Underwater static pressure self-balancing visual tactile perception device and method

Abstract

The invention provides an underwater static pressure self-balancing visual touch perception device and method. The device comprises an elastic body, a hexagonal tray, a light source, a camera, a protective cover, a base, a water inlet pipe, a water outlet pipe and a computer. The elastomer is used for contacting an object, and a plurality of micron-sized through holes for balancing internal and external water pressure are formed in the surface; the elastomer is arranged on the inner side of the hexagonal tray, and the RGB three-color light source is arranged on the outer side to realize uniform distribution of light; the camera collects a light intensity image reflected by the elastomer; the protective cover is fixedly connected to the base and used for isolating external light and fixing the hexagonal tray. A camera is mounted on the base; the computer is used to model the acquired images or videos as different haptic features.

Description

Technical Field

[0001] The present invention relates to the technical field of underwater robot tactile perception, and more particularly, to an underwater hydrostatic pressure self-balanced visual tactile perception device and method. Background Art

[0002] Touch is one of the important senses for understanding the real world, especially important in underwater situations where vision is limited or unavailable. Compared with vision, tactile perception provides more direct, accurate, and rich proprioception. Underwater, the environment is complex and variable (such as the irregularity of the seabed terrain), which brings huge challenges to industrial operations such as underwater operations and ocean exploration. Underwater tactile perception mainly relies on various tactile sensors, which are assembled on underwater robots to achieve refined operation and perception tasks. Existing electronic-based tactile sensors (piezoelectric, piezoresistive, capacitive, and triboelectric) mainly use external touch pressure to generate electrical signals to sense multi-dimensional physical information such as pressure, temperature, and shape. However, these sensors have many problems underwater, such as poor compressive resistance, low resolution, susceptibility to environmental influences (humidity, temperature, etc.), and difficulty in simultaneously obtaining multi-dimensional information. This will lead to a decline in the performance of the sensors underwater and even inability to work properly. Visual tactile sensors have received extensive attention due to their sensitivity to subtle contact forces and the ability to provide real-time, high-resolution contact geometry depth maps at low cost, and can sense the properties of objects themselves: surface texture, physical shape, hardness estimation, and the state of contact: normal force, shear force, relative sliding, object pose, etc., endowing the robot with extraordinary dexterity and precise perception. However, the biggest problem in the current application of visual tactile sensors underwater is that the elastomer film inside the existing visual tactile sensors has a hollow structure, and when working underwater, the hydrostatic pressure will cause the elastomer film to stretch, rupture, or harden, thus affecting the normal operation of the sensor.

[0003] In summary, it is necessary to design a visual tactile perception device that can be reliably used underwater to solve the above technical problems. Summary of the Invention

[0004] According to the above-mentioned technical problem that the existing visual tactile sensor is not resistant to hydrostatic pressure underwater, an underwater hydrostatic pressure self-balanced visual tactile perception device and method are provided. The present invention realizes the underwater multi-dimensional information tactile perception of the visual tactile sensor for the first time, and has outstanding advantages such as low cost and being unaffected by hydrostatic pressure.

[0005] The technical means adopted by the present invention are as follows:

[0006] An underwater hydrostatic pressure self-balanced visual tactile perception device, comprising: an elastomer, a hexagonal tray, a red light source, a green light source, a blue light source, a camera, a protective cover, a base, a water inlet pipe, a water outlet pipe, and a computer device, wherein:

[0007] The elastomer is embedded in the upper half of the hexagonal tray and is used to provide real-time feedback on the physical information of the contacted object, including the texture and shape of the surface of the contacted object;

[0008] The hexagonal tray has the elastomer placed inside and a light source installed outside to achieve a uniform light distribution effect;

[0009] The red light source, green light source, and blue light source are used to provide parallel RGB three-color illumination;

[0010] The camera is used to collect the light intensity images of the elastomer deforming under pressure;

[0011] The protective cover is used to isolate external light and fix the overall structure of the hexagonal tray and the underwater hydrostatic pressure self-balanced visual-tactile perception device;

[0012] The base is used to support and protect the camera, and the protective cover is fixedly connected to the base with screws;

[0013] The water inlet pipe and the water outlet pipe are respectively connected to both ends below the side wall of the protective cover and are used to balance the water pressure on both sides inside and outside the device;

[0014] The computer device is connected to the camera through the communication interface on the camera, is used to control the camera to take pictures and receive the collected images or videos, and uses computer vision algorithms to model the collected images or videos into different tactile features.

[0015] Further, the elastomer includes transparent silica gel and a reflective film, wherein:

[0016] The transparent silica gel is hexagonal prism-shaped and is provided with a plurality of micro through-holes;

[0017] The reflective film is a solution obtained by uniformly stirring titanium dioxide nano-powder, colored silica gel A / B, and a solvent in a ratio of 1:5:5:30, and is evenly sprayed on the surface of the transparent silica gel multiple times by an airbrush to form a matte coating, enhancing the reflection of light and isolating external ambient light, and improving the sensitivity of the sensor.

[0018] Further, the hexagonal tray is printed with transparent resin and has a light transmittance of more than 90%.

[0019] Further, the red light source, green light source, and blue light source are composed of three parallel light sources of three different colors of red, green, and blue, wherein:

[0020] Each parallel light source is composed of a 2×4 array of LEDs; the LED array is installed at an inclination of 71° to the outside of the hexagonal tray to illuminate the entire surface of the elastomer. Due to rotational symmetry, the illumination intensities of R, G, and B near the center of the elastomer surface are equal.

[0021] Furthermore, the red light source, green light source, blue light source, and camera are all sealed with sealant.

[0022] Furthermore, the inlet pipe and the outlet pipe have the same diameter and each contains two 90° elbows to guide the water flow to enter / exit the device stably and effectively block the external ambient light.

[0023] The present invention also provides an underwater hydrostatic pressure self-balanced visual-tactile perception method implemented based on the underwater hydrostatic pressure self-balanced visual-tactile perception device, including:

[0024] S1. Using the light intensity images of the elastomer film before and after deformation collected by the camera as the tactile reference image and the tactile deformation image, wherein the tactile deformation image represents the deformation information generated by the contact between the elastomer and the object;

[0025] S2. Adopting an image processing algorithm and a light compensation algorithm to correct the abnormal image points caused by the existence of multiple micro through-holes on the surface of the elastomer;

[0026] S3. Since the light intensity image is distorted due to the refraction attenuation effect when light propagates in multiple media such as water and silica gel, a geometric refraction correction method is used to correct the distortion of the light intensity image caused by the refraction at the silica gel-water interface;

[0027] S4. Applying known pressures at various positions of the elastomer using a calibration ball with a known radius and a force sensor indenter respectively, and capturing the deformed light intensity image through the camera;

[0028] S5. Constructing a look-up table respectively to obtain the mappings from color to gradient and from color to force, and collecting the color, position, gradient, and pressure of each pixel point;

[0029] S6. When the elastomer contacts an unknown object, capturing the light intensity images before and after deformation by the camera, performing differential processing after correction by the geometric refraction correction method, and then performing edge detection to extract the contour to accurately obtain the position of the contact area;

[0030] S7. According to the known RGB color distribution, force magnitude, and indenter diameter at the contact position, constructing a look-up table. When the elastomer contacts an unknown object, by capturing the color distribution in the image, looking up and obtaining the corresponding pressure magnitude in the look-up table;

[0031] S8. According to the known RGB color distribution and calibration ball diameter at the contact position, using spherical harmonic functions to calculate the illumination gradient distribution of all pixel points on the pressed spherical surface, constructing a look-up table. When the elastomer contacts an unknown object, by capturing the color distribution in the image, looking up and obtaining the light intensity gradient of each pixel point in the look-up table;

[0032] S9. Based on the obtained light intensity gradient, use the fast Poisson solver in the photometric stereo algorithm to calculate the pressing depth of each pixel, and finally accurately obtain the texture and shape of the contact area.

[0033] Further, step S4 specifically includes:

[0034] S41. Place a calibration ball with a diameter of 3 mm at various positions of the elastomer, press the calibration ball, the elastomer deforms, and the camera synchronously captures the deformation information generated by the contact.

[0035] S42. Use a indenter with a diameter of 4 mm to press the elastomer successively, with the applied force ranging from 0.1 N to 5 N, increasing by 0.1 N each time; when pressing the elastomer at different positions, the elastomer deforms, and the camera synchronously captures the deformation information of the contact area.

[0036] Further, step S5 specifically includes:

[0037] S51. Through the deformation information generated by the contact captured in step S41, combined with the known calibration ball diameter and camera resolution, calculate the illumination gradient distribution of all pixels on the pressing sphere through spherical harmonic functions.

[0038] S52. Through the deformation information of the contact area captured in step S42, combined with the known indenter diameter, camera resolution, pressing force and depth, establish the corresponding relationship between the image features and the force by combining the values of the pixels in the RGB color channels.

[0039] Further, step S6 specifically includes:

[0040] S61. Perform image smoothing processing on the corrected tactile reference image and tactile deformed image to remove noise, and through threshold processing, filter out smaller changes and retain significant deformation areas.

[0041] S62. Perform pixel-level subtraction operations on the processed tactile reference image and tactile deformed image to obtain a difference image.

[0042] S63. Based on the difference image, use the Sobel or Canny edge detection algorithm to extract the deformed boundary and obtain the position of the underwater contact area.

[0043] Compared with the prior art, the present invention has the following advantages:

[0044] 1. Inspired by the deep-sea biological mechanism of balancing hydrostatic pressure, for the first time, underwater multi-dimensional information tactile perception of a visual tactile sensor is realized, with outstanding advantages such as resistance to hydrostatic pressure and low cost.

[0045] 2. The underwater hydrostatic pressure self-balancing visual and tactile perception device provided by the present invention is provided with through holes of different diameters on the lower side of the side wall of the protective cover and on the surface of the elastic body, which can quickly maintain the pressure balance inside and outside the device. Compared with the existing visual and tactile sensors, it has the advantage of being resistant to high hydrostatic pressure.

[0046] Based on the above reasons, the present invention can be widely promoted in the fields of visual and tactile sensors, underwater visual and tactile perception, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0048] Figure 1 It is a schematic diagram of the overall device for underwater hydrostatic pressure self-balancing visual and tactile perception of the present invention.

[0049] Figure 2 It is a schematic diagram of the disassembled device for underwater hydrostatic pressure self-balancing visual and tactile perception of the present invention.

[0050] Figure 3 It is a schematic diagram of the overall system for underwater hydrostatic pressure self-balancing visual and tactile perception of the present invention.

[0051] Figure 4 It is a detailed schematic diagram of the elastic body of the present invention.

[0052] Figure 5 It is a flowchart of the method for underwater hydrostatic pressure self-balancing visual and tactile perception of the present invention.

[0053] In the figure: 1. Elastic body; 2. Hexagonal tray; 3. Red light source; 4. Red light source; 5. Red light source; 6. Camera; 7. Protective cover; 8. Base; 9. Water inlet pipe; 10. Water outlet pipe; 11. Computer; 12. Transparent silica gel; 13. Reflective film; 14. Micro through hole. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0054] It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other. The following will refer to the drawings and combine the embodiments to detail the present invention.

[0055] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are only a part rather than all of the embodiments of the present invention. The description of at least one exemplary embodiment below is actually only illustrative and in no way restricts the present invention and its application or use. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0056] It should be noted that the terms used herein are for the purpose of describing specific embodiments only and are not intended to limit the exemplary embodiments of the present invention. As used herein, unless the context clearly dictates otherwise, the singular forms are also intended to include the plural forms. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they specify the presence of the described features, steps, operations, devices, components, and / or combinations thereof.

[0057] Unless otherwise specifically stated, the relative arrangements, numerical expressions, and numerical values of the components and steps set forth in these embodiments do not limit the scope of the present invention. At the same time, it should be clear that, for the convenience of description, the dimensions of the various parts shown in the drawings are not drawn in actual proportional relationships. Technologies, methods, and devices known to those of ordinary skill in the relevant art may not be discussed in detail, but where appropriate, the said technologies, methods, and devices should be regarded as part of the authorized specification. In all the examples shown and discussed here, any specific values should be construed as merely exemplary and not as a limitation. Therefore, other examples of the exemplary embodiments may have different values. It should be noted that like reference numerals and letters denote like items in the following drawings, and thus, once an item is defined in one drawing, further discussion thereof is not required in subsequent drawings.

[0058] In the description of the present invention, it should be understood that the orientation or positional relationships indicated by the orientation terms such as "front, rear, upper, lower, left, right", "lateral, vertical, perpendicular, horizontal" and "top, bottom", etc. are usually based on the orientation or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description. Without contrary description, these orientation terms do not indicate and imply that the device or element referred to must have a specific orientation or be constructed and operated in a specific orientation, and thus should not be construed as limiting the protection scope of the present invention. The orientation terms "inner, outer" refer to the inside and outside relative to the contour of each component itself.

[0059] For ease of description, spatial relative terms such as "above", "over", "on the upper surface", "upper" etc. can be used here to describe the spatial positional relationship of a device or feature shown in the figure with other devices or features. It should be understood that spatial relative terms are intended to encompass different orientations in use or operation in addition to the orientation described in the figure for the device. For example, if the device in the attached drawing is inverted, a device described as "above or over other devices or structures" will then be positioned "below or under other devices or structures". Thus, the exemplary term "above" can include both the orientations of "above" and "below". The device can also be positioned in other different ways (rotated 90 degrees or in other orientations), and corresponding interpretations are made for the spatial relative descriptions used here.

[0060] In addition, it should be noted that the use of terms such as "first", "second" etc. to limit components is only for the convenience of differentiating the corresponding components. Without additional statements, the above terms have no special meanings, and thus should not be construed as limiting the protection scope of the present invention.

[0061] As Figure 1 、 2 shown in FIGS. 3 etc., the present invention provides an underwater hydrostatic pressure self-balanced visual and tactile perception device, including: an elastomer 1, a hexagonal tray 2, a red light source 3, a green light source 4, a blue light source 5, a camera 6, a protective cover 7, a base 8, an inlet pipe 9, an outlet pipe 10, and a computer device 11, where:

[0062] The elastomer 1 is embedded in the upper half of the hexagonal tray 2 and is used to provide real-time feedback on the physical information of the contacted object, including the texture and shape of the surface of the contacted object;

[0063] The hexagonal tray 2 has the elastomer 1 placed inside and the red light source 3, green light source 4, and blue light source 5 installed outside, achieving a uniform light distribution effect;

[0064] The red light source 3, green light source 4, and blue light source 5 are used to provide parallel RGB three-color illumination;

[0065] The camera 6 is used to collect the light intensity images of the elastomer 1 deformed by touch pressure;

[0066] The protective cover 7 is used to isolate external light and fix the hexagonal tray 2 and protect the overall structure of the underwater hydrostatic pressure self-balanced visual and tactile perception device;

[0067] The base 8 is used to support and protect the camera 6, and the protective cover 7 is fixedly connected to the base 8 using screws;

[0068] The water inlet pipe 9 and the water outlet pipe 10 are respectively connected to both ends below the side wall of the protective cover 7, and are used to balance the water pressure inside and outside the device;

[0069] The computer 11 is connected to the camera 6 through the communication interface on the camera 6, and is used to control the camera 6 to take pictures and receive the collected images or videos, and use computer vision algorithms to model the collected images or videos into different tactile features.

[0070] The underwater hydrostatic self-balancing visual-tactile perception device provided by the present invention has the following working principle:

[0071] When the device reaches hydrostatic balance, the elastic body 1 film is not affected by the hydrostatic pressure and is deformed or torn. Turn on the red light source 3, the green light source 4, the blue light source 5 and the camera 6. The visual-tactile sensor captures the deformed light intensity image generated by contact through the camera 6, and then uses image processing algorithms and geometric refraction correction methods to remove the abnormal points and distortions existing in the light intensity image caused by hydrostatic balance; obtain the mapping from color to gradient and force by pre-calibrating and constructing a look-up table. Based on this, accurately obtain the position, pressure, texture and shape of the contact area according to the edge detection algorithm and the photometric stereo algorithm.

[0072] In specific implementation, as a preferred implementation manner of the present invention, as Figure 4 shown, the elastic body 1 includes a transparent silica gel 12 and a reflective film 13, wherein:

[0073] The transparent silica gel 12 is hexagonal prism-shaped and is provided with a plurality of micron-level through holes 14; in this embodiment, preferably, the diameter of the through holes 14 is 100 μm;

[0074] The reflective film 13 is a solution obtained by uniformly stirring titanium dioxide nano-powder, coloring silica gel A / B and a solvent in a ratio of 1:5:5:30, and is uniformly sprayed on the surface of the transparent silica gel 12 by an airbrush for multiple times to form a matte coating, which enhances the reflection of light signals and isolates external ambient light, and improves the sensitivity of the sensor.

[0075] In this embodiment, one layer of marking points on the elastic body in the traditional visual-tactile sensor is removed, and the effective tactile perception area is increased.

[0076] In specific implementation, as a preferred implementation manner of the present invention, continue to refer to Figure 2 , the hexagonal tray 2 is printed with transparent resin, and the light transmittance is above 90%.

[0077] In specific implementation, as a preferred implementation manner of the present invention, continue to refer to Figure 2 , the red light source 3, the green light source 4, and the blue light source 5 are composed of three parallel light sources in three different directions of red, green, and blue, wherein:

[0078] Each parallel light source is composed of an LED 2×4 array; the LED array is installed at an inclination of 71° to the outside of the hexagonal tray 2 to illuminate the entire surface of the elastomer 1. Due to rotational symmetry, the light intensities of R, G, and B near the center of the surface of the elastomer 1 are equal.

[0079] In specific implementation, as a preferred implementation manner of the present invention, continue to refer to Figure 2 , the red light source 3, the green light source 4, the blue light source 5, and the camera 6 are all sealed with sealant.

[0080] In specific implementation, as a preferred implementation manner of the present invention, continue to refer to Figure 2 , the diameters of the water inlet pipe 9 and the water outlet pipe 10 are the same, and both include two 90° elbows to guide the water to flow into / out of the device stably and effectively block the external ambient light.

[0081] As shown in Figure 5 , the present invention also provides an underwater hydrostatic pressure self - balancing visual - tactile perception method implemented based on the above - mentioned underwater hydrostatic pressure self - balancing visual - tactile perception device, including:

[0082] S1. Using the light intensity images of the elastomer film before and after deformation collected by the camera 6 as the tactile reference image and the tactile deformation image, wherein the tactile deformation image represents the deformation information generated by the contact between the elastomer 1 and the object;

[0083] S2. Since there are multiple micro - through - holes 14 on the surface of the elastomer 1, light is reflected out from the transparent silica gel 12, resulting in abnormal points at the positions of the through - holes 14 in the light intensity image. An image processing algorithm and a light compensation algorithm are used to correct the abnormal points;

[0084] S3. Since the light is refracted and attenuated when propagating in multiple media such as water and silica gel, resulting in distortion of the light intensity image, a geometric refraction correction method is used to correct the distortion of the light intensity image caused by the refraction at the silica gel - water interface;

[0085] S4. Applying known pressures at various positions of the elastomer 1 using a calibration ball with a known radius and a force sensor indenter respectively, and capturing the deformed light intensity images through the camera;

[0086] S5. Constructing lookup tables respectively to obtain the mappings from color to gradient and from color to force, and collecting the color, position, gradient, and pressure of each pixel point;

[0087] S6. When the elastomer contacts an unknown object, the camera 6 captures the light intensity images before and after deformation. After being corrected by the geometric refraction correction method, differential processing is performed, and then edge detection is carried out to extract the contour, and the position of the contact area is accurately obtained;

[0088] S7. Construct a lookup table based on the known RGB color distribution, force magnitude, and indenter diameter at the contact position. When the elastomer 1 contacts an unknown object, capture the color distribution in the image and look up and obtain the corresponding pressure magnitude in the lookup table.

[0089] S8. Based on the known RGB color distribution and calibration ball diameter at the contact position, use spherical harmonic functions to calculate the illumination gradient distribution of all pixel points on the pressed spherical surface, and construct a lookup table. When the elastomer 1 contacts an unknown object, capture the color distribution in the image and look up and obtain the light intensity gradient of each pixel point in the lookup table.

[0090] S9. Based on the obtained light intensity gradient, use the fast Poisson solver in the photometric stereo algorithm to calculate the pressing depth of each pixel point (i.e., the distance from the unpressed reference plane to the pressed surface), and finally accurately obtain the texture and shape of the contact area.

[0091] In specific implementation, as a preferred implementation manner of the present invention, step S4 specifically includes:

[0092] S41. Place a calibration ball (bearing ball) with a diameter of 3 mm at various positions of the elastomer 1, press the calibration ball, the elastomer 1 deforms, and the camera 6 synchronously captures the deformation information generated by the contact.

[0093] S42. Use an indenter with a diameter of 4 mm (from a pressure sensor) to press the elastomer 1 successively, with the applied force range from 0.1 N to 5 N, increasing by 0.1 N each time; when pressing the elastomer 1 at different positions, the elastomer deforms, and the camera 6 synchronously captures the deformation information of the contact area.

[0094] In specific implementation, as a preferred implementation manner of the present invention, step S5 specifically includes:

[0095] S51. Calculate the illumination gradient distribution of all pixel points on the pressed spherical surface through the deformation information generated by the contact captured in step S41, and in combination with the known calibration ball diameter and the resolution of the camera 6, using spherical harmonic functions.

[0096] S52. Establish the correspondence between the image features and the force through the deformation information of the contact area captured in step S42, and in combination with the known indenter diameter, the resolution of the camera 6, the pressing force, and the depth, and combining the values of the RGB color channel pixels.

[0097] In specific implementation, as a preferred implementation manner of the present invention, step S6 specifically includes:

[0098] S61. Perform image smoothing processing on the corrected tactile reference image and the tactile deformation image to remove noise, and through threshold processing, filter out smaller changes and retain significant deformation regions.

[0099] S62. Perform a pixel-level subtraction operation on the processed tactile reference image and the tactile deformation image to obtain a difference image;

[0100] S63. Based on the difference image, use the Sobel or Canny edge detection algorithm to extract the boundary of the deformation and obtain the position of the underwater contact area.

[0101] In summary, the underwater hydrostatic pressure self-balanced visual tactile perception device and method provided by the present invention achieve hydrostatic pressure balance of the tactile sensor device by designing the water inlet pipe 9, the water outlet pipe 10 and a plurality of micro through holes 14, and the elastic body 1 film is not deformed by the influence of water pressure. When the elastic body 1 deforms, the red light source 3, the green light source 4, and the blue light source 5 provide three-color illumination, and the camera 6 captures the light intensity image of the film and observes it in real time on the computer 11. By using image processing algorithms and geometric refraction correction methods, abnormal points and distortions existing in the light intensity image caused by hydrostatic pressure balance are removed; the mapping from color to gradient and force is obtained by pre-calibration and constructing a look-up table. Based on this, the position, pressure, texture and shape of the contact area are accurately obtained according to the edge detection algorithm and the photometric stereo algorithm, and the underwater multi-dimensional information tactile perception of the visual tactile sensor is realized for the first time, which has significant advantages such as anti-hydrostatic pressure, low cost and the ability to obtain multi-dimensional information simultaneously.

[0102] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An underwater hydrostatic pressure self-balanced visual and tactile perception device, characterized in that Comprising: An elastomer (1), a hexagonal tray (2), a red light source (3), a green light source (4), a blue light source (5), a camera (6), a protective cover (7), a base (8), a water inlet pipe (9), a water outlet pipe (10) and a computer device (11), wherein: The elastomer (1) is embedded in the upper half of the hexagonal tray (2) and is used to real-time feedback the physical information of the contacted object, including the texture and shape of the surface of the contacted object; The hexagonal tray (2) has the elastomer (1) placed inside and the light source (3) installed outside, achieving a uniform light distribution effect; The red light source (3), green light source (4), and blue light source (5) are used to provide parallel RGB three-color illumination; The camera (6) is used to collect the light intensity image of the elastomer (1) deforming under pressure; The protective cover (7) is used to isolate external light and fix the hexagonal tray (2) and protect the overall structure of the underwater static pressure self-balanced visual-tactile perception device; The base (8) is used to support and protect the camera (6), and the protective cover (7) is fixedly connected to the base (8) using screws; The water inlet pipe (9) and the water outlet pipe (10) are respectively connected to both ends below the side wall of the protective cover (7) and are used to balance the water pressure on both sides inside and outside the device; The computer device (11) is connected to the camera (6) through the communication interface on the camera (6), used to control the camera (6) to take pictures and receive the collected images or videos, and use computer vision algorithms to model the collected images or videos into different tactile features.

2. The underwater hydrostatic pressure self-balanced visual and tactile perception device according to claim 1, wherein The elastomer (1) includes transparent silica gel (12) and a reflective film (13), wherein: The transparent silica gel (12) is hexagonal prism-shaped and is provided with a plurality of micron-level through holes (14); The reflective film (13) is a solution of titanium dioxide nano-powder, colored silica gel A / B and a solvent stirred evenly in a ratio of 1:5:5:30, and is evenly sprayed on the surface of the transparent silica gel (12) multiple times to form a matte coating, enhancing light reflection and isolating external ambient light, and improving the sensitivity of the sensor.

3. The underwater hydrostatic pressure self-balancing visual and tactile perception device according to claim 1, wherein The hexagonal tray (2) is printed with transparent resin and has a light transmittance of more than 90%.

4. The underwater hydrostatic pressure self-balanced visual and tactile perception device according to claim 1, characterized in that The red light source (3), green light source (4), and blue light source (5) are composed of three parallel light sources of three different colors of red, green, and blue, wherein: Each parallel light source is composed of an LED 2×4 array; the LED array is installed at an inclination of 71° to the outside of the hexagonal tray (2) to illuminate the entire surface of the elastomer (1). Due to rotational symmetry, the illumination intensities of R, G, and B near the center of the surface of the elastomer (1) are equal.

5. An underwater hydrostatic self-balanced visual and tactile perception device according to claim 1, characterized in that, The red light source (3), green light source (4), blue light source (5) and the camera (6) are all sealed with sealant.

6. The underwater hydrostatic self - balancing visual - tactile perception device according to claim 1, characterized in that, The water inlet pipe (9) and the water outlet pipe (10) have the same diameter and both include two 90° elbows to guide the water flow into and out of the device stably and effectively block external ambient light.

7. An underwater hydrostatic pressure self-balancing visual and tactile perception method implemented by the underwater hydrostatic pressure self-balancing visual and tactile perception device according to any one of claims 1-6, characterized in that, Comprising: S1. Using the light intensity images of the elastomer film before and after deformation collected by the camera (6) as the tactile reference image and the tactile deformation image, wherein the tactile deformation image reflects the deformation situation generated by the contact between the elastomer (1) and the object; S2. Use image processing algorithms and light compensation algorithms to correct the image abnormal points caused by multiple micro-vias (14) on the surface of the elastomer (1); S3. Use the geometric refraction correction method to correct the distortion of the light intensity image caused by the refraction at the silicone-water interface; S4. Apply known pressures at various positions of the elastomer (1) using a calibration ball with a known radius and a force sensor indenter respectively, and capture the deformed light intensity images through a camera; S5. Construct look-up tables respectively to obtain the mappings from color to gradient and from color to force, and collect the color, position, gradient, and pressure of each pixel point; S6. The camera (6) captures the light intensity images before and after the deformation of the elastomer. After being corrected by the geometric refraction correction method, differential processing is performed, and then edge detection is carried out to accurately obtain the position of the contact area; S7. According to the known RGB color distribution, force magnitude, and indenter diameter at the contact position, construct a look-up table. When the elastomer (1) contacts an unknown object, by capturing the color distribution in the image, look up and obtain the corresponding pressure magnitude in the look-up table; S8. According to the known RGB color distribution and calibration ball diameter at the contact position, use the spherical harmonic function to calculate the illumination gradient distribution of all pixel points on the pressed spherical surface, construct a look-up table. When the elastomer (1) contacts an unknown object, by capturing the color distribution in the image, look up and obtain the light intensity gradient of each pixel point in the look-up table; S9. Based on the obtained light intensity gradient, use the fast Poisson solver in the photometric stereo algorithm to calculate the pressing depth of each pixel point, and finally accurately obtain the texture and shape of the contact area.

8. An underwater hydrostatic self-balanced visual and tactile perception method according to claim 7, characterized in that Step S4 specifically includes: S41. Place a calibration ball with a diameter of 3 mm at various positions of the elastomer (1), press the calibration ball, the elastomer (1) deforms, and the camera (6) synchronously captures the deformation information generated by the contact; S42. Use an indenter with a diameter of 4 mm to press the elastomer (1) successively, with the applied force range from 0.1 N to 5 N, increasing by 0.1 N each time; when pressing the elastomer (1) at different positions, the elastomer deforms, and the camera (6) synchronously captures the deformation information of the contact area.

9. The underwater hydrostatic pressure self-balanced visual and tactile perception method according to claim 7, characterized in that, Step S5 specifically includes: S51. Based on the deformation information generated by the contact captured in step S41, and combined with the known calibration ball diameter and the resolution of the camera (6), calculate the illumination gradient distribution of all pixel points on the pressed spherical surface through the spherical harmonic function; S52. Based on the deformation information of the contact area captured in step S42, and combined with the known indenter diameter, the resolution of the camera (6), the pressing force, and the depth, establish the corresponding relationship between the image features and the force by combining the values of the pixels in the RGB color channels.

10. An underwater hydrostatic self - balancing visual - tactile perception method according to claim 7, characterized in that, Step S6 specifically includes: S61. Perform image smoothing processing on the corrected tactile reference image and tactile deformation image to remove noise, and through threshold processing, filter out the smaller changes and retain the significant deformation areas; S62. Perform pixel-level subtraction operation on the processed tactile reference image and tactile deformation image to obtain a differential image; S63. Based on the differential image, use the Sobel or Canny edge detection algorithm to extract the boundary of the deformation and obtain the position of the underwater contact area.

Images