imaging unit 20. The measurement unit 45 performs the same process for all pixel blocks 102 (target pixel block TB1) on the first image 100. Once the distances for all pixel blocks 102 have been acquired, the measurement unit 45 transmits this distance information to an external device via the communication interface 47.

[0048] The distance measuring device 1 having the above configuration can be used in a fixed position, or, for example, installed on the end effector (gripping part, etc.) of a robot arm performing work operations in a factory. In this case, the control unit 46 of the distance measuring device 1 receives a distance acquisition instruction from the robot controller via the communication interface 47 during the robot arm's work process. In response to this instruction, the control unit 46 causes the measuring unit 45 to measure the distance between the position of the end effector and the surface of the object A1 to be worked on, and transmits the measurement result to the robot controller via the communication interface 47. The robot controller then provides feedback control to the operation of the end effector based on the received distance information. When the distance measuring device 1 is installed on an end effector in this way, it is desirable that the distance measuring device 1 be small and lightweight.

[0049] Figure 5(a) is a schematic diagram showing the configuration of the filter 35 in Figure 2. Figure 5(b) is a magnified view of a part of Figure 5(a). Figures 5(a) and (b) show the filter 35 as viewed from the light incident surface 35a side.

[0050] As shown in Figures 5(a) and (b), multiple types of filter regions 351 to 354 are formed on the incident surface 35a of the filter 35 in a predetermined pattern. In Figures 5(a) and (b), the types of filter regions 351 to 354 are shown with different hatching patterns. Filter regions 351 to 353 selectively transmit light of different wavelength bands. Here, the transmitted wavelength bands of filter regions 351 to 353 correspond to the wavelength bands of light emitted from light sources 31 to 33, respectively.

[0051] In other words, filter region 351 has high transmittance mainly for the wavelength band of light from light source 31 and low transmittance to other wavelength bands. Filter region 352 has high transmittance mainly for the wavelength band of light from light source 32 and low transmittance to other wavelength bands. Filter region 353 has high transmittance mainly for the wavelength band of light from light source 33 and low transmittance to other wavelength bands.

[0052] The filter region 354 is set to have a low transmittance for all wavelengths of light from light sources 31 to 33. In other words, the filter region 354 effectively blocks light from light sources 31 to 33.

[0053] The size of each filter region 351 to 354 is set to approximately correspond to the size of one pixel on the image sensors 12 and 22. For example, region B1 shown by the dashed line in Figure 5(b) corresponds to the region of a pixel block consisting of 3 vertical and 3 horizontal pixels on the image sensors 12 and 22 (pixel blocks 102 and 202 used for the stereo correspondence point search described above). That is, when the distance D0 to the surface of object A1 is at a reference distance (for example, the intermediate distance of the distance measurement range), the light from region B1 is projected onto the region of the pixel block consisting of 3 vertical and 3 horizontal pixels on the image sensors 12 and 22.

[0054] However, the size of each filter region 351-354 is not necessarily limited to the size corresponding to one pixel. The size of each filter region 351-354 may be larger or smaller than the size corresponding to one pixel. Also, in Figure 5(b), each filter region 351-354 is rectangular and the same size as the others, but the sizes of each filter region 351-354 may be different, and their shapes may also be other shapes such as squares or circles.

[0055] It is preferable that the filter regions 351 to 354 are arranged such that each region B1, which corresponds to all pixel blocks used in stereo correspondence point search, contains different types of filter regions, and it is even more preferable that each of these regions B1 contains all types of filter regions 351 to 354. Furthermore, it is preferable that the arrangement pattern of the filter regions included in region B1 corresponding to a pixel block is unique (random) for each pixel block at each search position within the search range R0 in stereo correspondence point search.

[0056] When filter regions 351 to 354 are arranged in this manner, as described later, the brightness of the light passing through filter regions 351 to 354 can be made different from one another, thereby making the brightness distribution of light within each pixel block unique. This improves the accuracy of stereo correspondence point search, and consequently, improves the accuracy of distance measurement.

[0057] Filter regions 351 to 354 are formed, for example, by the following process.

[0058] First, a color resist is applied to the surface of a transparent glass substrate to form the filter region 351. Next, with the areas other than the filter region 351 masked, ultraviolet light is irradiated to insolve the color resist in the area corresponding to the filter region 351. Once insolubilization is complete, the mask is removed, unwanted color resist is removed with an alkaline developer, and then a post-bake treatment is performed to cure the color resist in the filter region 351. This forms the filter region 351 on the glass substrate.

[0059] The above process is performed sequentially on filter regions 352 to 354. This sequentially forms filter regions 352 to 354 on the glass substrate. In this way, all of the filter regions 351 to 354 are formed on the glass substrate. After that, a protective film is formed on the surface of the filter regions 351 to 354. This completes the formation of the filter 35.

[0060] Figures 6(a), (b) and 7(a), (b) schematically show the optical regions of the light that passes through the filter regions 351 to 354 when light from light sources 31 to 33 is incident on the entire area of ​​Figure 5(b).

[0061] Figure 6(a) shows the distribution of light (dot light DT1) transmitted through filter region 351 in Figure 5(b), and Figure 6(b) shows the distribution of light (dot light DT2) transmitted through filter region 352 in Figure 5(b). Furthermore, Figure 7(a) shows the distribution of light (dot light DT3) transmitted through filter region 353 in Figure 5(b), and Figure 7(b) shows the distribution of light in the area blocked by filter region 354 in Figure 5(b) (lightless dot DT4).

[0062] From the region shown in Figure 5(b), the dot lights DT1-DT3 and the unlit dot DT4 shown in Figures 6(a) to 7(b) are integrated and projected. From other regions of the filter 35, the dot lights DT1-DT3 and the unlit dot DT4 are also projected in a distribution corresponding to the distribution of the filter regions 351-354. In this way, the dot lights DT1-DT3 and the unlit dot DT4 projected from the filter 35 are irradiated onto the surface of object A1 as pattern light 30a. Subsequently, the dot lights DT1-DT3 and the unlit dot DT4 are reflected from the surface of object A1 and then captured by the first imaging unit 10 and the second imaging unit 20. As a result, the first image 100 and the second image 200, on which the dot lights DT1-DT3 and the unlit dot DT4 are projected, are obtained.

[0063] Here, the light emission amounts of light sources 31 to 33 are set so that the maximum brightness of the dot lights DT1 to DT3 and the unlit dot DT4 on the second image 200 differs from each other. More specifically, the light emission amounts of light sources 31 to 33 are set so that the maximum brightness of the dot lights DT1 to DT3 and the unlit dot DT4 on the second image 200 differs approximately equally in order of decreasing brightness.

[0064] Figures 8(a) to 8(e) are diagrams illustrating how to set the light emission levels of light sources 31 to 33.

[0065] Figure 8(a) is a graph showing the spectral output of light sources 31-33. The spectral output of light sources 31-33 is shown by solid, dotted, and dashed lines, respectively. Here, the vertical axis of the graph is normalized by the maximum output of light source 31.

[0066] Light source 31 emits light with a central wavelength of approximately 610 nm and an emission bandwidth of approximately 80 nm. Light source 32 emits light with a central wavelength of approximately 520 nm and an emission bandwidth of approximately 150 nm. Light source 33 emits light with a central wavelength of approximately 470 nm and an emission bandwidth of approximately 100 nm.

[0067] Figure 8(b) is a graph showing the spectral transmittance of filter regions 351 to 353. The spectral transmittances of filter regions 351 to 353 are shown by solid, dotted, and dashed lines, respectively. Here, the vertical axis of the graph is normalized by the maximum transmittance of filter region 351.

[0068] In filter region 351, the transmittance increases with increasing wavelength from around 570 nm, and maintains maximum transmittance above around 650 nm. Filter region 352 has a maximum transmittance of around 520 nm and spectral characteristics with a transmission bandwidth of approximately 150 nm. Filter region 353 has a maximum transmittance of around 460 nm and spectral characteristics with a transmission bandwidth of approximately 150 nm.

[0069] Note that the spectral transmittance of filter region 354 is omitted from the illustration. The spectral transmittance of filter region 354 is approximately zero near the emission band of light sources 31-33 (here, 400-650 nm).

[0070] Figure 8(c) is a graph showing the spectral reflectance of the surface of object A1, which is the measurement surface. Here, the case where the reflectance of the measurement surface is constant regardless of wavelength, that is, the reflectance of the measurement surface does not depend on wavelength, is illustrated. The vertical axis of the graph is normalized by the maximum reflectance.

[0071] Figure 8(d) is a graph showing the spectral sensitivity of the first imaging unit 10 and the second imaging unit 20. The spectral sensitivity of the first imaging unit 10 and the second imaging unit 20 is mainly determined by the spectral transmittance of the imaging lenses 11 and 21 and the spectral sensitivity of the image sensors 12 and 22. The vertical axis of the graph is normalized by the maximum sensitivity. Here, the spectral sensitivity is maximum at around 600 nm.

[0072] Figure 8(e) is a graph showing the maximum brightness of dot lights DT1-DT3 and blank dot DT4 in the second image 200, when the spectral output of light sources 31-33, the spectral transmittance of filter regions 351-354, the spectral reflectance of the measurement surface (surface of object A1), and the spectral sensitivity of the first imaging unit 10 and the second imaging unit 20 have the characteristics shown in Figures 8(a)-(d), respectively. Here, the vertical axis of the graph is normalized by the maximum brightness of dot light DT1.

[0073] In this case, the maximum brightness of dot light DT3 is about one-third of the maximum brightness of dot light DT1, and the maximum brightness of dot light DT2 is about two-thirds of the maximum brightness of dot light DT1. In other words, if the reflectivity of the measurement surface does not depend on wavelength, by setting the peak values ​​of the spectral output of light sources 31 to 33 as shown in Figure 8(a), the maximum brightness of dot lights DT1 to DT3 based on light from light sources 31 to 33 can be made to differ approximately equally in order of brightness.

[0074] As shown in Figure 2, the brightness adjustment unit 44 initializes the light emission amount (drive current value) of the light sources 31 to 33 so that the maximum brightness of the dot lights DT1 to DT3, based on the light from the light sources 31 to 33, differs approximately uniformly in order of brightness magnitude. As a result, if the reflectance of the measurement surface (surface of object A1) is not wavelength dependent, the maximum brightness of the dot lights DT1 to DT3 on the first image 100 and the second image 200 will have approximately uniform gradation differences. Therefore, during the stereo correspondence point search described above, a correlation value that significantly peaks at the search position of the matching pixel block MB2 is calculated. Thus, the position of the matching pixel block MB2 can be accurately identified, and as a result, distance measurement can be performed accurately.

[0075] On the other hand, when the light emission amount (drive current value) of light sources 31-33 is set to the initial value in this way, if the reflectance of the measurement surface (surface of object A1) is wavelength-dependent, the maximum brightness of the dot lights DT1-DT3 on the first image 100 and the second image 200 will no longer have a nearly uniform gradation difference.

[0076] Figure 9(c) is a graph showing the spectral reflectance of the reflectance of the measurement surface (the surface of object A1) when the reflectance of the measurement surface has wavelength dependence, and Figure 9(e) is a graph showing the maximum brightness of the dot lights DT1~DT3 and the blank dot DT4 in the second image 200 in that case. Figures 9(a), (b), and (d) are the same as Figures 8(a), (b), and (d).

[0077] If the reflectance of the measurement surface has the spectral reflectance shown in Figure 9(c), and the light sources 31-33 have the spectral output shown in Figure 9(a), then, as shown in Figure 9(e), the gradation difference between the maximum brightness of dot light DT2 and the maximum brightness of dot light DT1 becomes small. As a result, in the second image 200, the regions of dot light DT1 and dot light DT2 become less distinguishable by brightness, and these regions are more likely to be integrated into a single region for detection. Consequently, the specificity of the dot distribution in the pixel block decreases, and the search accuracy in stereo correspondence point search decreases.

[0078] However, in this case, while the specificity due to dot light DT1 and DT2 in the pixel block decreases, the specificity due to dot light DT3 is maintained. Furthermore, even if the regions of dot light DT1 and DT2 are integrated into a single region as described above, the pixel positions in which this region is distributed within each pixel block tend to differ between pixel blocks. Therefore, in this case as well, the specificity of the dot pattern in each pixel block tends to be maintained. Thus, in this case as well, by driving each light source with the initial settings, the search accuracy in stereo correspondence point search can be maintained at a high level.

[0079] Furthermore, in order to perform stereo correspondence point search with higher accuracy, when the reflectance of the measurement surface is wavelength-dependent, it is preferable to change the amount of light emitted from light sources 31 to 33 (driving current value) from the initial setting value according to the spectral reflectance of the reflectance of the measurement surface, thereby ensuring a difference in brightness between the maximum brightness of the dot lights.

[0080] Figure 10(a) is a graph showing how to adjust the output of light sources 31-33 in this case. Figures 10(b)-(d) are the same as Figures 9(b)-(d).

[0081] Here, the light output (drive current value) of light source 32 is set lower than in the case of Figure 9(a). As a result, as shown in Figure 10(e), the maximum brightness of dot light DT2 is reduced, and the brightness gradation difference between dot light DT1 and dot light DT2 is maintained as in the case of Figure 8(e). In this way, the maximum brightness of dot lights DT1 to DT3 based on the light from light sources 31 to 33 differs almost equally in order of brightness.

[0082] This ensures that the unique patterns of the dot light DT1-DT3 and the unlit dot DT4 in each pixel block are maintained, as in the case of Figure 8(e). Therefore, stereo correspondence point search can be performed with high accuracy.

[0083] Figure 11 is a flowchart showing the process for setting the light emission amount (driving current) of light sources 31-33. This process is performed by the brightness adjustment unit 44 in Figure 2 before the actual distance measurement to object A1.

[0084] The brightness adjustment unit 44 sets the drive current values ​​of the light sources 31 to 33 to their initial settings (S101). The initial settings for each light source are set so that, assuming there is no wavelength dependence on the reflectance of the surface of object A1, the maximum brightness based on the light from light sources 31 to 33 differs approximately uniformly in order of brightness, as shown in Figure 8(e). Furthermore, the initial settings for each light source are set so that, assuming the reflectance of the surface of object A1 is at a predetermined value (a presumed standard value), the maximum brightness based on the light from light sources 31 to 33 falls appropriately within the range of gradations (for example, 0 to 255) defined by the first imaging processing unit 41 and the second imaging processing unit 42. For example, the initial settings for each light source are set so that the maximum brightness of the largest light source 31 is slightly smaller than the maximum gradation within the range of gradations that define brightness (for example, about 80 to 90% of the maximum gradation).

[0085] Next, the brightness adjustment unit 44 sets one of the light sources 31 to 33 as the target light source and drives this light source with a drive current value set for this light source (S102). For example, light source 31 is set as the target light source. With only the target light source emitting light in this state, the brightness adjustment unit 44 causes either the first imaging unit 10 or the second imaging unit 20 to perform imaging (S103). In this embodiment, the imaging in step S103 is performed by the second imaging unit 20.

[0086] The brightness adjustment unit 44 obtains the maximum brightness of a pixel from the captured image (S104). Here, since imaging is performed by the second imaging unit 20, the brightness adjustment unit 44 obtains the maximum brightness of a pixel from the second image 200 acquired by the second imaging unit 20. As a result, the maximum brightness is obtained from the brightness output from the pixel to which dot light (in this case, dot light DT1) from the target light source (light source 31) is incident on the second image 200.

[0087] Subsequently, the brightness adjustment unit 44 determines whether or not the processing in steps S102 to S104 has been performed for all of the light sources 31 to 33 (S105). If there are any light sources that have not been processed (S105: NO), the brightness adjustment unit 44 sets the next light source as the target light source and drives this light source with the initial setting value (current value) corresponding to this light source (S102). For example, light source 32 is set as the target light source. Subsequently, the brightness adjustment unit 44 similarly performs the processing in steps S103 and S104 to obtain the maximum brightness of a pixel from the second image 200. This obtains the maximum brightness among the brightness output from pixels on the second image 200 that are incident on dot light (in this case, dot light DT2) from the target light source (light source 32).

[0088] In this case as well, since there is still an unprocessed light source (light source 33) remaining (S105:NO), the brightness adjustment unit 44 sets the next light source as the target light source and drives this light source with the initial setting value (current value) corresponding to this light source (S102). As a result, the last light source 33 is set as the target light source. After that, the brightness adjustment unit 44 performs the same processing as in steps S103 and S104 to obtain the maximum brightness of a pixel from the second image 200. As a result, the maximum brightness is obtained from the brightness output from the pixel into which the dot light (in this case, dot light DT3) from the target light source (light source 33) is incident on the second image 200.

[0089] Once the maximum brightness based on the initial settings has been obtained for all of the light sources 31 to 33 (S105: YES), the brightness adjustment unit 44 determines whether the balance of the obtained maximum brightness is appropriate (S106). Specifically, the brightness adjustment unit 44 determines whether the maximum brightness obtained when the light sources 31 to 33 emit light differs in an approximately equal manner in order of brightness, as shown in Figure 8(e).

[0090] Specifically, the brightness adjustment unit 44 determines whether the ratio of the maximum brightness obtained when light source 32 emits light (corresponding to the maximum brightness of dot light DT2) to the maximum brightness obtained when light source 31 emits light (corresponding to the maximum brightness of dot light DT1) falls within a predetermined tolerance range centered around 66%. Furthermore, the brightness adjustment unit 44 determines whether the ratio of the maximum brightness obtained when light source 33 emits light (corresponding to the maximum brightness of dot light DT3) to the maximum brightness obtained when light source 31 emits light (corresponding to the maximum brightness of dot light DT1) falls within a predetermined tolerance range centered around 33%.

[0091] These tolerances are set to a range in which the maximum brightness of adjacent pixels in the size direction can be distinguished, that is, in a pixel block, the dot light DT1 to DT3 can be distinguished by brightness, and the pattern of dot light DT1 to DT3 can maintain its uniqueness. For example, these tolerances are set to a range of approximately ±10% from the aforementioned 66% and 33%.

[0092] The brightness adjustment unit 44 terminates the process shown in Figure 11 if the maximum brightness values ​​obtained when the light sources 31 to 33 emit light are found to differ approximately uniformly in order of brightness, as shown in Figure 8(e) (S106: YES). In this case, the actual distance measurement to object A1 is performed by driving the light sources 31 to 33 according to their respective initial settings.

[0093] On the other hand, if the maximum brightness obtained when the light sources 31 to 33 emit light is not substantially equal in order of brightness (S106: NO), the brightness adjustment unit 44 executes a process to reset the drive current values ​​of the light sources 31 to 33 (S107).

[0094] Specifically, the brightness adjustment unit 44 readjusts the drive current values ​​of light sources 31 to 33 based on the relationship between the brightness and drive current values ​​that has been held in advance, and the current maximum brightness of each light source, so that the maximum brightness based on the light emission of light source 31 becomes slightly less than the maximum gradation (for example, about 80-90% of the maximum gradation), and the maximum brightness based on the light emission of light sources 32 and 33 is about 66% and 33% of this maximum brightness, respectively.

[0095] At this time, the brightness adjustment unit 44 also determines whether any of the three maximum brightness values ​​obtained in step S104 are saturated, that is, whether they have not reached the maximum gradation of the gradation that defines the brightness (for example, 0 to 255). If any of the maximum brightness values ​​are saturated, the brightness adjustment unit 44 sets the drive current value for the light source from which this maximum brightness was obtained to be lower by a predetermined gradation than the drive current value obtained from the relationship between brightness and drive current value and the maximum gradation. In this case as well, the brightness adjustment unit 44 readjusts the drive current values ​​of the light sources 31 to 33 so that the maximum brightness values ​​based on the light emission of the light sources 31 to 33 differ in an approximately equal manner in order of brightness magnitude.

[0096] After resetting the drive current values ​​for the light sources 31-33, the brightness adjustment unit 44 returns to step S102 to obtain the maximum brightness for each light source at the time of illumination using the reset drive current values ​​(S102-S105). The brightness adjustment unit 44 then compares the three newly obtained maximum brightness values ​​and determines whether these maximum brightness values ​​differ in roughly equal proportions in order of brightness (S106).

[0097] If this determination is YES, the brightness adjustment unit 44 terminates the process shown in Figure 11. In this case, the actual distance measurement to object A1 is performed by driving the light sources 31 to 33, each according to the reset drive current value.

[0098] On the other hand, if the determination in step S106 is NO, the brightness adjustment unit 44 resets the drive current values ​​for the light sources 31 to 33 based on the relationship between brightness and drive current value, as described above, using the three maximum brightness values ​​acquired this time (S107), and returns the process to step S102. The brightness adjustment unit 44 resets the drive current values ​​for the light sources 31 to 33 until the maximum brightness values ​​acquired by the light emitted by each of the light sources 31 to 33 differ in an approximately equal manner in order of brightness magnitude (S106: NO, S107). When these maximum brightness values ​​differ in an approximately equal manner in order of brightness magnitude (S106: YES), the brightness adjustment unit 44 terminates the process shown in Figure 11. As a result, the light sources 31 to 33 are driven by the finally set drive current values, and the actual distance measurement to object A1 is performed.

[0099] <Effects of the Embodiment> According to the above embodiment, the following effects are achieved.

[0100] As shown in Figures 6(a) to 7(a), pattern light 30a, in which multiple types of light regions (dot light DT1 to DT3) with different wavelength bands are distributed in a predetermined pattern, is projected onto the surface of object A1. Therefore, even if the surface of object A1 has a low reflectivity or a high light absorption rate for any of these wavelength bands, patterns from light in the other wavelength bands are included in the images captured by the first imaging unit 10 and the second imaging unit 20. As a result, the uniqueness of each pixel block 102 is maintained by the distribution pattern of light in the other wavelength bands, and stereo correspondence point search can be performed with high accuracy. Thus, the distance to the surface of object A1 can be measured with high accuracy.

[0101] As shown in Figure 8(e), the maximum brightness differs among multiple types of light regions (dot light DT1 to DT3). This allows these light regions (dot light DT1 to DT3) to be distinguished by brightness, and the distribution of these light regions (dot light DT1 to DT3) enhances the uniqueness of each pixel block 102. Therefore, stereo correspondence point search can be performed with high accuracy, and the distance to the surface of object A1 can be measured with high accuracy.

[0102] As shown in Figures 2 and 5(a) and (b), the projection unit 30 includes a filter 35 in which multiple types of filter regions 351 to 353 for generating multiple types of light regions (dot light DT1 to DT3) are distributed in a pattern similar to the pattern of the light regions (dot light DT1 to DT3). This makes it easy to generate patterned light 30a in which multiple types of light regions (dot light DT1 to DT3) are distributed in a desired pattern. Furthermore, since there is no variation in diffraction efficiency (variation in brightness gradation) due to manufacturing errors or assembly errors, as is the case with diffractive optical elements, patterned light in which multiple types of light regions (dot light DT1 to DT3) are distributed in a desired pattern can be stably generated.

[0103] As shown in Figure 2, the projection unit 30 includes a plurality of light sources 31-33 that emit light in different wavelength bands, and an optical system 34 that guides the light emitted from the plurality of light sources 31-33 to the filter 35. This makes it easy to irradiate the filter 35 with light to generate multiple types of light regions (dot light DT1-DT3).

[0104] As shown in Figures 8(a) and (b), multiple light sources 31 to 33 are arranged corresponding to multiple types of filter regions 351 to 353, and each filter region 351 to 353 selectively extracts light from the corresponding light source 31 to 33. This allows for the efficient generation of multiple types of light regions (dot light DT1 to DT3).

[0105] As shown in Figure 8(e), the maximum brightness based on the light from each light source 31-33, acquired based on the pixel signal from the second imaging unit 20, differs from one another. This allows for the division of multiple types of light regions (dot light DT1-DT3) based on brightness, and the distribution of these light regions (dot light DT1-DT3) enhances the uniqueness of each pixel block 102. Therefore, stereo correspondence point search can be performed with high accuracy, and the distance to the surface of object A1 can be measured with high accuracy.

[0106] As shown in Figure 11, the brightness adjustment unit 44 sets the light emission amount (drive current value) of the multiple light sources 31 to 33 so that the maximum brightness based on the light from each of the light sources 31 to 33 acquired based on the pixel signal from the second imaging unit 20 differs from one another (S101, S107). This makes it possible to make the maximum brightness based on the light from each of the light sources 31 to 33 differ from one another, even if the reflectance or light absorption rate of the surface of object A1 is wavelength-dependent. Therefore, even if the reflectance or light absorption rate of the surface of object A1 is wavelength-dependent, multiple types of light regions (dot light DT1 to DT3) can be divided by brightness, and the specificity of each pixel block 102 can be enhanced by the distribution of these light regions (dot light DT1 to DT3). Thus, stereo correspondence point search can be performed with high accuracy, and the distance to the surface of object A1 can be measured with high accuracy.

[0107] In steps S101 and S107 of Figure 11, the brightness adjustment unit 44 sets the light emission amount (driving current) of the multiple light sources 31 to 33 so that the maximum brightness based on the light from each light source 31 to 33 acquired based on the pixel signal from the second imaging unit 20 differs substantially in order of brightness, as shown in Figure 10(e). This makes it possible to make the maximum brightness based on the light from each light source 31 to 33 differ significantly from one another. As a result, multiple types of light regions (dot light DT1 to DT3) can be clearly distinguished by brightness, and the uniqueness of each pixel block 102 can be significantly enhanced by the distribution of these light regions (dot light DT1 to DT3). Therefore, stereo correspondence point search can be performed with greater accuracy, and the distance to the surface of object A1 can be measured with greater accuracy.

[0108] In the above embodiment, the light sources 31 to 33 are light-emitting diodes. This suppresses the superposition of speckle noise on the captured images (first image 100, second image 200) of the pattern light 30a. Therefore, stereo correspondence point search can be performed with high accuracy, and the distance to the surface of object A1 can be measured with high accuracy.

[0109] As shown in Figure 7(b), the pattern light 30a includes a region without light (non-light dot DT4). This increases the variation in brightness gradation of the light regions (dot light DT1-DT3, non-light dot DT4), and further enhances the uniqueness of each pixel block 102 due to the distribution of these light regions (dot light DT1-DT3, non-light dot DT4). In addition, the non-light region (non-light dot DT4) can suppress the overlap of light regions (dot light DT1-DT3) with different frequency bands, and the brightness gradation based on these light regions (dot light DT1-DT3) can be properly maintained. Therefore, stereo correspondence point search can be performed with greater accuracy, and the distance to the surface of object A1 can be measured with greater accuracy.

[0110] <Example of change 1> In the above embodiment, three light sources 31 to 33 corresponding to dot lights DT1 to DT3 are arranged in the projection unit 30, but in modified example 1, only one light source is arranged in the projection unit 30.

[0111] Figure 12 shows the configuration of the distance measuring device 1 according to modification example 1.

[0112] The projection unit 30 comprises a light source 37, a collimator lens 38, a filter 35, and a projection lens 36. The light source 37 emits light in a wavelength band that includes multiple types of selectable wavelength bands 351 to 353. The light source 37 is, for example, a white laser diode. The collimator lens 38 aligns the light emitted from the light source 37. The collimator lens 38 constitutes an optical system that guides the light from the light source 37 to the filter 35. The configuration of the filter 35 and the projection lens 36 is the same as in the above embodiment. Furthermore, the configuration other than the projection unit 30 is the same as the configuration in Figure 2.

[0113] Figure 13(a) is a graph showing the spectral output of the light source 37, and Figure 13(b) is a graph showing the spectral transmittance of the filter region 351-353. Figures 13(c)-(e) are the same as Figures 8(c)-(e).

[0114] If the light source 37 has the spectral output characteristics shown in Figure 13(a) and the filter regions 351 to 353 have the spectral transmittance characteristics shown in Figure 13(b), then if the spectral reflectance of the measurement surface (surface of object A1) and the spectral sensitivity of the first imaging unit 10 and the second imaging unit 20 have the characteristics shown in Figures 13(c) and (d), respectively, then the maximum brightness of the dot lights DT1 to DT3 will differ approximately uniformly in order of brightness, as shown in Figure 13(e).

[0115] Therefore, according to the configuration of Modification Example 1, the maximum brightness of the dot lights DT1 to DT3 can be made to differ from each other simply by emitting light from the light source 37, and these maximum brightnesses can be made to differ approximately equally in order of brightness. Thus, as with the above embodiment, the distance to the surface of object A1 can be measured with high accuracy. In addition, the number of parts in the projection unit 30 can be reduced, and the configuration of the projection unit 30 can be simplified.

[0116] However, in the configuration of Modification Example 1, since there is no light source for each dot light DT1 to DT3, it is not possible to adjust the amount of light from dot lights DT1 to DT3 according to the wavelength dependence of the reflectance of the surface of object A1, as in the above embodiment. Therefore, in order to perform stereo correspondence point search with greater accuracy when the reflectance of the surface of object A1 is wavelength dependent, it is preferable to arrange light sources 31 to 33 for each dot light DT1 to DT3, as in the above embodiment.

[0117] In the configuration of Modification Example 1, the brightness adjustment unit 44 adjusts the amount of light emitted by the light source 37 (drive current value) so that the maximum brightness based on the dot lights DT1 to DT3 does not saturate, and these maximum brightness levels fall within the range of tonal gradations (for example, 0 to 255) defined by the first imaging processing unit 41 and the second imaging processing unit 42. In this case, before distance measurement, the brightness adjustment unit 44 emits light from the light source 37 at its initial value to acquire the second image 200, and obtains the maximum brightness from the second image 200. If the maximum brightness is saturated or too low, the drive current value of the light source 37 is reset based on the relationship between brightness and drive current value so that the maximum brightness is slightly less than the highest tonal gradation. The maximum brightness of the dot lights DT1 to DT3 will fall within the range of tonal gradations (for example, 0 to 255) defined by brightness.

[0118] <Example of change 2> In modification example 2, a light-shielding wall is formed at the boundary between adjacent filter regions on filter 35.

[0119] Figure 14(a) is a schematic diagram showing the configuration of filter 35 in modification example 2. Figure 14(b) is a magnified view of a portion of Figure 14(a).

[0120] As shown in Figures 14(a) and (b), in modification example 2, a light-shielding wall 355 is formed at the boundary between adjacent filter regions on the filter 35. The height of the light-shielding wall 355 is the same as the thickness of the filter regions 351 to 354. The light-shielding wall 355 is formed in a matrix on the glass substrate that constitutes the filter 35. One square in the matrix corresponds to one filter region. The filter regions 351 to 354 are formed on the glass substrate on which the light-shielding wall 355 is formed in this manner by the process described above. This results in the filter 35 with the configuration shown in Figures 14(a) and (b).

[0121] When light-shielding walls 355 are formed in this manner, it is possible to suppress the overlapping of dot light due to leakage when passing through adjacent filter regions, and to generate a good pattern light 30a in which each type of dot light is clearly distinguishable. As a result, stereo correspondence point search can be performed with greater accuracy, and the distance to the surface of object A1 can be measured with greater accuracy.

[0122] <Other examples of changes> In the above modification example 1, one light source 37 having spectral output spanning the wavelength bands of spectral transmittance of three filter regions 351 to 353 was arranged in the projection unit 30, but in the modification example 1, a light source having spectral output spanning the wavelength bands of spectral transmittance of two filter regions 352 and 353 and filter region 351 of A light source having a spectral output corresponding to the wavelength band of spectral transmittance may be arranged, or two filter regions 351, 352 of A light source having a spectral output spanning the wavelength band of spectral transmittance and a light source having a spectral output corresponding to the wavelength band of spectral transmittance in the filter region 353 may be arranged.

[0123] In this case, an optical system that integrates the light from these two light sources and directs it to the filter 35 is arranged in the projection unit 30. The light source having spectral output spanning two spectral transmittance wavelength bands only needs to have spectral output characteristics such that the maximum brightness based on the light from these two wavelength bands differs in the same way as shown in Figure 8(e), and the output of the other light source only needs to be set so that the maximum brightness based on its light differs from the maximum brightness based on the other light in the same way as shown in Figure 8(e).

[0124] Furthermore, in the above embodiment, as shown in Figures 5(a) and (b), four types of filter regions 351 to 354 are arranged in the filter 35, but the types of filter regions arranged in the filter 35 are not limited to these. For example, two types of filter regions may be arranged in the filter 35, or five or more types of filter regions may be arranged in the filter 35.

[0125] In this case, multiple light sources may be arranged so that there is a one-to-one correspondence between the types of filter regions, or multiple light sources having spectral outputs corresponding to the spectral transmittances of multiple types of filter regions may be arranged. That is, the number of light sources may be set to be less than the number of types of filter regions, and dot light of different wavelength bands may be generated from multiple types of filter regions based on light from a single light source. In this case as well, the spectral output of each light source and the spectral transmittance of each filter region should be set so that the maximum brightness of the dot light generated by each type of filter region is different from one another. More preferably, the spectral output of each light source and the spectral transmittance of each filter region should be set so that the maximum brightness of these dot lights differs substantially equally in order of brightness.

[0126] Furthermore, the various spectral characteristics are not limited to those shown in Figures 8(a)-(d), 9(a)-(d), 10(a)-(d), and 13(a)-(d). The spectral output of each light source and the spectral transmittance of each filter region can be appropriately changed, as long as the maximum brightness of the dot light generated by each filter region differs from one another. The wavelength bands of each light source and each type of filter region are also not limited to those shown in the above embodiments and their modifications.

[0127] Furthermore, the arrangement patterns of each type of filter region are not limited to those shown in Figures 5(a) and (b), but can be changed as appropriate. In this case as well, it is sufficient that the arrangement patterns of each type of filter region are set such that, at least within the search range R0, the arrangement pattern of each type of dot light in each pixel block is unique (random).

[0128] Furthermore, although a transmissive filter 35 was exemplified in the above embodiment and its modifications, a reflective filter may also be used. In this case, for example, a reflective film is formed between the glass substrate constituting the filter 35 and the material layer forming each filter region.

[0129] Furthermore, in the above embodiment and its modified examples, multiple types of optical regions with different wavelength bands were represented as dot optical regions DT1 to DT3. However, these optical regions do not necessarily have to be dots. As long as there is some uniqueness (randomness) in the distribution pattern of the optical regions for each pixel block within the search range R0, the multiple types of optical regions may have shapes other than dots.

[0130] Furthermore, in the above embodiment, in step S103 of Figure 11, the surface of object A1 was imaged by the second imaging unit 20. However, in step S103 of Figure 11, the surface of object A1 may be imaged by the first imaging unit 10, and the maximum brightness acquisition process in step S104 may be performed using the first image 100 acquired by the first imaging unit 10.

[0131] Furthermore, in the above embodiment and its modifications, two imaging units, the first imaging unit 10 and the second imaging unit 20, were used, but three or more imaging units may be used. In this case, these imaging units are arranged so that their fields of view overlap, and pattern light 30a is projected onto the overlapping area of ​​their fields of view. Stereo correspondence point search is performed between the pair of imaging units.

[0132] Furthermore, the usage of the distance measuring device 1 is not limited to the usage shown in Figure 1 or the usage where it is installed on the end effector of a robot arm, but may also be used in other systems that perform predetermined control using the distance to the surface of an object. Also, the configuration of the distance measuring device 1 is not limited to the configuration shown in the above embodiment, but for example, a photosensor array in which multiple photosensors are arranged in a matrix may be used as the image sensors 12 and 22.

[0133] In addition, the embodiments of the present invention can be modified in various ways as appropriate within the scope of the technical idea set forth in the claims. [Explanation of Symbols]

[0134] 1. Distance measuring device 10 First Imaging Section 10a, 20a field of view 20 Imaging Section 2 30 Projection part 30a Patterned Light 31~33, 37 light source 34 Optical system 35 filters 38. Collimator lens (optical system) 44 Brightness adjustment section 45 Measurement section 351-354 Filter region DT1~DT3 Dot Light (Light Region) DT4 Non-illuminated dot (non-illuminated area)

Claims

1. The system consists of a first imaging unit and a second imaging unit, each equipped with a monochrome image sensor, arranged side by side so that their fields of view overlap, A projection unit projects patterned light, in which multiple types of light regions with different wavelength bands and areas without light are distributed in a predetermined pattern, onto the overlapping area of ​​the field of view. The system includes a measuring unit that performs stereo correspondence point search processing on images acquired by the first imaging unit and the second imaging unit, respectively, to identify the second pixel block in the other image whose pixel value best matches that of each first pixel block in one of the images, and measures the distance to the surface of the object onto which the pattern light is projected by triangulation based on the pixel misalignment between the first and second pixel blocks. A distance measuring device characterized by the following features.

2. In the distance measuring device according to claim 1, The maximum brightness differs among the aforementioned multiple types of light regions. A distance measuring device characterized by the following features.

3. In the distance measuring device according to claim 1 or 2, The projection unit includes a filter in which multiple types of filter regions for generating each of the multiple types of light regions are distributed in a pattern similar to the pattern of the light region. A distance measuring device characterized by the following features.

4. In the distance measuring device according to claim 3, The projection unit is, Multiple light sources that emit light in different wavelength bands, The system comprises an optical system that guides light emitted from the plurality of light sources to the filter, A distance measuring device characterized by the following features.

5. In the distance measuring device according to claim 4, The multiple light sources are arranged to correspond to each of the multiple types of filter regions. Each of the aforementioned filter regions selectively extracts light from the corresponding light source. A distance measuring device characterized by the following features.

6. In the distance measuring device according to claim 5, The maximum brightness values ​​obtained based on the light from each of the light sources, which are acquired based on the pixel signals from the first imaging unit or the second imaging unit, are different from each other. A distance measuring device characterized by the following features.

7. In the distance measuring device according to claim 5, The system includes a brightness adjustment unit that adjusts the amount of light emitted from each of the aforementioned multiple light sources, The brightness adjustment unit sets the amount of light emitted from the plurality of light sources such that the maximum brightness based on the light from each of the light sources acquired based on the pixel signals from the first imaging unit or the second imaging unit is different from that of the others. A distance measuring device characterized by the following features.

8. In the distance measuring device according to claim 7, The brightness adjustment unit sets the amount of light emitted from the plurality of light sources such that the maximum brightness based on the light from each of the light sources acquired based on the pixel signals from the first imaging unit or the second imaging unit differs substantially equally in order of brightness magnitude. A distance measuring device characterized by the following features.

9. In the distance measuring device according to claim 3, The projection unit is, A light source that emits light in a wavelength band including the selected wavelength band of the multiple types of filter regions, The system comprises an optical system that guides light from the light source to the filter, The light source has a spectral output such that the maximum brightness of the light transmitted through the multiple types of filter regions differs from one another. A distance measuring device characterized by the following features.

10. In the distance measuring device according to claim 4, The light source is a light-emitting diode. A distance measuring device characterized by the following features.

Images