A height information detection method and detection device of multi-structured light projection

Abstract

The application discloses a kind of height information detection methods of multi-structure light projection, comprising the following steps: the object to be measured is placed in the field of view range of camera, multiple projection light sources are started in turn, and multiple height information matrices M1-Mn are obtained;Wave crest luminance proportion matrix G1-Gn is calculated;Wave crest score matrix GS1-GSn is calculated, wherein GSn=Gn / Max{G1,G2,G3,G4…Gn};Height difference matrix D1-Dn is calculated;Height difference score matrix DS1-DSn is calculated;Set weighted proportion a, b, calculate the height quality score matrix S1-Sn of each projection light source, wherein Sn=a*GSn+b*DSn, a+b=1;Set height quality threshold T;The average value of height information matrix greater than or equal to T in S1-Sn is taken to obtain target height information matrix.The application also discloses a kind of height information detection devices of multi-structure light projection.The application proposes the height information detection method and device of multi-structure light projection with better applicability, integrity and accuracy;The accuracy of height is maintained.

Description

Technical Field

[0001] This invention belongs to the field of three-dimensional scanning technology, and in particular relates to a method and device for detecting height information of multi-structured light projection. Background Technology

[0002] 3D scanning technology is widely used in fields such as human-computer interaction, object recognition, facial recognition, product quality inspection, and 3D modeling. Classified by the basic physical principles of the measurement methods, they mainly include triangulation, time-of-flight measurement, focusing height measurement, confocal measurement, and interferometry.

[0003] Structured light technology using area arrays is a popular 3D scanning technology in recent years. Compared with point structured light and line structured light, its advantage is that it can obtain the height information of a surface without the need for mechanical motion scanning, and the system integration requirements are low.

[0004] Most existing structured light sensors use a monocular approach, consisting of a light source emitter and a receiver, with the receiver and emitter on the same horizontal axis, as shown in the schematic diagram. Figure 1-1 As shown.

[0005] The structured light emitted by the transmitter is typically phase-shifted sinusoidal fringes or linear fringes. Taking linear fringes as an example... Figure 1-2 The image shown is one of a series of straight stripe images. For each pixel in the receiver, the grayscale values ​​of the multiple straight stripe images form a pulse waveform, as shown below. Figure 1-3 By fitting the peak position of the pulse, the corresponding transmitter pixel can be decoded. Then, combining the spatial geometry of the transmitter and receiver with the lens calibration parameters, the pixel height can be reconstructed. For most diffuse reflective surfaces, this method allows the receiver to correctly receive the structured light information returned by the measured surface and reconstruct the height information. However, in some scenarios, such as projection onto the edge of an object or projection occlusion, the receiver cannot accurately obtain the structured light information returned by the measured surface. In these cases, the reconstructed height will be inaccurate or invalid.

[0006] Taking the edge as an example, for such Figure 1-4 As shown, at the edge of the object, straight stripes are projected, with brightness on the inner side (right side) of the edge and no brightness on the outer side (left side), as... Figure 1-5 As shown. Therefore, a sudden drop in gray level on one side of the grayscale waveform peak causes a deviation in the fitted peak position, as shown. Figure 1-6 As shown, the height of the reconstructed pixels also has a deviation.

[0007] Therefore, when faced with complex shooting scenarios, the integrity and accuracy of height information of a monocular structured light sensor often fail to meet application requirements. Summary of the Invention

[0008] To overcome the shortcomings of the prior art, the present invention provides a method and device for detecting height information of multi-structured light projection with multiple light sources arranged at multiple angles, which has higher applicability, better integrity and higher accuracy.

[0009] The technical solution adopted by this invention to solve its technical problem is: a method for detecting height information of multi-structured light projection, comprising the following steps:

[0010] Place the object to be tested within the field of view of the camera, and activate multiple projection light sources in sequence to obtain multiple height information matrices M1 to Mn;

[0011] Calculate the peak brightness ratio matrix G1~Gn for each projection light source;

[0012] Calculate the peak score matrix GS1~GSn for each projection light source, where GSN=Gn / Max{G1,G2,G3,G4…Gn};

[0013] Calculate the height difference matrix D1~Dn for each projection light source, where D1=Min{|M1-M2|,|M1-M3|,|M1-M4|…|M1-Mn|}; where Dm=Min{|Mm-M1|,|Mm-M2|,…,|Mm-Mm-1|,|Mm-Mm+1|,…,|Mm-Mn|}, m≠1, and m≠n; where Dn=Min{|Mn-M1|,|Mn-M2|,…|Mn–Mn-1|};

[0014] Calculate the height difference scoring matrix DS1~DSn for each projection light source, where DSn=Min{D1,D2,D3,D4…Dn} / Dn;

[0015] Set weighting ratios a and b, and calculate the height quality score matrix S1~Sn for each projection light source, where Sn=a*GSn+b*DSn, a+b=1;

[0016] Set a height quality threshold T;

[0017] Take the average value of the height information matrices greater than or equal to T from S1 to Sn to obtain the target height information matrix.

[0018] Furthermore, the peak brightness ratio matrix is ​​calculated as follows: peak brightness ratio of each pixel = maximum phase shift fringe brightness / total phase shift fringe brightness.

[0019] Furthermore, the values ​​of the elements GS1 to GSN range from 0 to 1.0.

[0020] Furthermore, the values ​​of the elements DS1 to DSn range from 0 to 1.0.

[0021] Furthermore, the values ​​of the elements S1 to Sn range from 0 to 1.0.

[0022] Furthermore, in a single region, among the multiple height information matrices M1 to Mn, only one is valid height information, while the rest are invalid.

[0023] Furthermore, the method includes the following steps: Before calculating the peak brightness ratio matrix G1 to Gn for each projection light source, the validity of the region height information matrix is ​​manually judged. If the region height information matrix is ​​invalid, the region mask setting is increased to eliminate interference.

[0024] Furthermore, the mask setting step involves setting one or more rectangular wrapping regions to wrap the regions where the height information matrix is ​​invalid.

[0025] This invention also discloses a height information detection device for multi-structured light projection, comprising:

[0026] The camera is positioned above the surface of the object being measured.

[0027] Multiple projection light sources are positioned at the same horizontal level as the camera and spaced apart around the camera.

[0028] The control unit controls the activation and deactivation of multiple projection light sources;

[0029] The output unit is used to output the height information matrix corresponding to multiple projection light sources and the target height information matrix.

[0030] The calculation unit is used to calculate the peak score matrix and height difference score matrix for each projection light source to eliminate invalid height information;

[0031] The peak rating matrix is ​​GS1~GSn, where GSn=Gn / Max{G1,G2,G3,G4…Gn}, and G1~Gn is the peak brightness ratio matrix of each projection light source;

[0032] The height difference scoring matrix is ​​DS1~DSn, where DSn=Min{D1,D2,D3,D4…Dn} / Dn, D1~Dn is the height difference matrix of each projection light source, where D1=Min{|M1-M2|,|M1-M3|,|M1-M4|…|M1-Mn|}; where Dm=Min{|Mm-M1|,|Mm-M2|,…,|Mm-Mm-1|,|Mm-Mm+1|,…,|Mm-Mn|}, m≠1, and m≠n; where Dn=Min{|Mn-M1|,|Mn-M2|,…|Mn–Mn-1|}.

[0033] Furthermore, there are four projection light sources, which are evenly spaced along the outer periphery of the camera.

[0034] The beneficial effects of this invention are: 1) It proposes a height information detection method and device for multi-structured light projection that can remove malfunctions caused by occlusion and weak light reception through a peak scoring matrix, and remove malfunctions caused by multiple reflections and edge errors through a height scoring matrix. Finally, by weighted summation of the two scores, it proposes a method and device for height information detection with better applicability, completeness and accuracy; 2) When the weighted scoring is interfered with by complex factors such as projection occlusion and projection angle, and when the number of projection light sources with correct height information is small, a mask setting can be added to eliminate interference, thus maintaining the accuracy of the height. Attached Figure Description

[0035] Figure 1-1 This is a schematic diagram of a monocular structure in the prior art.

[0036] Figure 1-2 This is a schematic diagram of linear stripe structured light in the prior art.

[0037] Figure 1-3 This is a schematic diagram of a pulse waveform formed by the grayscale values ​​of multiple straight-line stripe images in the prior art.

[0038] Figure 1-4 This is a schematic diagram of a straight stripe projected onto the inner edge of a bright area in the prior art.

[0039] Figure 1-5 This is a schematic diagram of straight stripes projected onto the outer edge of the existing technology with no brightness.

[0040] Figure 1-6 A schematic diagram illustrating the deviation in the fitted peak position present in existing technologies.

[0041] Figure 2 This is a three-dimensional structural diagram of the camera and projection light source in the detection device of the present invention.

[0042] Figure 3 This is a schematic diagram of the camera and projection light source in the detection device of the present invention from a low angle.

[0043] Figure 4 This refers to the grayscale image of the object under test captured by individually activating four projection light sources in this invention.

[0044] Figure 5 This is a schematic diagram of the height matrix of the object under test when four projection light sources are turned on individually in this invention.

[0045] Figure 6 This is a schematic diagram of the height matrix after removing the height of the projection light source 21 when only the peak rating matrix is ​​used.

[0046] Figure 7 This is a schematic diagram of the height matrix after removing the height of the projection light source 22 when only the height difference scoring matrix is ​​used.

[0047] Figure 8 This is a schematic diagram of the height matrix using weighted scoring in the detection method of the present invention.

[0048] Figure 9 This is a schematic diagram of the height matrix using weighted scoring and masking methods in the detection method of the present invention. Detailed Implementation

[0049] To enable those skilled in the art to better understand the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0050] A height information detection device for multi-structured light projection, comprising:

[0051] Camera 1 is positioned above the surface of the object being measured;

[0052] Multiple projection light sources 2 are set at the same horizontal height as camera 1 and are spaced out around camera 1.

[0053] The control unit controls the opening and closing of multiple projection light sources 2;

[0054] The output unit is used to output the height information matrix corresponding to multiple projection light sources 2 and the target height information matrix;

[0055] The calculation unit is used to calculate the peak score matrix and height difference score matrix for each projection light source 2, thereby eliminating invalid height information.

[0056] A method for detecting height information of multi-structured light projection, based on the aforementioned detection device, includes the following steps:

[0057] Place the object to be measured within the field of view of the camera, and activate multiple projection light sources in sequence to obtain multiple height information matrices M1 to Mn; for a single region, only one of the multiple height information matrices M1 to Mn is valid, and the rest of the height information is invalid;

[0058] The validity of the region height information matrix is ​​manually judged. If the region height information matrix is ​​invalid, a mask is added to the region to eliminate interference. The mask setting step involves setting one or more rectangular wrapping regions to wrap the regions with invalid height information matrices.

[0059] Calculate the peak brightness ratio matrix G1~Gn for each projection light source; the above peak brightness ratio matrix is ​​calculated as follows: peak brightness ratio of each pixel = maximum phase shift fringe brightness / total phase shift fringe brightness.

[0060] Calculate the peak score matrix GS1~GSn for each projection light source, where GSn=Gn / Max{G1,G2,G3,G4…Gn}; where the values ​​of the elements of GS1~GSn range from 0 to 1.0;

[0061] Calculate the height difference matrix D1~Dn for each projection light source, where D1=Min{|M1-M2|,|M1-M3|,|M1-M4|…|M1-Mn|}; where Dm=Min{|Mm-M1|,|Mm-M2|,…,|Mm-Mm-1|,|Mm-Mm+1|,…,|Mm-Mn|}, m≠1, and m≠n; where Dn=Min{|Mn-M1|,|Mn-M2|,…|Mn–Mn-1|};

[0062] Calculate the height difference scoring matrix DS1~DSn for each projection light source, where DSn=Min{D1,D2,D3,D4…Dn} / Dn; and the values ​​of the elements of DS1~DSn range from 0 to 1.0.

[0063] Set weighting ratios a and b, and calculate the height quality score matrix S1 to Sn for each projection light source, where Sn = a*GSn + b*DSn, a + b = 1; and the values ​​of the elements in S1 to Sn range from 0 to 1.0.

[0064] Set a height quality threshold T;

[0065] Take the average value of the height information matrices greater than or equal to T from S1 to Sn to obtain the target height information matrix.

[0066] Specifically, such as Figure 2 , Figure 3 As shown, the example is that there are four projection light sources 2, which are evenly spaced along the outer periphery of the camera 1.

[0067] Place the object to be tested within the field of view of camera 1, and activate four projection light sources 2 in sequence to obtain four height information matrices M1 to M4;

[0068] like Figure 4 As shown, a, b, c, and d are grayscale images captured when the four projection light sources 2 (projection light source 21, projection light source 22, projection light source 23, and projection light source 24) are turned on individually in sequence. Figure 5The height matrix is ​​captured when the four projection light sources 2 are turned on individually. The height range (-1, 1) is represented by a color transitioning from blue to green to red, while black indicates invalid height and white indicates height outside the range (-1, 1).

[0069] The true shape of the object to be measured is obtained using a three-dimensional coordinate instrument and compared with the previously obtained true shape. Figure 5 The data shows that when projection light source 21 is turned on, the height at mark 1 does not match the actual shape. Mark 1 is a notch, and its actual height should be around -0.5, but the obtained height is greater than 1. Therefore, the height obtained by projection light source 21 at mark 1 is incorrect. When projection light source 22 is turned on, the heights at marks 2 and 3 do not match the actual shape. The actual heights at marks 2 and 3 should both be around 0.58, but the height obtained at mark 2 is around 1.35 and the height obtained at mark 3 is around 0.78. Therefore, the heights obtained by projection light source 22 at marks 2 and 3 are incorrect.

[0070] In this case, marker 1 is valid in the height matrix of both projection light source 21 and projection light source 23. However, projection light source 21 is in shadow at this location, hence its peak score is low. Specifically, the peak brightness ratios of projection light source 21 and projection light source 23 at this location are G1 = 0.15 and G3 = 0.69, respectively. Therefore, their peak scores are GS1 = G1 / G3 = 0.22 and GS3 = G3 / G3 = 1 (projection light sources 22 and 24 are invalid, so GS2 = 0 and GS4 = 0).

[0071] The heights of projection light source 21 and projection light source 23 at this location are: M1 = 1.117 and M3 = -0.495, respectively; the height difference values ​​are: D1 = |M1 - M3| = 1.612 and D3 = |M3 - M1| = 1.612, respectively; the height difference scores are: DS1 = D1 / D1 = 1 and DS3 = D1 / D3 = 1 (the heights of projection light source 22 and projection light source 24 are invalid, so DS2 = 0 and DS4 = 0). Using weights a = 1, b = 0, and a threshold T = 0.61, where the threshold T = 1 / 2(GS1 + GS3), then S1 = a*GS1 + b*DS1 = 0.22, S3 = a*GS3 + b*DS3 = 1, i.e., S1 < T, S3 > T. Therefore, the height matrix at mark 1 is taken as the height information matrix M3 obtained by the projection light source 23. Thus, when calculating the target height matrix, only the peak score matrix is ​​used to remove the height of the projection light source 21. The result is as follows... Figure 6 .

[0072] Compared to the pre-obtained true shape, marker 2 is valid in the height matrix of all projections. However, projection light source 22 is affected by multiple reflections at this location, resulting in an incorrect height, but its peak score is the highest, so it cannot be evaluated using the peak score matrix. Nevertheless, the height of projection light source 22 differs greatly from the heights of other projections at this location, so only the height difference score can be used to remove the height of projection light source 22. Specifically, the peak brightness ratios of projection light sources 21 to 24 at this location are: G1 = 0.32, G2 = 0.87, G3 = 0.29, G4 = 0.37, respectively. Therefore, the peak scores are: GS1 = G1 / G2 = 0.368, GS2 = G2 / G2 = 1, GS3 = G3 / G2 = 0.333, GS4 = G4 / G2 = 0.425. The heights of projection light sources 21 to 24 at this point are: M1 = 0.563, M2 = 1.351, M3 = 0.586, M4 = 0.584. Therefore, the height difference values ​​are: D1 = |M1 - M4| = 0.021, D2 = |M2 - M3| = 0.765, D3 = |M3 - M4| = 0.002, D4 = |M4 - M3| = 0.002. The height difference scores are: DS1 = D3 / D1 = 0.095, DS2 = D3 / D2 = 0.003, DS3 = D3 / D3 = 1, DS4 = D3 / D4 = 1. Using weights a=0, b=1, and a threshold T=0.52, where threshold T=1 / 4(DS1+DS2+DS3+DS4), then S1=a*GS1+b*DS1=0.095, S2=a*GS2+b*DS2=0.003, S3=a*GS3+b*DS3=1, S4=a*GS4+b*DS4=1, i.e., S1<T, S2<T, S3>T, S4>T. Therefore, the height matrix at mark 2 is the average of the height information matrices obtained by projection light sources 23 and 24, and the result is as follows: Figure 7 However, it can be seen that it cannot reasonably evaluate point 1, because only two projection heights are valid at point 1, and the height difference score cannot be reasonably evaluated.

[0073] In summary, when calculating the target height matrix, using only the peak rating matrix can obtain the correct height information matrix at mark 1, but the height information matrix at mark 2 is incorrect; similarly, using only the height difference rating matrix can obtain the correct height information matrix at mark 2, but the height information matrix at mark 1 is incorrect. Therefore, using a single rating matrix cannot optimally obtain the target height matrix. Using the weighted rating proposed in this invention, specifically using weight a = 0.5 and weight b = 0.5, then at mark 1:

[0074] S11=0.5*GS1+0.5*DS1=0.5*0.22+0.5*1=0.61;

[0075] S12 = 0;

[0076] S13=0.5*GS3+0.5*DS3=0.5*1+0.5*1=1;

[0077] S14 = 0;

[0078] Mark 2:

[0079] S21=0.5*GS1+0.5*DS1=0.5*0.368+0.5*0.095=0.232;

[0080] S22=0.5*GS2+0.5*DS2=0.5*1+0.5*0.003=0.502;

[0081] S23=0.5*GS3+0.5*DS3=0.5*0.333+0.5*1=0.667;

[0082] S24=0.5*GS4+0.5*DS4=0.5*0.425+0.5*1=0.713;

[0083] The threshold T = 1 / 2(S11 + S23), that is, T = 0.64;

[0084] Then, S11<T, S13>T, S21<T, S22<T, S23>T, S24>T;

[0085] The height matrix at mark 1 is taken as the height information matrix M3 obtained by projection light source 23, and the height matrix at mark 2 is taken as the average of the height information matrices obtained by projection light sources 23 and 24. This allows for a reasonable assessment of the heights at both marks 1 and 2 simultaneously, as shown in the following results. Figure 8 .

[0086] However, it can be seen that the measurement at mark 3 still cannot be reasonably evaluated. It uses the incorrect height of projection light source 22, while the height of projection light source 24 is correct. Because mark 3 is the edge of the part, and the edge direction is parallel to the projected stripes of projection light source 22, the stripe grayscale will change abruptly, causing a deviation in the fitted peak position, and ultimately resulting in a deviation in the reconstructed pixel height (e.g., ...). Figure 5(Higher than the surface of the part). Specifically, the height of projection light sources 21 and 23 is ineffective at this point. The peak brightness ratios of projection light sources 22 and 24 at this point are G2 = 0.38 and G4 = 0.36, respectively, so the peak scores are GS2 = G2 / G2 = 1 and GS4 = G4 / G2 = 0.95. The heights of projection light sources 22 and 24 at this point are M2 = 0.778 and M4 = 0.589, respectively, so the height difference values ​​are D2 = |M2-M3| = 0.189 and D4 = |M4-M3| = 0.189. So the height difference scores are DS2 = D2 / D2 = 1 and DS4 = D2 / D4 = 1. Using weights a = 0.5 and b = 0.5, mark 3 locations:

[0087] S2=0.5*GS2+0.5*DS2=0.5*1+0.5*1=1;

[0088] S4=0.5*GS4+0.5*DS4=0.5*0.95+0.5*1=0.975;

[0089] Therefore, S2>S4. For any threshold T, it is impossible to exclude the height of projection light source 22 while retaining the height of projection light source 24. Thus, this type of error cannot be evaluated using the aforementioned scoring method. Therefore, the masking method proposed in this invention is used. A mask region is set at mark 3, specifically X=2240, Y=1210, Width=80, Height=400, to mask the height of projection 2. The height matrix at mark 3 is then taken as the height information matrix M4 of projection light source 24. The final result is as follows. Figure 9 Its height matrix matches the actual object.

[0090] The above specific embodiments are used to explain and illustrate the present invention, but not to limit the present invention. Any modifications and changes made to the present invention within the spirit and scope of the claims shall fall within the protection scope of the present invention.

Claims

1. A method for detecting height information using multi-structured light projection, characterized in that, Includes the following steps: Place the object to be tested within the field of view of the camera, and activate multiple projection light sources in sequence to obtain multiple corresponding height information matrices M1~Mn; Calculate the peak brightness ratio matrix G1~Gn for each projection light source; The peak brightness ratio matrix is ​​calculated as follows: peak brightness ratio of each pixel = maximum phase shift fringe brightness / total phase shift fringe brightness. Calculate the peak score matrix GS1~GSn for each projection light source, where GSN=Gn / Max{G1, G2, G3, G4…Gn}; Calculate the height difference matrix D1~Dn for each projection light source, where D1 = Min{|M1 - M2|, |M1 - M3|, |M1 - M4|…|M1 - Mn|}; where Dm = Min{|Mm - M1|, |Mm - M2|, …,|Mm - Mm-1|,|Mm -Mm+1|, …,|Mm - Mn|}, m≠1 and m≠n; where Dn = Min{|Mn - M1|, |Mn - M2|, …|Mn –Mn-1|}; Calculate the height difference scoring matrix DS1~DSn for each projection light source, where DSn = Min{D1, D2, D3, D4…Dn} / Dn; Set weighting ratios a and b, and calculate the height quality rating matrix S1~Sn for each projection light source, where Sn = a*GSn + b*DSn, a+b = 1; Set a height quality threshold T; Take the average value of the height information matrices greater than or equal to T from S1 to Sn to obtain the target height information matrix.

2. The height information detection method for multi-structured light projection according to claim 1, characterized in that: The values ​​of the elements GS1 to GSn range from 0 to 1.

0.

3. The height information detection method for multi-structured light projection according to claim 1, characterized in that: The values ​​of the elements DS1 to DSn range from 0 to 1.

0.

4. The height information detection method for multi-structured light projection according to claim 1, characterized in that: The values ​​of the elements S1 to Sn range from 0 to 1.

0.

5. The height information detection method for multi-structured light projection according to claim 1, characterized in that: For a single region, among multiple height information matrices M1~Mn, only one is valid height information, and the rest are invalid.

6. The height information detection method for multi-structured light projection according to claim 1, characterized in that, It also includes the following steps: Before calculating the peak brightness ratio matrix G1~Gn for each projection light source, the validity of the region height information matrix is ​​manually judged. If the region height information matrix is ​​invalid, the region mask setting is increased to eliminate interference.

7. The height information detection method for multi-structured light projection according to claim 6, characterized in that: The mask setting step involves setting one or more rectangular wrapping regions to wrap the regions where the height information matrix is ​​invalid.

8. A height information detection device for multi-structured light projection, characterized in that, include: The camera is positioned above the surface of the object being measured. Multiple projection light sources are positioned at the same horizontal level as the camera and spaced apart around the camera. The control unit controls the activation and deactivation of multiple projection light sources; The output unit is used to output the height information matrices M1~Mn corresponding to multiple projection light sources and the target height information matrix; The calculation unit is used to calculate the peak score matrix and height difference score matrix for each projection light source to exclude invalid height information; the peak score matrix is ​​GS1~GSn, where GSn=Gn / Max{G1, G2, G3, G4…Gn}, and G1~Gn is the peak brightness ratio matrix for each projection light source; the peak brightness ratio matrix is ​​calculated as follows: peak brightness ratio of each pixel = maximum phase shift fringe brightness / total phase shift fringe brightness. The height difference scoring matrix is ​​DS1~DSn, where DSn = Min{D1, D2, D3, D4…Dn} / Dn, and D1~Dn is the height difference matrix of each projection light source, where D1 = Min{|M1 - M2|, |M1 - M3|, |M1 - M4|…|M1 - Mn|}; where Dm = Min{|Mm - M1|, |Mm - M2|, …,|Mm - Mm-1|,|Mm - Mm+1|, …,|Mm - Mn|}, m≠1, and m≠n; where Dn = Min{|Mn - M1|, |Mn - M2|, …|Mn – Mn-1|}; Set weighting ratios a and b, and calculate the height quality score matrix S1~Sn for each projection light source, where Sn = a*GSn + b*DSn, a+b = 1; set a height quality threshold T; take the average value of the height information matrices greater than or equal to T in S1~Sn to obtain the target height information matrix.

9. The height information detection device for multi-structured light projection according to claim 8, characterized in that: There are four projection light sources, which are evenly spaced along the outer perimeter of the camera.

Images