A pressure sensing device and detection method based on image recognition

Abstract

The present invention relates to the field of pressure detection technology, and in particular to a pressure sensing device and detection method based on image recognition. The device includes a flexible grid marker plate, a rigid transparent substrate, a camera mechanism, and a processor. The flexible grid marker plate includes a lower flexible transparent layer, a marking texture, and an upper flexible transparent layer; the flexible grid marker plate is placed on the rigid transparent substrate; the camera mechanism includes a central camera and several side cameras; and the processor is connected to the camera mechanism. The method of using a flexible grid marker plate and a camera mechanism to realize optical vision for pressure measurement greatly reduces the hardware cost compared to the traditional pressure sensor method, which is conducive to market promotion and application; the flexible grid marker plate can completely fit the surface of the force-applying object, so the data obtained is closer to the actual situation of the force applied by the measuring body and has high accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of pressure detection, and in particular to a pressure sensing device and a detection method based on image recognition. Background Art

[0002] With the rise of artificial intelligence and the development and application of conductive nanomaterials such as graphene, flexible pressure sensors have become a hot topic of research at home and abroad. Pressure sensors generally used to measure static forces include piezoresistive, capacitive, and photosensitive. Piezoresistive sensors use array-type lead arrangements, with many internal cross points, relatively difficult decoupling, and high costs. Capacitive sensors are small in size and suitable for precise measurement of small areas, but are easily interfered by strong parasitic capacitance. The shapes of these two sensors cannot fit irregular force-bearing surfaces and cannot accurately collect regional pressure distribution data. Photosensitive pressure sensors use optical signals to sense and detect external forces, which can solve problems such as electromagnetic interference, complex wiring, and crosstalk. Most of them use special optical fibers with high costs. A patent proposes an image-based tactile sensing method. By integrating the light source into the pressure sensor, the bending and deformation of the microcolumn can be clearly captured, and the intelligent algorithm is used based on the deformation image to complete the decoupling of the three-dimensional force, realizing the miniaturization, high integration, and precision of the sensor. However, it cannot meet the measurement and analysis of large-area pressure distribution, and can only measure planes.

[0003] Therefore, there is an urgent need for a low-cost, customizable, and flexible pressure sensing device and detection method that is not easily restricted by the application environment and is suitable for different usage scenarios. Summary of the invention

[0004] In order to overcome the deficiencies in the prior art, the present invention provides a pressure detection method based on image recognition, which has high stability, is not easily affected by the application environment, and obtains accurate and high-precision pressure data.

[0005] To achieve the above-mentioned purpose, the technical solution adopted by the present invention is as follows: a pressure sensing device based on image recognition, characterized in that: it includes a flexible grid marking plate, a rigid transparent substrate, a camera mechanism, and a processor, the flexible grid marking plate includes a lower flexible transparent layer and a marking texture, the marking texture is arranged on the upper side of the lower flexible transparent layer, and the marking texture has marking points that form a grid; the flexible grid marking plate is placed on the rigid transparent substrate; the camera mechanism is located below the rigid transparent substrate, and includes a central camera and several side cameras, the central camera is located below the center position of the flexible grid marking plate and is vertically facing the flexible grid marking plate; the side cameras are located below the flexible grid marking plate and are inclined toward the flexible grid marking plate, the side cameras are arranged in pairs, and the corresponding two groups of side cameras are symmetrically distributed relative to the center of the central camera; the processor is connected to the camera mechanism.

[0006] Furthermore, the grid composed of marking points is a regular grid, and the marking points include reference square points and array circular points. The reference square points have a square shape, and there are four reference square points, which are respectively arranged at the four corners of the grid composed of marking points. The array circular points have a circular shape, and each array circular point is arranged at equal intervals.

[0007] Furthermore, the flexible grid marking plate also includes an upper flexible transparent layer, and the marking texture is located between the upper flexible transparent layer and the lower flexible transparent layer.

[0008] Furthermore, the ratio of the thickness of the upper flexible transparent layer to the thickness of the lower flexible transparent layer is 1:1-1:100000.

[0009] Furthermore, the flexible grid marking plate, the rigid transparent substrate, and the camera mechanism form a group of measuring sections, and multiple groups of measuring sections are set, which also include several auxiliary sections. A measuring section is located between two auxiliary sections. The auxiliary section includes a rigid substrate and a flexible panel. The flexible panel is arranged on the rigid substrate. The material and thickness of the flexible panel are the same as those of the flexible grid marking plate. The flexible panel borders on the flexible grid marking plate, and the top surface of the flexible panel is flush with the top surface of the flexible grid marking plate.

[0010] A pressure detection method based on image recognition, characterized in that it comprises the following steps:

[0011] Step 1: In the initial state, the central camera and the side cameras of the camera mechanism take pictures to form multiple reference images, and transmit the image information to the processor;

[0012] Step 2: In the processor, each reference image is subjected to image noise reduction processing, the features of the marker points are enhanced, the features of each marker point are extracted from each reference image based on the SIFT algorithm, the feature marker points are matched according to the feature vector, the matched image is reconstructed using the reconstruction algorithm, the coordinates of each marker point are generated according to the interpolation calculation, and the coordinates of each marker point of the reconstructed reference model are output;

[0013] Step 3: In the measurement state, the measuring body applies pressure to the top surface of the flexible grid marking plate. The pressure applied by the measuring body causes the flexible grid marking plate to deform, that is, the positions of the marking points of the marking texture change; the central camera and the side cameras of the camera mechanism take pictures to form multiple detection images, and transmit the image information to the processor;

[0014] Step 4: In the processor, each detection image is subjected to image denoising, the features of the marker points are enhanced, the features of each marker point are extracted from each detection image based on the SIFT algorithm, the feature marker points are matched according to the feature vector, the matched image is reconstructed using the reconstruction algorithm, the coordinates of each marker point are generated according to the interpolation calculation, and the coordinates of each marker point of the reconstructed detection model are output;

[0015] Step 5: Calculate the displacement of the marker point based on the coordinates of the marker point in the reconstructed reference model and the coordinates of the reconstructed detection model, and integrate the displacements of each marker point to form a marker point displacement distribution map;

[0016] Step six, obtaining a grid pressure value distribution map according to the displacement pressure value correspondence table formed by pre-measurement.

[0017] Furthermore, between step five and step six, the displacement of the four marking points constituting the cell is calculated as an average value to obtain the displacement of the unit grid, the displacement of each cell is integrated to output a grid displacement distribution map, and the grid pressure value distribution map is obtained by corresponding to the displacement pressure value corresponding table formed by pre-measurement; or the marking point is drawn at the center of the grid, the cell displacement is calculated with the displacement of the marking point, and the displacement of each cell is integrated and output to obtain the grid pressure value distribution map.

[0018] Furthermore, the method for making the flexible grid marking plate is as follows:

[0019] A customized mold container is prepared with a flexible base material solution and placed in a vacuum box to remove bubbles; the flexible base material solution is poured into the mold container to a predetermined liquid level, and allowed to stand until the material solidifies to form a lower flexible transparent layer; a layer of characteristic marking point material is printed on the upper surface of the lower flexible transparent layer to form a marking texture; the flexible base material solution is poured into the mold container again to a predetermined liquid level height, the marking texture is encapsulated and an upper flexible transparent layer is formed; the material is allowed to stand until it is completely solidified and then taken out to form a flexible grid marking plate.

[0020] Furthermore, the displacement pressure value correspondence table is formed as follows:

[0021] There are multiple standard parts, each of which has the same shape and is densely and evenly distributed inside the standard parts. The masses of the standard parts are different and known. Each standard part is placed individually on the surface of the flexible grid marking plate for multiple times, and the grid displacement distribution diagram corresponding to different standard parts is obtained through the camera mechanism and processor processing. A unit grid displacement distribution diagram and pressure association model is established to form a displacement pressure value correspondence table.

[0022] Furthermore, the method for making the standard part is as follows:

[0023] Metal powders of different densities are selected and mixed in different proportions to form multiple portions of mixed metal powders of the same volume but different masses. The mixed metal powders are mixed evenly and multiple standard parts of the same shape but different masses are manufactured through compression molding.

[0024] It can be seen from the above description of the present invention that, compared with the prior art, the pressure detection method based on image recognition provided by the present invention has the following advantages:

[0025] The pressure applied by the measuring body is reflected by the deformation of the flexible grid marker plate. The coordinate changes of the marking points of the marking texture grid can finely display the pressure conditions of each position of the flexible grid marker plate in a concrete digital form. Through the shooting of the camera mechanism and the processing of the processor, accurate and detailed pressure information can be obtained.

[0026] The flexible grid marker board has no electronic components and is not easily affected by external environment such as temperature or electromagnetic interference. It has high adaptability and reliability. During the measurement process, only the flexible grid marker board is subject to loss. However, the production cost of the flexible grid marker board is low, and the accuracy can be restored by replacing it, thus achieving low-cost maintenance.

[0027] The method of optical vision using a flexible grid marker plate and a camera mechanism to measure pressure greatly reduces the hardware cost compared to the traditional pressure sensor method, which is conducive to market promotion and application; the flexible grid marker plate can completely fit the surface of the force-applying object, so the data obtained is closer to the actual situation of the force applied by the measuring body and has high accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 This is a schematic structural diagram of a pressure sensing device based on image recognition according to the present invention.

[0029] Figure 2 It is a schematic diagram of the application scenario structure in an embodiment of the present invention.

[0030] Figure 3 The present invention is a schematic flowchart of a pressure detection method based on image recognition.

[0031] Figure 4 It is a schematic diagram of the displacement change of the marking points of the flexible grid marking plate of the present invention.

[0032] Figure 5 This is a schematic diagram of the marking texture of the present invention.

[0033] Figure 6 Schematic diagram of unit grid displacement of the present invention.

[0034] Figure 7 This is the distribution diagram of the grid displacement of the present invention.

[0035] Figure 8 This is a schematic diagram of the manufacturing process of the flexible grid marking plate of the present invention.

[0036] Fig. 9 The displacement pressure value corresponding curve and color code table of the present invention.

[0037] Fig.10 The figure is a schematic diagram of the process of obtaining a displacement pressure value correspondence table through standard parts in the present invention.

[0038] Fig.11 A schematic diagram of the grid pressure value distribution generation according to the present invention.

[0039] The symbols in the figure correspond to the following: 1. Flexible grid marking plate, 11. Lower flexible transparent layer, 12. Marking texture, 121. Marking point, 121a. Reference square point, 121b. Array dot, 13. Upper flexible transparent layer, 2. Rigid transparent substrate, 3. Camera mechanism, 31. Center camera, 32. Side camera, 4. Processor, 5. Auxiliary segment, 51. Rigid substrate, 52. Flexible panel. DETAILED DESCRIPTION

[0040] The present invention is further described below through specific implementation modes.

[0041] Reference Figures 1 to 11 As shown, a pressure sensing device based on image recognition includes a flexible grid marking plate 1, a rigid transparent substrate 2, a camera mechanism 3, a processor 4, and an auxiliary segment 5.

[0042] The flexible grid marking plate 1 comprises a lower flexible transparent layer 11, a marking texture 12, and an upper flexible transparent layer 13, wherein the thickness ratio of the upper flexible transparent layer 13 to the thickness of the lower flexible transparent layer 11 is 1:1-1:100000, and the marking texture 12 is disposed between the upper flexible transparent layer 13 and the lower flexible transparent layer 11. In another embodiment, the upper flexible transparent plate is not provided, and the marking texture 12 is directly disposed on the upper side of the lower flexible transparent layer 11.

[0043] The marking texture 12 has marking points 121 forming a grid, and the grid formed by the marking points 121 is a regular grid, such as a square grid, a diamond grid, a triangular grid, etc. The marking points 121 include reference square points 121a and array dots 121b. The reference square points 121a have a square shape. There are four reference square points 121a, which are respectively arranged at the four corners of the grid formed by the marking points 121. The array dots 121b have a circular shape, and each array dot 121b is arranged at equal intervals; the flexible grid marking plate 1 is placed on the rigid transparent substrate 2; the camera mechanism 3 is located below the rigid transparent substrate 2, and includes a central camera 31 and a plurality of side cameras 32. The central camera 31 is located below the center position of the flexible grid marking plate 1 and is vertically facing the flexible grid marking plate 1; the side camera position 32 is located below the flexible grid marking plate 1 and is tilted toward the flexible grid marking plate 1. The side cameras 32 are arranged in pairs, and the corresponding two groups of side cameras 32 are symmetrically distributed relative to the center of the central camera 31; the processor 4 is connected to the camera mechanism.

[0044] The flexible grid marking plate 1, the rigid transparent substrate 2, and the camera mechanism 3 form a group of measuring sections. Multiple groups of measuring sections are set, and also include a number of auxiliary sections 5. A measuring section is located between two auxiliary sections 5. The auxiliary section 5 includes a rigid substrate 51 and a flexible panel 52. The flexible panel 52 is arranged on the rigid substrate 51. The material and thickness of the flexible panel 52 are the same as those of the flexible grid marking plate 1. The flexible panel 52 is adjacent to the flexible grid marking plate 1, and the top surface of the flexible panel 52 is flush with the top surface of the flexible grid marking plate 1.

[0045] A pressure detection method based on image recognition comprises the following steps:

[0046] Step 1, in the initial state, the internal parameters of the central camera 31 and each side camera 32 of the camera mechanism 3 are obtained, the central camera 31 and each side camera 32 are fixedly set, the central camera 31 is used as the reference calibration coordinate, and the relative position between the side camera 32 and the central camera 31 is measured to obtain the calibration coordinates of each side camera 32; in the initial state, the flexible grid marking plate 1 is in an unpressurized state, and the central camera 31 and each side camera 32 of the camera mechanism 3 shoot to form a plurality of reference images, and transmit the image information to the processor;

[0047] Step 2: In the processor, each reference image is pre-processed and image noise reduction processing based on algorithms such as gamma transform is performed to enhance the features of the marker point 121. The four corners of the marker point 121 are the reference square points 121a. The reference square point 121a has a square shape and contains direction information, which plays a role in further assisting positioning. The features of each marker point 121 are extracted from each reference image based on the SIFT algorithm, and the feature marker points are matched according to the feature vector. The matched image is reconstructed using a reconstruction algorithm, and the coordinates of each marker point 121 are generated according to the interpolation calculation, and the coordinates of each marker point of the reconstructed reference model are output; for example, a marker point 121 in the reconstructed reference model is p, and the output coordinates are p(xyz);

[0048] Step 3: In the measurement state, the measuring body applies pressure to the top surface of the flexible grid marking plate 1. The pressure applied by the measuring body causes the flexible grid marking plate 1 to deform, that is, the positions of the marking points 121 of the marking texture 12 change; the central camera 31 and the side cameras 32 of the camera mechanism 3 take pictures to form multiple detection images, and transmit the image information to the processor 4;

[0049] In step four, in the processor 4, each detection image is preprocessed and subjected to image noise reduction processing based on algorithms such as gamma transform to enhance the features of the marker point 121. The four corners of the marker point 121 are the reference square points 121a. The reference square point 121a has a square shape and contains directional information, which plays a role in further assisting positioning. The features of each marker point 121 are extracted from each detection image based on the SIFT algorithm, and the feature marker points are matched according to the feature vector. The matched image is reconstructed using a reconstruction algorithm, and the coordinates of each marker point 121 are generated according to the interpolation calculation, and the coordinates of each marker point of the reconstructed detection model are output; for example, a marker point 121 in the reconstructed detection model is p′, and the output coordinates are p′(x′.y′.z′).

[0050] Step 5: Calculate the coordinates of the reconstructed reference model and the reconstructed detection model based on the coordinates of the marker point 121 The displacement of the marking point 121 is obtained, and the displacement of each marking point 121 is integrated to form a distribution diagram of the displacement of the marking point;

[0051] Step six, take the displacement of the four marking points 121 that make up a cell and calculate the average value to obtain the displacement of the unit grid, and integrate the displacement of each cell to output a grid displacement distribution map; or draw the marking point 121 at the center of the grid, calculate the cell displacement with the displacement of the marking point 121, and integrate the displacement of each cell to obtain a grid pressure value distribution map.

[0052] Step seven, obtain a grid pressure value distribution diagram according to the displacement pressure value correspondence table formed by pre-measurement, pre-set the pressure value corresponding color code table, fill in the corresponding color according to the grid pressure value, and realize a more intuitive visual pressure value distribution diagram.

[0053] The manufacturing method of the flexible grid marking plate 1 is as follows: a mold container is customized, a flexible base material solution is prepared and placed in a vacuum box to remove bubbles; the flexible base material solution is poured into the mold container to a predetermined liquid level, and the material is allowed to stand until the material solidifies to form a lower flexible transparent layer 11; a layer of characteristic marking point material is printed on the upper surface of the lower flexible transparent layer 11 to form a marking texture 12; the flexible base material solution is poured into the mold container again to a predetermined liquid level height, the marking texture is encapsulated and an upper flexible transparent layer 13 is formed; the material is allowed to stand until it is completely solidified and taken out to form a flexible grid marking plate 1. The flexible grid marking plate 1 does not contain electronic components and is not easily affected by external environments such as temperature or electromagnetic interference.

[0054] The displacement-pressure value correspondence table is formed as follows: there are multiple standard parts, each of which has the same shape, is densely and evenly distributed inside the standard parts, and has different and known masses. Each standard part is placed individually on the surface of the flexible grid marking plate for multiple times, and the grid displacement distribution diagram corresponding to different standard parts is obtained through the camera mechanism and processor processing, and a unit grid displacement distribution diagram and pressure association model is established to form a displacement-pressure value correspondence table.

[0055] The method for making standard parts is as follows: select metal powders of different densities, form multiple mixed metal powders of the same volume but different masses through different proportions, stir the mixed metal powders evenly, and make multiple standard parts with the same shape but different masses through molding.

[0056] SIFT, or Scale-invariant feature transform (SIFT), is a description used in the field of image processing. This description is scale-invariant and can detect key points in an image. It is a local feature descriptor.

[0057] The following further illustrates the detection method through a practical application scenario;

[0058] Monitoring of the force pressure and force posture of the soles of the feet of runners on the track during running:

[0059] A group of measuring segments includes a flexible grid marking plate 1, a rigid transparent substrate 2, and a camera mechanism 3, and a group of auxiliary segments 5 includes a rigid substrate 51 and a flexible panel 52; a plurality of measuring segments and a plurality of auxiliary segments 5 are arranged along the length direction, with a measuring segment located between two auxiliary segments 5, to form a measuring runway.

[0060] The athlete runs on the track, and the camera mechanism 3 of the measuring section works continuously. When the athlete passes through the measuring section, the deformation caused by the force exerted by the sole of the foot on the flexible grid marking plate 1 will be recorded. After calculation by the processor 4, the pressure value exerted by the athlete's sole at various positions on the track during the running process and the information on various postures of the sole of the foot can be obtained. Through the above information, the athlete's movements can be accurately grasped to provide a data basis for further scientific training.

[0061] The flexible grid marking plate 1 of the measuring section and the flexible panel 52 of the auxiliary section have the same material and top surface horizontal position, and the two are in a bordering state, so the feedback feeling obtained by the athlete running on the entire track is consistent, reducing measurement errors.

[0062] The above is only a specific implementation of the present invention, but the design concept of the present invention is not limited to this. Any non-substantial changes to the present invention using this concept shall be deemed as an infringement of the protection scope of the present invention.

Claims

1. A pressure sensing device based on image recognition, characterized in that: It includes a flexible grid marking plate, a rigid transparent substrate, a camera mechanism, and a processor. The flexible grid marking plate includes a lower flexible transparent layer and a marking texture. The marking texture is arranged on the upper side of the lower flexible transparent layer. The marking texture has marking points forming a grid. The flexible grid marking plate is placed on the rigid transparent substrate. The camera mechanism is located below the rigid transparent substrate, and includes a central camera and several side cameras. The central camera is located below the center of the flexible grid marking plate and is vertically facing the flexible grid marking plate; the side cameras are located below the flexible grid marking plate and are inclined toward the flexible grid marking plate. The side cameras are arranged in pairs, and the corresponding two groups of side cameras are symmetrically distributed relative to the center of the central camera; the processor is connected to the camera mechanism, and the flexible grid marking plate also includes an upper flexible transparent layer, and the marking texture is located between the upper flexible transparent layer and the lower flexible transparent layer.

2. The pressure sensing device based on image recognition according to claim 1, characterized in that: The grid composed of marking points is a regular grid, and the marking points include reference square points and array circular points. The reference square points have a square shape. There are four reference square points, which are respectively arranged at the four corners of the grid composed of marking points. The array circular points have a circular shape, and each array circular point is arranged at equal intervals.

3. The pressure sensing device based on image recognition according to claim 1, characterized in that: The ratio of the thickness of the upper flexible transparent layer to the thickness of the lower flexible transparent layer is 1:1-1:100000.

4. The pressure sensing device based on image recognition according to claim 1, characterized in that: The flexible grid marking plate, the rigid transparent substrate, and the camera mechanism form a group of measuring sections. Multiple groups of measuring sections are set, and several auxiliary sections are also included. A measuring section is located between two auxiliary sections. The auxiliary section includes a rigid substrate and a flexible panel. The flexible panel is arranged on the rigid substrate. The material and thickness of the flexible panel are the same as those of the flexible grid marking plate. The flexible panel is adjacent to the flexible grid marking plate, and the top surface of the flexible panel is flush with the top surface of the flexible grid marking plate.

5. A pressure detection method based on image recognition, characterized in that: A pressure sensing device based on image recognition according to any one of claims 1 to 4, comprising the following steps: Step 1: In the initial state, the central camera and the side cameras of the camera mechanism take pictures to form multiple reference images, and transmit the image information to the processor; Step 2: In the processor, each reference image is subjected to image noise reduction processing, the features of the marker points are enhanced, the features of each marker point are extracted from each reference image based on the SIFT algorithm, the feature marker points are matched according to the feature vector, the matched image is reconstructed using the reconstruction algorithm, the coordinates of each marker point are generated according to the interpolation calculation, and the coordinates of each marker point of the reconstructed reference model are output; Step 3: In the measurement state, the measuring body applies pressure to the top surface of the flexible grid marking plate. The pressure applied by the measuring body causes the flexible grid marking plate to deform, that is, the positions of the marking points of the marking texture change; the central camera and the side cameras of the camera mechanism take pictures to form multiple detection images, and transmit the image information to the processor; Step 4: In the processor, each detection image is subjected to image denoising, the features of the marker points are enhanced, the features of each marker point are extracted from each detection image based on the SIFT algorithm, the feature marker points are matched according to the feature vector, the matched image is reconstructed using the reconstruction algorithm, the coordinates of each marker point are generated according to the interpolation calculation, and the coordinates of each marker point of the reconstructed detection model are output; Step 5: Calculate the displacement of the marker point based on the coordinates of the marker point in the reconstructed reference model and the coordinates of the reconstructed detection model, and integrate the displacements of each marker point to form a marker point displacement distribution map; Step six, obtaining a grid pressure value distribution map according to the displacement pressure value correspondence table formed by pre-measurement.

6. The pressure detection method based on image recognition according to claim 5, characterized in that: Between step 5 and step 6, the displacements of the four marking points constituting the cell are calculated to average the displacements of the unit grid, and the displacements of each cell are integrated to output a grid displacement distribution map, which is matched according to the displacement pressure value corresponding table formed by pre-measurement to obtain a grid pressure value distribution map; Or the marking point is drawn at the center of the grid, the displacement of the cell is calculated by the displacement of the marking point, and the displacement of each cell is integrated and output to obtain the grid pressure value distribution map.

7. The pressure detection method based on image recognition according to claim 5, characterized in that: The method of making the flexible grid marker board is as follows: A customized mold container is prepared with a flexible base material solution and placed in a vacuum box to remove bubbles; the flexible base material solution is poured into the mold container to a predetermined liquid level, and allowed to stand until the material solidifies to form a lower flexible transparent layer; a layer of characteristic marking point material is printed on the upper surface of the lower flexible transparent layer to form a marking texture; the flexible base material solution is poured into the mold container again to a predetermined liquid level height, the marking texture is encapsulated and an upper flexible transparent layer is formed; the material is allowed to stand until it is completely solidified and then taken out to form a flexible grid marking plate.

8. The pressure detection method based on image recognition according to claim 5, characterized in that: The displacement pressure value correspondence table is formed as follows: There are multiple standard parts, each of which has the same shape and is densely and evenly distributed inside the standard parts. The masses of the standard parts are different and known. Each standard part is placed individually on the surface of the flexible grid marking plate for multiple times. The grid displacement distribution diagram corresponding to different standard parts is obtained through the camera mechanism and processor processing, and a unit grid displacement distribution diagram and pressure correlation model is established to form a displacement pressure value correspondence table.

9. The pressure detection method based on image recognition according to claim 8, characterized in that: The method for making the standard parts is as follows: Metal powders of different densities are selected and mixed in different proportions to form multiple portions of mixed metal powders of the same volume but different masses. The mixed metal powders are mixed evenly and multiple standard parts of the same shape but different masses are manufactured through compression molding.

Images