Measurement mark tracking method and device based on AI vision

Abstract

The invention belongs to the technical field of AI vision measurement, and particularly relates to a measurement mark tracking method and device based on AI vision, and the device comprises an AI vision perception module, an IMU inertial measurement module, a data fusion calculation module and an execution control module, the visual information acquisition device is deployed on a measurement object; through fusion of AI vision and IMU, interference of complex working conditions such as environment illumination change, dynamic motion blurring, camera motion and mark point blurring on measurement operation is overcome, the problems that single vision tracking is easily interfered by the environment and single IMU tracking has accumulative drifting are solved, real-time, stable and high-precision measurement in the complex environment is achieved, and the measurement precision is improved. The method can effectively adapt to various measurement scenes such as unmanned aerial vehicle inspection, industrial assembly, automatic driving calibration and the like.

Description

Technical Field

[0002] This invention belongs to the field of AI vision measurement technology, specifically a measurement mark tracking method and device based on AI vision. Background Technology

[0003] In complex operating conditions, vision-based measurement and positioning systems often face numerous challenges, such as changes in ambient lighting, dynamic motion blur, camera movement, and marker blurring, making stable and accurate target tracking extremely difficult. Traditional vision-based marker stabilization tracking methods typically rely on single-sensor data (such as visual images) for target localization, but in complex environments, the reliability and accuracy of a single sensor are limited.

[0004] In recent years, the integration of AI and vision technologies has brought about groundbreaking progress in sign tracking, especially with deep learning and computer vision algorithms enabling more accurate sign recognition and tracking. Meanwhile, the introduction of IMU technology has provided crucial support for dynamic motion and displacement measurement. However, each of these technologies still faces limitations, and how to efficiently fuse multi-sensor data from AI, vision, and IMU to improve the stability and tracking accuracy of signs under complex conditions remains a pressing technical challenge. Summary of the Invention

[0005] To overcome the shortcomings of existing technologies and solve the aforementioned technical problems, this invention proposes a measurement mark tracking method and device based on AI vision.

[0006] The technical solution adopted by the present invention to solve its technical problem is as follows: The present invention proposes a measurement mark tracking device based on AI vision, including an AI vision perception module, an IMU inertial measurement module, a data fusion calculation module and an execution control module. The AI ​​vision perception module includes a visual information acquisition device, which is deployed on the measurement object. The visual information acquisition device includes a mounting plate, on which a mounting cylinder is rotatably mounted in a mounting groove in the middle of the mounting plate. A measuring block is slidably disposed inside the mounting cylinder. The measuring block is connected to a telescopic device disposed on the inner wall of the bottom of the mounting cylinder. The mounting cylinder is connected to a rotating device disposed on the inner wall of the mounting groove, which drives the mounting cylinder to rotate relative to the mounting groove. The measuring block has evenly spaced measuring grooves on its sidewalls. A measuring camera is installed inside the measuring block, with the camera's acquisition end extending into the measuring groove. The lens surface of the camera's acquisition end is covered with a transparent protective film. The inside of the mounting cylinder is connected to an external inflation device.

[0007] Preferably, the inner wall of the measuring groove is provided with a movable groove, the measuring camera is slidably embedded in the movable groove, and is connected to the inner wall of the movable groove by an elastic element; An air guide hole is provided at the center of the bottom of the measuring block. The air guide hole extends upward and is connected to each moving slot through a connecting hole.

[0008] Preferably, the measuring camera has an annular block formed by an outward protrusion of the annular portion surrounding the lens on the acquisition end. The inner ring of the annular block is provided with a flushing hole, the opening of which points towards the surface of the protective film. The flushing hole communicates with the connecting hole through the guide channel in the outer housing of the measuring camera.

[0009] Preferably, a purification tank is provided on the inner wall of the mounting cylinder at the location corresponding to the acquisition end of the measuring camera. A purification block is provided on the inner wall of the purification tank. A limiting groove is provided in the middle of the purification block. The inner wall of the limiting groove has an arc-shaped cross-section and is recessed in the direction away from the measuring camera. A purification hole is provided on the purification block. The purification hole communicates with the purification chamber inside the purification block. The purification chamber is communicated with the air inlet of the air filling device through an air pipe.

[0010] Preferably, an air inlet is provided on the inner wall of the annular block at the location of the gap between the protective film and the surface of the measuring camera lens. The air inlet is connected to the flow channel, and a through hole is provided on the surface of the protective film to guide the airflow in the gap area between the protective film and the measuring camera lens.

[0011] Preferably, a retaining ring is provided at the end of the purification block corresponding to the annular block, and an annular adjustment groove is provided at the end of the annular block. An adjustment ring is slidably provided inside the adjustment groove. The adjustment ring covers the air inlet, and a through groove is provided at the part of the inner wall of the adjustment ring corresponding to the air guide hole. A filter screen is provided inside the through groove. The end of the adjustment ring extends from the opening of the adjustment groove to the outside, and the end of the adjustment ring located inside the adjustment groove is connected to the inner wall of the adjustment groove through an elastic element.

[0012] Preferably, the through hole is located in the middle of the protective film, the arc-shaped recess on the inner wall of the limiting groove corresponding to the measuring camera is the elastic part, and the purification holes are distributed in a ring on the surface of the elastic part near the retaining ring.

[0013] Preferably, a purification screen is provided inside the purification chamber. The purification screen is slidably connected to the inner wall of the purification chamber through a limiting ring provided on the outer ring, and the output end of the propulsion device provided on the inner wall of the purification chamber is connected to the limiting ring. An adjustment rod is installed in the middle of the purification net. The middle part of the adjustment rod is connected to the center of the elastic part. The outer surface of the middle part of the elastic part is raised to form a closed protrusion.

[0014] Preferably, the outer surface of the elastic part is uniformly provided with guide bumps, the guide bumps form a ring structure around the central closed bump, and the cross section of the guide bumps is triangular.

[0015] A measurement marker tracking method based on AI vision, wherein the measurement marker tracking method utilizes the aforementioned measurement marker tracking device, and the specific steps include: S1: Configure the visual information acquisition device and IMU sensor in the AI ​​visual perception module onto the measurement object, and collect image data and IMU data of the target point on the measurement object during the operation of the measurement object; S2: Denoise and edge enhancement are performed on the image data, invalid marker coordinates caused by environmental interference are removed, and zero drift compensation and low-pass filtering are performed on the IMU data to reduce the noise of the IMU sensor. S3: Input the preprocessed image data and IMU data into the tightly coupled fusion algorithm for calculation, and output the motion trajectory, velocity, and acceleration data of the positioning markers of the measured object in real time; S4: Based on the calculated motion trajectory, velocity, and acceleration data, a measurement report of the positioning markers is generated to serve as data support for the automatic navigation system of the measured object.

[0016] The beneficial effects of this invention are as follows: The present invention discloses a measurement marker tracking method and device based on AI vision. By integrating AI vision with IMU, it overcomes the interference of complex working conditions such as changes in ambient light, dynamic motion blur, camera movement, and marker blur on measurement operations. It also solves the problems of single vision tracking being susceptible to environmental interference and single IMU tracking having cumulative drift. It achieves real-time, stable, and high-precision measurement in complex environments and can be effectively adapted to various measurement scenarios such as UAV inspection, industrial assembly, and autonomous driving calibration. Attached Figure Description

[0017] The invention will now be further described with reference to the accompanying drawings.

[0018] Figure 1 This is a perspective view of the visual information acquisition device in this invention; Figure 2 This is a cross-sectional view of the visual information acquisition device in this invention; Figure 3 yes Figure 2 A magnified view of a section at point A in the middle; Figure 4 yes Figure 3 A magnified view of a section at point B in the middle; Figure 5 This is a flowchart of the measurement marker tracking method in this invention.

[0019] In the diagram: Visual information acquisition device 1, mounting plate 11, mounting groove 111, mounting cylinder 12, measuring block 13, measuring groove 131, moving groove 132, air guide hole 133, connecting hole 134, measuring camera 14, annular block 141, flushing hole 142, flow guide channel 143, air filling hole 144, adjusting groove 145, adjusting ring 146, filter screen 147, protective membrane 15, through hole 151, purification tank 16, purification block 161, limiting groove 162, purification hole 163, purification chamber 164, retaining ring 165, elastic part 166, sealing protrusion 167, flow guide protrusion 168, purification net 17, limiting ring 171, adjusting rod 172. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Example 1: As shown in the attached diagram of the instruction manual. Figures 1-4 As shown, this application proposes a measurement mark tracking device based on AI vision, including an AI vision perception module, an IMU inertial measurement module, a data fusion calculation module and an execution control module. The AI ​​vision perception module includes a visual information acquisition device 1, which is deployed on the measurement object. The visual information acquisition device 1 includes a mounting plate 11. A mounting cylinder 12 is rotatably mounted on a mounting groove 111 in the middle of the mounting plate 11. A measuring block 13 is slidably arranged inside the mounting cylinder 12. The measuring block 13 is connected to a telescopic device provided on the inner wall of the bottom of the mounting cylinder 12. The mounting cylinder 12 is connected to a rotating device provided on the inner wall of the mounting groove 111 to drive the mounting cylinder 12 to rotate relative to the mounting groove 111. Specifically, regarding the specific structure of the rotating device, this embodiment provides a possible implementation scheme. Specifically, a ring gear can be provided on the outer surface of the mounting cylinder 12, which meshes with the drive gear in the inner wall of the mounting groove 111. The drive gear is connected to the drive motor. When the drive motor starts, it drives the drive gear to rotate through the output end, and then drives the mounting cylinder 12 to rotate relative to the mounting groove 111 through meshing transmission, thereby realizing the rotation adjustment of the mounting cylinder 12. Measuring slots 131 are evenly distributed on the side wall of the measuring block 13. A measuring camera 14 is installed inside the measuring block 13. The acquisition end of the measuring camera 14 extends into the measuring slot 131. The lens surface of the acquisition end of the measuring camera 14 is covered with a transparent protective film 15. The inside of the mounting cylinder 12 is connected to the external inflation device.

[0022] Specific workflow: Addressing the issue that existing visual measurement, positioning, and navigation systems configured for drones, automobiles, and other equipment struggle to operate normally under complex conditions due to variations in ambient light, dynamic motion blur, camera movement, and blurred marker points, this application provides an AI vision-based measurement marker tracking device that effectively overcomes these problems. Specifically, the portable visual information acquisition device 1 from the AI ​​vision perception module of this application is placed on the target device, which serves as the measurement object. Image information of the positioning marker points on the moving target device is acquired from its moving perspective, while fixed cameras on surrounding road sections provide image information captured from different directions. Thus, the image information of the positioning marker points captured from different directions and angles is aggregated into the data fusion calculation module. Simultaneously, the IMU sensor configured in the IMU inertial measurement module is also configured on the target device to synchronously collect angular velocity, acceleration, and attitude angle data during the target device's movement, thus recording the target device's motion status in real time. This data is then transmitted to the data fusion and calculation module. The data fusion and calculation module preprocesses the image information and performs a tightly coupled fusion algorithm with the IMU data to calculate the trajectory, velocity, acceleration, and other parameters of the positioning markers on the target device. It continuously tracks these parameters, providing real-time and accurate data support for the target device's navigation operation based on this data. By integrating AI vision with IMU, this application solves the problems of single vision tracking being susceptible to environmental interference and single IMU tracking having cumulative drift, achieving real-time, stable, and high-precision measurement in complex environments, and is suitable for various measurement scenarios such as drone inspection, industrial assembly, and autonomous driving calibration.

[0023] Furthermore, in the above data acquisition process, in order to facilitate the normal operation of the visual information acquisition device 1, it is necessary to arrange the visual information acquisition device 1 of this application on the target device. Specifically, the mounting plate 11 is set on the front surface of the target device, and the mounting plate 11 and the target device are tightly connected by fasteners passing through the fixing holes on the mounting plate 11. At this time, the mounting cylinder 12 on the mounting plate 11 is located inside the target device. When the working environment around the target device is complex, the telescopic device located inside the mounting cylinder 12 can be activated to push the measuring block 13 inside the mounting cylinder 12 to move upward and extend out of the mounting groove 111 opening in the middle of the mounting cylinder 12, so that the measuring groove 131 on the measuring block 13 is exposed. The measuring camera 14 distributed circumferentially on the side wall of the measuring block 13 can normally capture image information of the surrounding environment after activation, and collect image data of the positioning markers and their surrounding environment on the target device. In order to expand the range, the mounting cylinder 12 can be rotated relative to the mounting plate 11 by controlling the rotating device to realize the rotation of the measuring block 13, expand the shooting range of the measuring camera 14, and make the collected image data more comprehensive. Furthermore, to better ensure the normal operation of the measuring camera 14 under complex working conditions, when there are factors in the surrounding environment that threaten the safety of the measuring camera 14, the measuring block 13 can be moved into the mounting cylinder 12 to be shielded and protected, isolating it from external threats. When the surrounding environment is dusty, the air filling device inside the mounting cylinder 12 can be activated to fill the mounting cylinder 12 with purified airflow. On the one hand, this increases the air pressure inside the mounting cylinder 12, preventing external airflow containing dust and impurities from penetrating inward. On the other hand, the higher air pressure inside the mounting cylinder 12 will cause the airflow inside the mounting cylinder 12 to permeate upward along the gap between the measuring block 13 and the inner wall of the mounting cylinder 12, thereby creating an airflow trend from the inside out around the measuring block 13, further preventing external air containing dust and impurities from approaching and affecting the normal operation of the measuring camera 14.

[0024] Example 2: Based on Embodiment 1, a movable groove 132 is provided on the inner wall of the measuring groove 131. The measuring camera 14 is slidably embedded into the movable groove 132 and is connected to the inner wall of the movable groove 132 by an elastic element, which can be a spring. A vent 133 is provided at the center of the bottom of the measuring block 13. The vent 133 extends upward and is connected to each moving slot 132 through a connecting hole 134.

[0025] The measuring camera 14 has an annular block 141 formed by the outward protrusion of the annular part around the lens on the acquisition end. The inner ring of the annular block 141 is provided with a flushing hole 142. The flushing hole 142 opens towards the surface of the protective film 15. The flushing hole 142 communicates with the connecting hole 134 through the guide channel 143 in the outer housing of the measuring camera 14.

[0026] Specific workflow: Based on the specific workflow in Embodiment 1, the measuring camera 14 located inside the measuring slot 131 can move along the moving slot 132. Therefore, when inside the mounting cylinder 12, the end of the measuring camera 14 is retracted into the moving slot 132. When the measuring camera 14 needs to work, the measuring block 13 is pushed upward and extends out of the mounting slot 111. At this time, the measuring slot 131 is exposed. In order to expand the shooting range, the mounting cylinder 12 can be inflated by the inflation device to increase the air pressure inside the mounting cylinder 12. The airflow flows in along the air guide hole 133 and enters the moving slot 132 through the connecting hole 134, which increases the air pressure in the moving slot 132 and pushes the measuring camera 14 to slide outward along the moving slot 132, so that the end of the measuring camera 14 can extend out of the opening of the measuring slot 131, expand the shooting range, and improve the efficiency of image data acquisition of the surrounding environment. Furthermore, during this process, some airflow can enter the guide channel 143. Since the guide channel 143 is located on the inner wall of the housing of the measuring camera 14 and surrounds the working components such as the processor, battery, and data transmission device inside the measuring camera 14, the airflow can absorb the heat generated during the operation of the working components through the inner wall of the guide channel 143 and discharge the heat to the outside, thereby achieving temperature control of the working components inside the measuring camera 14. Furthermore, as the airflow continues to flow out along the guide channel 143 and exits through the flushing hole 142 at the end of the annular block 141, because the opening of the air guide hole 133 points towards the surface of the protective film 15, the airflow forms a flushing airflow along the opening of the flushing hole 142 and acts on the surface of the protective film 15, carrying away dust and impurities that may adhere to the surface of the protective film 15 in complex working environments, ensuring the cleanliness of the surface of the protective film 15, thereby ensuring the clarity of the image information acquired by the lens of the measuring camera 14 covered by the protective film 15, and improving the accuracy of the final acquired positioning marker image information.

[0027] Example 3: Based on Embodiment 2, a purification tank 16 is provided on the inner wall of the mounting cylinder 12 at the corresponding position of the acquisition end of the measuring camera 14. A purification block 161 is provided on the inner wall of the purification tank 16. A retaining ring 165 is provided at the end of the purification block 161 at the corresponding position of the annular block 141. A purification hole 163 is provided on the purification block 161. The purification hole 163 communicates with the purification chamber 164 inside the purification block 161. The purification chamber 164 communicates with the air inlet of the air filling device through an air pipe. An inflation hole 144 is provided on the inner wall of the annular block 141 at the gap between the protective film 15 and the lens surface of the measuring camera 14. The inflation hole 144 communicates with the guide channel 143. An annular adjustment groove 145 is provided at the end of the annular block 141. An adjustment ring 146 is slidably provided inside the adjustment groove 145. The adjustment ring 146 covers the inflation hole 144. A through groove is provided on the inner wall of the adjustment ring 146 at the position corresponding to the air guide hole 133. A filter screen 147 is provided inside the through groove. The end of the adjustment ring 146 extends out from the opening of the adjustment groove 145 and extends to the outside. The end of the adjustment ring 146 located inside the adjustment groove 145 is connected to the inner wall of the adjustment groove 145 through an elastic element. Specific workflow: Based on the specific workflow in Example 2, in order to further purify the lens of the measuring camera 14, when it is found that the image information collected by the measuring camera 14 is blocked by dirt, the telescopic device is controlled to move the measuring block 13 downward, so that the measuring groove 131 faces the purification groove 16 located on the inner wall of the mounting cylinder 12; by activating the air extraction device, the measuring camera 14 is pushed to slide along the moving groove 132, so that the end of the measuring camera 14 extends out of the measuring groove 131 and enters the purification groove 16, until the annular block 141 at the end of the measuring camera 14 abuts against the retaining ring 165 on the purification block 161. At this time, the protruding part on the retaining ring 165 slides into the inner wall of the annular block 141 and blocks the flushing hole 142, preventing the airflow from flowing out of the opening of the flushing hole 142; Simultaneously, the moving retaining ring 165 pushes the adjusting ring 146 along the adjusting groove 145, causing the through groove on the adjusting ring 146 to move from the position corresponding to the flushing hole 142 to the position corresponding to the air filling hole 144. At this time, the air filling hole 144 remains unobstructed, while the flushing hole 142 is closed, allowing airflow to flow from the guide channel 143 into the air filling hole 144, and then from the air filling hole 144 into the gap area between the protective film 15 and the lens surface. This increases the air pressure in the gap area, causing the protective film 15 to expand and deform outward, and the airflow flushes the inner wall surface of the protective film 15 along the gap area. The airflow then flows out through the through hole 151 in the middle of the protective membrane 15 and enters the gap between the outer surface of the protective membrane 15 and the inner wall of the limiting groove 162 in the middle of the purification block 161. It then flows along the gap to the surrounding purification holes 163. By controlling the flow trajectory of the airflow, the airflow passes through the inner and outer wall surfaces of the protective membrane 15 in sequence. The vibration and impact on the protective membrane 15 can effectively remove dirt and impurities adhering to the surface of the protective membrane 15, thereby ensuring the cleanliness of the protective membrane 15. This, in turn, ensures that the image information acquired by the measuring camera 14 at the end is clear and accurate enough.

[0028] Example 4: Based on Embodiment 3, the through hole 151 is located in the middle of the protective film 15, the arc-shaped recessed part of the inner wall of the limiting groove 162 corresponding to the measuring camera 14 is the elastic part 166, and the purification hole 163 is distributed in a ring on the surface of the elastic part 166 near the retaining ring 165. The purification chamber 164 is equipped with a purification screen 17. The purification screen 17 is slidably connected to the inner wall of the purification chamber 164 through a limiting ring 171 on the outer ring. The output end of the propulsion device on the inner wall of the purification chamber 164 is connected to the limiting ring 171. The propulsion device can be a miniature electric telescopic rod device, which is controlled by an external controller. An adjusting rod 172 is provided in the middle part of the purification net 17. The middle part of the adjusting rod 172 is connected to the center part of the elastic part 166. The outer surface of the middle part of the elastic part 166 protrudes to form a closed protrusion 167. The outer surface of the elastic part 166 is uniformly provided with flow guiding protrusions 168. The flow guiding protrusions 168 form a ring structure around the central closed protrusion 167, and the cross section of the flow guiding protrusions 168 is triangular. Specific workflow: Based on the specific workflow in Embodiment 3, the inflation device includes a miniature air pump device. One end of the air pump device is connected to the interior of the upper purification chamber 164 through an air pipe, and the other end is connected to the area near the bottom of the installation cylinder 12 through an air pipe, thus realizing the circulation of airflow. Furthermore, the dust and impurities cleaned up by the circulating airflow are intercepted and accumulated inside the purification chamber 164 by the purification screen 17 set inside the purification chamber 164. The purification chamber 164 is opened periodically to clean the accumulated dust and impurities inside. Furthermore, when the airflow flows out from the through hole 151 in the middle of the protective film 15 and enters the gap area between the outer surface of the protective film 15 and the elastic part 166, it flows outward from the center of the gap area and passes through each guide protrusion 168. Because the guide protrusion 168 is conical, the airflow trajectory through the guide protrusion 168 is arc-shaped, so that the airflow presents a continuous curved flow trajectory along the gap direction, and forms a continuous impact on the corresponding surface of the protective film 15, causing the dust and impurities adhering to the surface of the protective film 15 to fall off under the continuous and violent impact. Moreover, the side surface of the guide protrusion 168 near the through hole 151 is inclined, and the side inclined on the side away from the through hole 151 is horizontal, ensuring that the airflow flows unidirectionally from the center to the edge, preventing the dust and impurities washed by the airflow from adhering to the surface of the protective film 15 again due to backflow. Furthermore, because the elastic part 166 on the inner wall of the limiting groove 162 is made of elastic material, as the airflow continues to enter the gap area, the air pressure in the gap area increases. Due to the pressure, the elastic part 166 tends to deform and indent away from the through hole 151, thereby increasing the space in the gap area. In order to further improve the cleaning efficiency of the protective film 15 surface, the propulsion device can be started by controlling the movement of the limiting ring 171 along the inner wall of the purification chamber 164. The moving purification net 17 drives the connected adjusting rod 172 to move, thereby causing the elastic part 166 to deform. When the purification net 17 drives the adjusting rod 172 to cause the elastic part 166 to bulge towards the through hole 151, the closing protrusion 167 in the middle of the elastic part 166 contacts the through hole 151, thus sealing the through hole 151. At the same time, the protruding elastic part 166 squeezes the airflow in the gap area, reducing the space and increasing the pressure, which causes the compressed airflow to accelerate the flow of the cleaned impurities into the purification hole 163. When the pushing device drives the purification net 17 to reset and retract, the elastic part 166 is recessed away from the through hole 151, increasing the space in the gap area and forming a negative pressure. This accelerates the airflow from the through hole 151 into the gap area between the protective membrane 15 and the elastic part 166. This process is repeated. The intermittent changes in air pressure and airflow impact in the gap area help to accelerate the cleaning of the protective membrane 15, ensuring the cleanliness of the surface of the protective membrane 15. This ensures the clarity of the acquired images during the normal operation of the measuring camera 14 and eliminates the adverse effects of surface-adhered dust and impurities on image data acquisition.

[0029] Example 5: Based on the above embodiments, as shown in the accompanying drawings of the specification. Figure 5 As shown, a measurement sign tracking method based on AI vision is described. This method utilizes the aforementioned measurement sign tracking device and includes the following specific steps: S1: Configure the visual information acquisition device 1 and IMU sensor in the AI ​​visual perception module onto the measurement object, and collect image data and IMU data of the target point on the measurement object during the operation of the measurement object; S2: Denoise and edge enhancement are performed on the image data, invalid marker coordinates caused by environmental interference are removed, and zero drift compensation and low-pass filtering are performed on the IMU data to reduce the noise of the IMU sensor. S3: Input the preprocessed image data and IMU data into the tightly coupled fusion algorithm for calculation, and output the motion trajectory, velocity, and acceleration data of the positioning markers of the measured object in real time; S4: Based on the calculated motion trajectory, velocity, and acceleration data, a measurement report of the positioning markers is generated, serving as data support for the automatic navigation system of the measured object. The above describes and illustrates the basic principles, main features, and advantages of this invention. Those skilled in the art should understand that this invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to this invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A measurement marker tracking device based on AI vision, comprising an AI vision perception module, an IMU inertial measurement module, a data fusion calculation module, and an execution control module, characterized in that: The AI ​​visual perception module includes a visual information acquisition device (1), which is deployed on the object being measured. The visual information acquisition device (1) includes a mounting plate (11), a mounting cylinder (12) is rotatably mounted on the mounting groove (111) in the middle part of the mounting plate (11), a measuring block (13) is slidably arranged inside the mounting cylinder (12), the measuring block (13) is connected to the telescopic device provided on the bottom inner wall of the mounting cylinder (12), and the mounting cylinder (12) is connected to the rotating device provided on the inner wall of the mounting groove (111) to drive the mounting cylinder (12) to rotate relative to the mounting groove (111); Measuring slots (131) are evenly provided on the side wall of the measuring block (13). A measuring camera (14) is installed inside the measuring block (13). The acquisition end of the measuring camera (14) extends into the measuring slot (131). The lens surface of the acquisition end of the measuring camera (14) is covered with a transparent protective film (15). The inside of the mounting cylinder (12) is connected to the external inflation device.

2. The measurement mark tracking device based on AI vision according to claim 1, characterized in that: The inner wall of the measuring groove (131) is provided with a movable groove (132), and the measuring camera (14) is slidably embedded in the movable groove (132) and connected to the inner wall of the movable groove (132) by an elastic element; A vent hole (133) is provided at the center of the bottom of the measuring block (13). The vent hole (133) extends upward and is connected to each moving slot (132) through a connecting hole (134).

3. The measurement mark tracking device based on AI vision according to claim 2, characterized in that: The measuring camera (14) has an annular block (141) formed by the outward protrusion of the annular part around the lens on the acquisition end. The inner ring of the annular block (141) is provided with a flushing hole (142). The flushing hole (142) opens towards the surface of the protective film (15). The flushing hole (142) communicates with the connecting hole (134) through the guide channel (143) in the outer housing of the measuring camera (14).

4. The measurement mark tracking device based on AI vision according to claim 3, characterized in that: A purification tank (16) is provided on the inner wall of the mounting cylinder (12) at the location corresponding to the acquisition end of the measuring camera (14). A purification block (161) is provided on the inner wall of the purification tank (16). A limiting groove (162) is provided in the middle of the purification block (161). The inner wall of the limiting groove (162) has an arc-shaped cross section and is recessed in the direction away from the measuring camera (14). A purification hole (163) is provided on the purification block (161). The purification hole (163) is connected to the purification chamber (164) inside the purification block (161). The purification chamber (164) is connected to the air inlet of the air filling device through an air pipe.

5. The measurement mark tracking device based on AI vision according to claim 4, characterized in that: An air inlet (144) is provided on the inner wall of the annular block (141) at the gap between the protective film (15) and the lens surface of the measuring camera (14). The air inlet (144) is connected to the flow channel (143). A through hole (151) is provided on the surface of the protective film (15) to guide the airflow in the gap area between the protective film (15) and the lens of the measuring camera (14).

6. The measurement mark tracking device based on AI vision according to claim 5, characterized in that: A retaining ring (165) is provided at the end of the purification block (161) corresponding to the annular block (141), and an annular adjustment groove (145) is provided at the end of the annular block (141). An adjustment ring (146) is slidably provided inside the adjustment groove (145). The adjustment ring (146) covers the air inlet (144), and a through groove is provided at the part of the inner wall of the adjustment ring (146) corresponding to the air guide hole (133). A filter screen (147) is provided inside the through groove. The end of the adjustment ring (146) extends out of the opening of the adjustment groove (145) to the outside, and the end of the adjustment ring (146) located inside the adjustment groove (145) is connected to the inner wall of the adjustment groove (145) through an elastic element.

7. The measurement mark tracking device based on AI vision according to claim 6, characterized in that: The through hole (151) is located in the middle of the protective film (15). The arc-shaped recessed part of the inner wall of the limiting groove (162) corresponding to the measuring camera (14) is the elastic part (166). The purification hole (163) is distributed in a ring on the surface of the elastic part (166) near the retaining ring (165).

8. The measurement mark tracking device based on AI vision according to claim 7, characterized in that: A purification screen (17) is provided inside the purification chamber (164). The purification screen (17) is slidably connected to the inner wall of the purification chamber (164) through a limiting ring (171) on the outer ring. The output end of the propulsion device provided on the inner wall of the purification chamber (164) is connected to the limiting ring (171). An adjusting rod (172) is provided in the middle part of the purification net (17). The middle part of the adjusting rod (172) is connected to the center part of the elastic part (166). The outer surface of the middle part of the elastic part (166) is raised to form a closed protrusion (167).

9. The measurement mark tracking device based on AI vision according to claim 8, characterized in that: The outer surface of the elastic part (166) is uniformly provided with flow guiding protrusions (168). The flow guiding protrusions (168) form a ring structure around the central closed protrusion (167), and the cross section of the flow guiding protrusions (168) is triangular.

10. A measurement sign tracking method based on AI vision, wherein the measurement sign tracking method uses the measurement sign tracking device according to any one of claims 1-9, and the specific steps include: S1: Configure the visual information acquisition device (1) and IMU sensor in the AI ​​visual perception module onto the measurement object, and collect image data and IMU data of the target point on the measurement object during the operation of the measurement object; S2: Denoise and edge enhancement are performed on the image data, invalid marker coordinates caused by environmental interference are removed, and zero drift compensation and low-pass filtering are performed on the IMU data to reduce the noise of the IMU sensor. S3: Input the preprocessed image data and IMU data into the tightly coupled fusion algorithm for calculation, and output the motion trajectory, velocity, and acceleration data of the positioning markers of the measured object in real time; S4: Based on the calculated motion trajectory, velocity, and acceleration data, a measurement report of the positioning markers is generated to serve as data support for the automatic navigation system of the measured object.

Images