Different-spectrum structured light measurement method for rapid region mapping

Abstract

The invention discloses a different-spectrum structured light measurement method for rapid region mapping, which can realize acquisition of a region contour of a measured object only according to a square pattern, and further complete region coordinate mapping of a projection equipment surface. Compared with a current method for obtaining the phase through transverse and longitudinal stripes, a large number of stripe images are not needed for coordinate conversion any more, efficiency can be remarkably improved, and rapid measurement of a dynamic scene is achieved. Different spectrums are set in different areas for simultaneous projection, and compared with a current method of sequentially projecting different areas, the measurement efficiency can be obviously improved. And feasibility is provided for dynamic measurement of a mutual reflection scene.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of three-dimensional measurement, and in particular to a fast region mapping heterospectral structured light measurement method. BACKGROUND

[0002] Structured light three-dimensional measurement technology is widely used in industry, medicine, entertainment and other fields due to its non-contact, high precision, high speed and other advantages. However, when the measured scene contains both diffuse reflection and specular reflection objects, the traditional structured light measurement method has limitations. Because the reflected light on the surface of the specular reflection object at some angles will exactly illuminate the surface of the diffuse reflection object, at this time the mutual reflection between objects occurs, causing the structured light image collected by the image acquisition device to have stripe aliasing interference, affecting the accuracy of phase extraction and three-dimensional reconstruction.

[0003] To overcome this problem, the method of dividing the projection area is currently used. However, implementing region positioning requires projecting horizontal and vertical stripes on the surface of the measured object in advance to obtain the phase in two directions of the measured object, and then mapping the corresponding object region on the surface of the projection device according to the phase. When the position of the measured object moves, a large number of horizontal and vertical stripes need to be projected again to obtain the region contour, so this method cannot meet the dynamic demand. SUMMARY

[0004] The purpose of the present application is to overcome the shortcomings of the prior art and provide a fast region mapping heterospectral structured light measurement method that can significantly improve efficiency, reduce phase error, and be suitable for high-precision measurement in dynamic scenes.

[0005] To achieve the above purpose, the technical solution provided by the present application is as follows:

[0006] A fast region mapping heterospectral structured light measurement method, comprising:

[0007] S1, projecting a preset multi-color checkerboard pattern onto the surface of a target object by a projection device;

[0008] S2, acquiring a multi-color checkerboard image modulated by the surface of the object using an image acquisition device;

[0009] S3, using RGB three-channel separation technology and combining color coding to identify multiple color blocks; then, marking feature points according to a preset rule, and establishing an accurate mapping relationship between the coordinate system of the projection device and the coordinate system of the image acquisition device through the spatial distribution relationship of the feature points;

[0010] S4, dynamically generating a binary projection mask matched with the geometric features of the surface of the object according to a feature matching algorithm;

[0011] S5, divide the projection mask area into two groups according to the spatial position relationship, and generate longitudinal sinusoidal phase shift fringe patterns in the two groups of mask areas;

[0012] S6, encode the two groups of sinusoidal phase shift fringe patterns into the red and blue channels of the RGB image respectively, and reserve the green channel for color balance to form the heterospectral phase shift fringe pattern;

[0013] S7, project the encoded fringe image to the target object through the HDMI interface at a set refresh rate, and synchronously trigger the image acquisition device to capture images;

[0014] S8, use the collected heterospectral phase shift fringe pattern to extract phase and reconstruct three-dimensional topography.

[0015] Further, the multicolor square pattern projected by the projection device to the surface of the target object is a seven-color square pattern, which has an m×n pixel resolution, and is divided into seven color domain blocks in the vertical direction according to the equal division principle, and each block has a height of pixels.

[0016] The color domain arrangement order and color composition of the seven-color square pattern are as follows:

[0017] The first color domain block: pure red (R, RGB value [255, 0, 0]);

[0018] The second color domain block: pure green (G, RGB value [0, 255, 0]);

[0019] The third color domain block: pure blue (B, RGB value [0, 0, 255]);

[0020] The fourth color domain block: red-green mixed color (RG / Yellow, RGB value [255, 255, 0]);

[0021] The fifth color domain block: red-blue mixed color (RB / Magenta, RGB value [255, 0, 255]);

[0022] The sixth color domain block: green-blue mixed color (GB / Cyan, RGB value [0, 255, 255]);

[0023] The seventh color domain block: red-green-blue mixed color (RGB / White, RGB value [255, 255, 255]).

[0024] Further, in step S3, the RGB three-channel separation technology is adopted, and the seven color blocks are identified by combining color coding, and the process includes:

[0025] The image is RGB three-channel separated, and red channel (R), green channel (G) and blue channel (B) are extracted to generate three binary images; based on the three binary images, each The color gamut block with high pixel can be uniquely identified by its corresponding 3-bit binary code, wherein the pure red (R) block is coded as 100 (R=1, G=0, B=0), the pure green (G) block is coded as 010 (R=0, G=1, B=0), the pure blue (B) block is coded as 001 (R=0, G=0, B=1), the red-green mixed (RG / Yellow) block is coded as 110 (R=1, G=1, B=0), the red-blue mixed (RB / Magenta) block is coded as 101 (R=1, G=0, B=1), the green-blue mixed (GB / Cyan) block is coded as 011 (R=0, G=1, B=1), and the red-green-blue mixed (RGB / White) block is coded as 111 (R=1, G=1, B=1); by scanning the 3-bit code of the image pixel and matching with the preset code sequence, the color gamut block to which the feature point belongs can be accurately located.

[0026] Further, in step S3, the feature points are marked according to the preset rule, and the accurate mapping relationship between the projection device coordinate system and the image acquisition device coordinate system is established through the spatial distribution relationship of the feature points, and the process includes:

[0027] The RGB three-channel separation technology is used to divide the seven-color checkerboard pattern with m*n pixels into seven equal height areas (the height of each area is pixels), and each area is constructed into a regular checkerboard matrix by alternately filling the main color and 8*8 pixel black squares. Each color block area (containing checkerboard blocks) is subdivided into 7 sub-areas (each area contains checkerboard blocks) along the row direction, and a circular feature marker point is set at the center position of each sub-area. The color configuration of the marker point needs to meet the following requirements: 1) one of the seven colors coded above, 2) seven marker points corresponding to the features of the remaining seven main color areas, 3) the color arrangement order strictly follows the spatial distribution rule of the seven color areas from top to bottom; the center coordinates of the feature marker points are normalized by the checkerboard blocks, and the first marker point is located at the geometric center of the first sub-area with coordinates The subsequent same-color marker points are equally distributed in the horizontal direction with checkerboard blocks (equivalent to pixel points in the horizontal direction), and equally distributed in the vertical direction with checkerboard blocks (equivalent to pixel points in the vertical direction), and the center coordinates of the feature points in the first color block area are where k is the result of color coding, which is equal to the binary number corresponding to the color region in value, and the seven feature points in the first color block correspond to the seven coordinates one by one.

[0028] Further, the process of step S4 includes:

[0029] First, the multicolor grid image modulated by the surface of the object obtained by the image acquisition device is sequentially separated into red, green and blue color channels to obtain I R , I G , I B , I R , I G , I B , which are binarized and encoded in RGB order to obtain sequentially encoded I R , I G , I B The three-digit code corresponds to the color block region one by one, so as to obtain the horizontal coordinate of the feature point, where k1 is the decimal color coding value corresponding to the color block region.

[0030] Then, the inner center circle of the feature point is separated into red, green and blue color channels to obtain C R , C G , C B , C R , C G , C B , which are binarized and encoded in RGB order to obtain C R C G C B According to the color coding, it is confirmed that the feature point is the first feature point in the color block region, so as to obtain the vertical coordinate of the feature point, where k2 is the decimal color coding value of the feature point in the color block region.

[0031] Finally, a new m×n grid map is created, the nearest feature point of the contour point is identified, and all contour points are filled into the grid according to the coordinates of the feature points. Any contour point E i =(x i ,y i ) has a grid coordinate (m i ,n i ) calculated by projection:

[0032]

[0033] E k is the coordinate of the nearest feature point of the corner point, where Δx=Δy=8, v x and v y are reference vectors, which satisfy orthogonality vx v y =0;

[0034]

[0035] Where E1 is the feature point E k The nearest corner point in the horizontal direction, E2 is the feature point E k The nearest corner point in the vertical direction; obtain the grid coordinates (m) of all contour points. i ,n i After that, they are connected together to form a binary projection mask M of the target object. p .

[0036] Further, step S5 includes:

[0037] All projection mask regions are labeled with connected components. Then, the projection mask regions are divided into two groups according to the connected component index. The connected component indexes of these two groups are guaranteed to be discontinuous.

[0038] The first group of mask regions is named M. R The second set of mask regions is named M. B , respectively in M R and M B The three-step phase-shifting fringe pattern P is generated. i R and P i B Both are calculated using the following formula:

[0039]

[0040] (u p ,v p ) represents the coordinates of the projected pixel, A and B are preset constants, F represents the fringe frequency, and i represents the i-th fringe pattern in the three-step phase-shifting method, i = 1, 2, 3.

[0041] Further, step S6 includes:

[0042] First, create a 24-bit all-black background image. i_color Using (0,0,0) as a carrier, the first group of eight-bit deep three-step phase-shifted fringe patterns P i R Encoded to the red channel, second group of three-step phase-shift fringe pattern P i B Encode the image into the blue channel, leaving the green channel at zero, to form the composite image I. i_color =(P i R ,0,P i B Three sets of such images were generated using the three-step phase-shifting method.i_color = (P i R ,0,P i B ).

[0043] Further, the step S8 comprises:

[0044] First, the spectral channel separation obtains the fringe images of the red and blue single color channels, and then the phase extraction is carried out on the red and blue single color channel images respectively, as shown in the following formula:

[0045]

[0046] wherein I1, I2, I3 are three different phase shift fringe images in the single color channel, is the wrapped phase of the channel calculated by the I1, I2, I3 phase shift fringe images of the channel, and finally, the phases obtained by the two different color channel images are fused, and the phase-depth calibration formula

[0047]

[0048] to obtain the final three-dimensional topography.

[0049] Compared with the prior art, the principle and advantages of the technical solution are as follows:

[0050] 1. The proposed fast regional mapping method can realize the acquisition of the region contour of the measured object according to only one square pattern, and then complete the regional coordinate mapping of the projection device plane. Compared with the current method of obtaining phase by horizontal and vertical fringes, a large number of fringe images are no longer needed for coordinate conversion, which can significantly improve the efficiency and realize the fast measurement of dynamic scenes.

[0051] 2. The proposed heterospectral projection method simultaneously projects different regions with different spectra, which can significantly improve the measurement efficiency compared with the current method of sequentially projecting different regions, and provides feasibility for dynamic measurement of mutual reflection scenes. BRIEF DESCRIPTION OF DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the services needed to be used in the embodiment or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0053] Figure 1 is the working principle diagram of the structured light projection three-dimensional measurement system used in the present application;

[0054] Figure 2 A principle flow chart of a fast region mapping heterospectral structured light measurement method of the present application;

[0055] Figure 3 A seven-color checkerboard image;

[0056] Figure 4 A color-coded image;

[0057] Figure 5 A feature point image in a single region;

[0058] Figure 6 A gray-scale image of the seven-color checkerboard image taken in R, G and B channels respectively. DETAILED DESCRIPTION

[0059] The present application will be further described below in combination with specific embodiments:

[0060] In this example, a 896x1120 pixel image is used as the projection image. Before this, a structured light projection three-dimensional measurement system used in the fast region mapping heterospectral structured light measurement method described in the embodiments of the present application is described in detail:

[0061] As shown in Figure 1 , the structured light projection three-dimensional measurement system comprises a computer 1, a projection light machine 2, a color industrial camera 3 and a measured object 4.

[0062] As shown in Figure 2 , the fast region mapping heterospectral structured light measurement method described in the embodiments is implemented by using the structured light projection three-dimensional measurement system, and the working process is as follows:

[0063] S1, the projection light machine 2 projects a preset multi-color checkerboard pattern (generated by the computer 1) onto the surface of the measured object 4;

[0064] As shown in Figure 3 , the seven-color checkerboard pattern projected by the projection light machine 2 onto the surface of the measured object 4 has a 896x1120 pixel resolution, and is divided into seven color domain blocks in the vertical direction according to the equal division principle, and each block has a height of 160 pixels.

[0065] The color domain arrangement order and color composition of the seven-color checkerboard pattern are as follows:

[0066] The first color domain block: pure red (R, RGB value [255, 0, 0]);

[0067] The second color domain block: pure green (G, RGB value [0, 255, 0]);

[0068] The third color domain block: pure blue (B, RGB value [0, 0, 255]);

[0069] Fourth color gamut block: red-green mixed color (RG / Yellow, RGB value [255, 255, 0]);

[0070] Fifth color gamut block: red-blue mixed color (RB / Magenta, RGB value [255, 0, 255];

[0071] Sixth color gamut block: green-blue mixed color (GB / Cyan, RGB value [0, 255, 255]);

[0072] Seventh color gamut block: red-green-blue mixed color (RGB / White, RGB value [255, 255, 255]).

[0073] S2, acquire the multi-color checkerboard image modulated by the surface of the measured object 4 by using the color industrial camera 3;

[0074] S3, the computer 1 uses RGB three-channel separation technology and combines color coding to identify a plurality of color blocks; then, according to a preset rule, mark feature points, and through the spatial distribution relationship of these feature points, establish an accurate mapping relationship between the projection light machine 2 coordinate system and the color industrial camera 3 coordinate system;

[0075] The specific process of this step includes:

[0076] S3-1, respectively extract the red channel (R), the green channel (G) and the blue channel (B) to generate three binary images; based on the three binary images, each 160-pixel-high color gamut block can be uniquely identified by its corresponding 3-bit binary code, wherein the pure red (R) block code is 100 (R=1, G=0, B=0), the pure green (G) block code is 010 (R=0, G=1, B=0), the pure blue (B) block code is 001 (R=0, G=0, B=1), the red-green mixed (RG / Yellow) block code is 110 (R=1, G=1, B=0), the red-blue mixed (RB / Magenta) block code is 101 (R=1, G=0, B=1), the green-blue mixed (GB / Cyan) block code is 011 (R=0, G=1, B=1), and the red-green-blue mixed (RGB / White) block code is 111 (R=1, G=1, B=1); by scanning the 3-bit code of the image pixel and matching with the preset code sequence, the feature point belonging to the color gamut block can be accurately located. The coding chart is shown in Figure 4 .

[0077] S3-2. Using RGB three-channel separation technology, the 896×1120 pixel seven-color square pattern is divided into seven equal-height regions (each region is 160 pixels high) according to the color gamut. Each region is constructed into a regular square matrix of 128 rows × 160 columns by alternating filling of the main color and 8×8 pixel black squares. Then, each color block region (containing 128×20 square blocks) is subdivided into 7 sub-regions along the row direction (each region contains 16×20 square blocks), and a circular feature marker is set at the center of each sub-region. The color configuration of the marker must meet the following requirements: (1) one of the seven colors of the above color coding, (2) the seven markers correspond to the features of the other seven main color regions respectively, and (3) the color arrangement The order strictly follows the top-down spatial distribution rule of the seven-color regions; the center coordinates of the feature markers are represented by a grid normalization method. The first marker is located at the geometric center of the first sub-region (containing 16×20 grids), with coordinates (8, 10). Subsequent markers of the same color are equally spaced in the vertical direction at intervals of 20 grids (equivalent to 160 pixels in the vertical direction) and in the horizontal direction at intervals of 16 grids (equivalent to 128 pixels in the vertical direction). The center coordinates of the feature points in the first color block are (8+16k, 10), where k is the result obtained by color encoding, which is numerically equal to the binary number corresponding to the color region. The seven feature points in the first color block correspond one-to-one with these seven coordinates.

[0078] S4. Computer 1 dynamically generates a binary projection mask that matches the geometric features of the object's surface based on a feature matching algorithm.

[0079] The specific process of this step includes:

[0080] S4-1. First, the multi-color grid image modulated by the object surface obtained by the color industrial camera 3 is sequentially separated into R, G, and B color channels to obtain I... R I G I B , will I R I G I B Binarize and encode in RGB order (if the region is not 0, I will be encoded). R I G I B Consider it as 1), and we get I. R I G I B The three-digit code corresponds one-to-one with the color block area, thus obtaining the horizontal coordinate of the feature point as 8+16k1, where k1 is the decimal color code value corresponding to the color block area.

[0081] S4-2. Then, perform R, G, and B color channel separation on the center circle inside the feature point to obtain C.R , C G , C B , also C R , C G , C B binary, and encoded in RGB order (C R , C G , C B viewed as 1), C R C G C B , according to the color code to confirm that the feature points are in the color block area within the first feature point, so as to obtain the longitudinal coordinate of the feature point 10+20k2, wherein k2 is the decimal color code value of the corresponding feature point in the color block area;

[0082] S4-3, a new 896x1120 grid map is created, the nearest feature point of the contour point is identified, and all contour points are filled into the grid according to the coordinates of the feature points. Any contour point E i =(x i ,y i ) Its grid coordinates (m i ,n i ) are calculated by projection:

[0083]

[0084] Wherein, E k is the nearest feature point coordinate of the corner point, wherein Δx=Δy=8, v x and v y are reference vectors, which satisfy the orthogonality v x v y =0.

[0085]

[0086] Wherein, E1 is the nearest corner point in the horizontal direction of the feature point E k , E2 is the nearest corner point in the vertical direction of the feature point E k . After obtaining the grid coordinates (m i ,n i ) of all contour points, they are connected to form the binary projection mask M p of the target object.

[0087] S5, the computer 1 divides the projection mask area into two groups according to the spatial position relationship, and generates longitudinal sinusoidal phase shift fringe in the two groups of mask areas;

[0088] This step specifically includes:

[0089] The connected domain marking is performed on all the projection mask regions, and then the projection mask regions are divided into two groups according to the connected domain serial numbers, and the connected domain serial numbers of the two groups of regions are ensured to be discontinuous;

[0090] The first group of mask regions is named M R , and the second group of mask regions is named M B , and three-step phase shift fringe patterns P R , P B and P i R are generated in M i B respectively, both of which adopt the following calculation formula:

[0091]

[0092] (u p ,v p ) represents the projection pixel coordinates, A and B are both preset constants, F represents the fringe frequency, i represents the i-th fringe pattern in the three-step phase shift method, and i=1, 2, 3.

[0093] S6, the computer 1 encodes the two groups of sinusoidal phase shift fringe patterns into the red and blue channels of the RGB image respectively, and the green channel is reserved for color balance to form a heterospectral phase shift fringe pattern;

[0094] This step specifically includes:

[0095] First, a 24-bit all-black background image I i_color (0, 0, 0) is created as a carrier, and the first group of eight-bit deep three-step phase shift fringe patterns P i R are encoded into the red channel, and the second group of three-step phase shift fringe patterns P i B are encoded into the blue channel, and the green channel is kept at zero value to form a composite image I i_color =(P i R , 0, P i B ); three groups of such images I i_color =(P i R , 0, P i B ) are generated according to the three-step phase shift method.

[0096] S7, the projection light machine 2 projects the encoded fringe image to the measured object 4 at a refresh rate of ≥60Hz, and synchronously triggers the color industrial camera 3 to capture images;

[0097] S8, the collected heterospectral phase shift fringe patterns are used for phase extraction and three-dimensional topography reconstruction.

[0098] This step specifically includes:

[0099] First, the spectral channel separation obtained red and blue monochrome channel fringe image, then the red and blue monochrome channel image for independent phase extraction, as shown below:

[0100]

[0101] Wherein, I1, I2, I3 is the three different phase shift in monochrome channel, fringe, The wrapped phase of the channel calculated by I1, I2, I3 phase shift fringe of the channel, finally, the phase fusion obtained by calculating two different color channel image, through the phase-depth calibration formula

[0102]

[0103] Obtain the final three-dimensional topography.

[0104] The above described embodiments of the present application are only the preferred embodiments of the present application, not to limit the scope of the present application, so that any changes made in accordance with the shape, principle of the present application, should be covered within the scope of the present application.

Claims

1. A fast region mapping heterospectral structured light measurement method, characterized in that, The method comprises the following steps: S1, projecting a preset multi-color checkerboard pattern onto the surface of a target object by a projection device; S2, acquiring a multi-color checkerboard image modulated by the surface of the object by an image acquisition device; S3, identifying the multi-color blocks by using RGB three-channel separation technology combined with color coding, then marking feature points according to a preset rule, and establishing an accurate mapping relationship between the coordinate system of the projection device and the coordinate system of the image acquisition device according to the spatial distribution relationship of the feature points; S4, dynamically generating a binary projection mask matched with the geometric features of the surface of the object according to a feature matching algorithm; S5, dividing the projection mask area into two groups according to the spatial position relationship, and generating longitudinal sinusoidal phase shift fringe patterns in the two groups of mask areas; S6, encoding the two groups of sinusoidal phase shift fringe patterns into the red and blue channels of the RGB image respectively, and reserving the green channel for color balance to form a heterospectral phase shift fringe pattern; S7, projecting the encoded fringe image onto the target object at a set refresh rate, and synchronously triggering the image acquisition device to capture images; S8, using the acquired heterospectral phase shift fringe pattern to extract the phase and reconstruct the three-dimensional topography.

2. The fast region-mapped heterospectral structured light measurement method according to claim 1, characterized in that, The multicolor square pattern projected by the projection device to the target object surface is a seven-color square pattern, which has an m×n pixel resolution, is divided into seven color domain blocks according to the equal division principle in the vertical direction, and each block has a height of Pixels; The color gamut arrangement order and color composition of the seven-color checkerboard pattern are as follows: First color gamut block: pure red; Second color gamut block: pure green; Third color gamut block: pure blue; Fourth color gamut block: red-green mixed color; Fifth color gamut block: red-blue mixed color; Sixth color gamut block: green-blue mixed color; Seventh color gamut block: red-green-blue mixed color.

3. The fast region-mapped heterospectral structured light measurement method of claim 2, wherein, In step S3, seven color blocks are identified by using RGB three-channel separation technology combined with color coding, which includes the following steps: The image is RGB three-channel separated, and red, green and blue channels are extracted to generate three binary images; based on the three binary images, each The color gamut block with a pixel height is uniquely identified by its corresponding 3-bit binary code, wherein the pure red block is coded as 100, the pure green block is coded as 010, the pure blue block is coded as 001, the red-green mixed block is coded as 110, the red-blue mixed block is coded as 101, the green-blue mixed block is coded as 011, and the red-green-blue mixed block is coded as 111; by scanning the 3-bit code of the image pixel and matching with the preset code sequence, the color gamut block to which the feature point belongs is accurately positioned.

4. The fast region-mapped heterospectral structured light measurement method of claim 2, wherein, In step S3, the feature points are marked according to a preset rule, and the accurate mapping relationship between the coordinate system of the projection device and the coordinate system of the image acquisition device is established according to the spatial distribution relationship of the feature points, which includes the following steps: The RGB three-channel separation technology is adopted to divide the m×n pixel seven-color checkerboard pattern into seven equal high areas according to the color gamut, and each area is constructed by alternately filling the main color and 8×8 pixel black checkerboard Row Column regular checkerboard matrix; then each color block area is subdivided into 7 sub-areas along the row direction, and a circular feature marker point is set at the center position of each sub-area, and the color configuration of the marker point needs to meet: 1) one of the seven colors coded by the above method, 2) the seven marker points correspond to the features of the remaining seven main color areas respectively, 3) the color arrangement order strictly follows the spatial distribution rule of the seven color areas from top to bottom; the center coordinates of the feature marker points are normalized by the checkerboard block, the first marker point is located at the geometric center of the first sub-area, and the coordinates are The subsequent same-color marker points are equally distributed with an interval of checkerboard blocks in the horizontal direction, and an interval of checkerboard blocks in the vertical direction, and the center coordinates of the feature points in the first color block area are Wherein k is the result obtained by color coding, and the value is equal to the binary number corresponding to the color area, and the seven feature points in the first color block area correspond to the seven coordinates one by one.

5. The fast region-mapped heterospectral structured light measurement method of claim 2, wherein, The process of step S4 includes: First, the multi-color square image modulated by the object surface obtained by the image acquisition device is sequentially separated into red, green and blue color channels to obtain I R 、 G 、 B , I R 、 G 、 B , I R 、 G 、 B , the three-digit code corresponds to the color block area one by one, so as to obtain the horizontal coordinate of the feature point where k1 is the decimal color encoding value corresponding to the color block area; Then the feature point inner center circle is separated into C R , C G , C B , C R , C G , C B , and binaryzation and encoding according to RGB order, C R C G C B , and according to color coding, the feature point is confirmed as the kth feature point in the color block region, so as to obtain the vertical coordinate of the feature point wherein k2 is the decimal color coding value of the corresponding feature point in the color block region; Finally, a new m x n grid map is created, the nearest feature point of the contour point is identified, and all contour points are filled into the grid according to the coordinates of the feature points. Any contour point E i = (x i ,y i ) has its grid coordinates (m i ,n i ) calculated by projection: E k is the coordinate of the nearest feature point to the corner point, where Δx = Δy = 8, v x and v y is the reference vector, satisfying the orthogonality v x v y = 0; E1 is the nearest corner point in horizontal direction k E2 is the nearest corner point in vertical direction k After getting the grid coordinates (m i ,n i ) of all the contour points, they are connected to form the binary projection mask M p of the target object.

6. The fast region-mapped heterospectral structured light measurement method of claim 2, wherein, Step S5 includes: All projection mask areas are labeled as connected domains, and then the projection mask areas are divided into two groups according to the connected domain serial numbers, and the connected domain serial numbers of the two groups of areas are ensured to be discontinuous; where the first set of mask regions is named M R , the second set of mask regions is named M B , three-step phase-shifted fringe patterns P R , P B and P i R and P i B are generated in M R and M B respectively, both of which are calculated by the following formula: (u p ,v p ) represents the projection pixel coordinates, A and B are both preset constants, F represents the fringe frequency, i represents the i-th fringe image in the three-step phase shift method, i = 1, 2, 3.

7. The fast region-mapped heterospectral structured light measurement method according to claim 6, characterized in that, Step S6 includes: First create a 24-bit all-black background image I i_color (0,0,0) as a carrier, a first set of eight-bit deep three-step phase-shift fringe patterns P i R encoded into the red channel, a second set of three-step phase-shift fringe patterns P i B encoded into the blue channel, the green channel remains zero, forming a composite image I i_color = (P i R ,0, P i B ); generate three sets of such images I i_color = (P i R ,0, P i B ) by three-step phase-shift method.

8. The fast region-mapped heterospectral structured light measurement method of claim 1, wherein, Step S8 includes: First, the red and blue monochrome channel fringe images are obtained by spectral channel separation, and then independent phase extraction is performed on the red and blue monochrome channel images respectively, as shown in the following formula: where I1, I2, I3 are three different phase-shifted fringe patterns in a monochrome channel, the wrapped phase of the channel calculated from the I1, I2, I3 phase-shifted fringe patterns of the channel, and finally, the phases calculated from the two different color channel images are fused through a phase-depth calibration formula The final three-dimensional topography is obtained.

Images