Gauge rule standard device based on machine vision and deep learning technology

Abstract

The utility model discloses a gauging rule standard device based on machine vision and deep learning technology, which belongs to the technical field of gauging rules and comprises a measuring mechanism, the measuring mechanism is mounted on an adjustable table, a visual detection unit is arranged on the right side of the measuring mechanism and mounted on the adjustable table, and an inertial measurement unit (IMU) is mounted on the adjustable table. A control module is mounted on the front side of the IMU, a gas distribution tool is mounted at the bottom of the adjustable table and on the rear side of the control module, the gas distribution tool is controlled by a gas pump, and nozzles are mounted at the bottoms of the two sides of the adjustable table. According to the device, the industrial camera is combined with the deep learning model to carry out visual recognition and analysis on the gauging rule, automatic scale recognition and accurate error measurement are achieved, a traditional manual visual inspection or comparison method is replaced, and the detection efficiency and accuracy are remarkably improved.

Description

TECHNICAL FIELD

[0001] The utility model relates to the technical field of track gauge ruler, more specifically, to the track gauge ruler standard based on machine vision and deep learning technology. BACKGROUND

[0002] With the continuous expansion of railway transportation network and the continuous improvement of train running speed, the accurate measurement and maintenance of track geometric parameters have become the key link to ensure the safety of railway operation. Among them, the track gauge ruler is a standard measuring tool for measuring the distance between steel rails (track gauge), super-geometric parameters, and is widely used in railway line maintenance and quality detection work.

[0003] At present, the traditional track gauge ruler calibration work generally adopts manual comparison, projection reading or mechanical caliper and other methods. These methods have the following shortcomings:

[0004] ①High degree of manual intervention, low measurement efficiency: the operator needs to manually compare the standard scale or measuring gauge, the process is tedious and easy to fatigue, and the efficiency is limited by manpower and working environment.

[0005] ②Large human error, poor repeatability: different testers have different judgments on reading position and angle, which may cause scale reading error and affect the consistency and accuracy of calibration.

[0006] ③Lack of posture compensation and environmental adaptability: during the calibration process on site, due to uneven laying of sleepers or inclination of platform, the traditional method is difficult to dynamically correct the instrument posture, which may cause distortion of measurement data.

[0007] ④Cannot realize intelligent management of data: most of the traditional measurement methods rely on manual data recording, lack of image archiving and digital output capability, and are not conducive to long-term tracking and intelligent operation and maintenance management.

[0008] Under the background of intelligent construction of new infrastructure and digital transformation of railway, it is urgent to develop a track gauge ruler calibration device integrating automatic identification, intelligent judgment, posture sensing and flexible structure to improve the standardization, intelligence and efficiency of the measurement process. Utility model content

[0009] 1. Technical problem to be solved

[0010] In view of the problems in the prior art, the utility model aims to provide a track gauge ruler standard based on machine vision and deep learning technology, which can sense the contact force between the track gauge ruler and the gauge in real time, automatically judge the pressing state, avoid measurement error caused by insufficient contact or excessive force, and has good portability and site adaptability.

[0011] 2. Technical scheme

[0012] To solve the above problems, the utility model adopts the following technical scheme.

[0013] The track gauge standard gauge based on machine vision and deep learning technology, including measuring mechanism, the measuring mechanism is installed on the adjustable table, the right side of the measuring mechanism is equipped with visual detection unit and is installed on the adjustable table, the adjustable table is installed with inertial measurement unit IMU, the front side of the inertial measurement unit IMU is installed with control module, the bottom of the adjustable table, the rear side of control module is installed with gas distribution tool, the gas distribution tool is controlled by air pump, the bottom of both sides of the adjustable table is installed with nozzle.

[0014] Further, the adjustable table includes a horizontal plate, the horizontal plate is provided with a slot, and the horizontal plate is connected with a supporting leg below through a hinge.

[0015] According to the above features, the visual detection unit includes a bracket, an LED group is installed on the bracket, and an industrial camera is installed on the LED group.

[0016] In some embodiments, the measuring mechanism includes a track gauge positioning surface, a gauge is installed below the track gauge positioning surface, a vertical plate is installed on the right side of the gauge, a sleeve slot is provided on the bottom of the vertical plate, a bottom plate is provided below the gauge, and the gauge and the vertical plate are in the same plane.

[0017] According to the above features, the connecting part between the gauge and the vertical plate is a connecting plate, a limiting cylinder is provided on the left side of the connecting plate to be sleeved on the gauge, and a sleeve frame is provided on the right side of the connecting plate to be sleeved on the vertical plate.

[0018] In some embodiments, a connecting shaft is coaxially connected to the bottom of the limiting cylinder, and a pressure sensor is installed below the connecting shaft.

[0019] According to the above features, the nozzle includes a sleeve plate, a spray head is installed on the bottom of the sleeve plate, and a connecting pipe is provided in the middle of the nozzle.

[0020] 3, beneficial effects

[0021] Compared with the prior art, the utility model has the advantages that:

[0022] 1) The device uses an industrial camera combined with a deep learning model to visually recognize and analyze the track gauge, realizes automatic identification of the scale and accurate measurement of the error, replaces the traditional manual visual inspection or comparison method, and significantly improves the detection efficiency and accuracy.

[0023] 2) The pressure sensor and the connecting shaft structure in the measuring mechanism can sense the contact force between the track gauge and the gauge in real time, automatically judge the pressing state, avoid measurement errors caused by insufficient contact or excessive force, and realize more scientific and stable measurement determination.

[0024] 3) The adjustable platform features foldable legs and adjustable slots, allowing the measurement module and vision system to be installed on the platform. It has good portability and field adaptability, and is suitable for various operating scenarios such as field, high-speed rail lines, and maintenance bases.

[0025] 4) By setting up a gas distribution tool and nozzle structure, the surface of the gauge ruler can be automatically directionally purged before measurement to remove dust, water stains and other contaminants in a timely manner, avoid interfering with the vision system's recognition, and further improve the reliability of the detection results.

[0026] 5) The measuring mechanism in this utility model has a flexible structure. Through the adjustable connection design between the gauge and the vertical plate (limiting cylinder and sleeve frame), it can be adapted to the calibration requirements of track gauges of different specifications, thereby improving the applicability and expansion capability of this device. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of the track gauge standard based on machine vision and deep learning technology of this utility model.

[0028] Figure 2 This is a schematic diagram of the adjustable platform of this utility model;

[0029] Figure 3 This is a schematic diagram of the structure of the visual inspection unit of this utility model;

[0030] Figure 4 This is a schematic diagram of the measuring mechanism of this utility model;

[0031] Figure 5 This is a schematic diagram of the structure of the gauge of this utility model;

[0032] Figure 6 This is a schematic diagram of the track gauge positioning surface of this utility model;

[0033] Figure 7 This is a schematic diagram of the nozzle structure of this utility model;

[0034] Explanation of the labels in the diagram:

[0035] 1. Vision inspection unit; 11. Industrial camera; 12. LED group; 13. Bracket; 2. Measuring mechanism; 21. Track gauge positioning surface; 211. Pressure sensor; 212. Coupling; 22. Gauge; 221. Contact; 222. Limiting cylinder; 223. Connecting plate; 224. Sleeve frame; 23. Base plate; 24. Vertical plate; 25. Sleeve slot; 3. Adjustable table; 31. Support leg; 32. Hinge; 33. Slot; 34. Horizontal plate; 4. Nozzle; 41. Spray head; 42. Connecting pipe; 43. Sleeve plate; 5. Gas distribution tool; 6. Air pump; 7. Control module; 8. Inertial measurement unit (IMU). DETAILED DESCRIPTION

[0036] The technical solutions in the embodiments of the utility model will be clearly and completely described below with reference to the drawings in the embodiments of the utility model. Obviously, the described embodiments are only part of the embodiments of the utility model, rather than all the embodiments. Based on the embodiments in the utility model, all other embodiments obtained by those skilled in the art without creative labor fall within the protection scope of the utility model.

[0037] Embodiment 1

[0038] The gauge standard for track gauge based on machine vision and deep learning technology comprises a measuring mechanism 2, the measuring mechanism 2 is installed on an adjustable table 3, a visual detection unit 1 is arranged on the right side of the measuring mechanism 2 and installed on the adjustable table 3, an inertial measurement unit IMU is installed on the adjustable table 3, a control module 7 is installed on the front side of the inertial measurement unit IMU, a gas distribution tool 5 is installed on the bottom of the adjustable table 3 and the rear side of the control module 7, the gas distribution tool 5 is controlled through a gas pump 6, and nozzles 4 are installed on the bottom of both sides of the adjustable table 3. The adjustable table 3 comprises a horizontal plate 34, the horizontal plate 34 is provided with a slot 33, and a supporting leg 31 is connected to the lower side of the horizontal plate 34 through a hinge 32. The visual detection unit 1 comprises a bracket 13, an LED group is installed on the bracket 13, and an industrial camera 11 is installed on the LED group.

[0039] In some embodiments, the measuring mechanism 2 comprises a track gauge positioning surface 21, a gauge 22 is installed below the track gauge positioning surface 21, a vertical plate 24 is installed on the right side of the gauge 22, a sleeve slot 25 is arranged on the bottom of the vertical plate 24, a bottom plate 23 is arranged below the gauge 22, and the gauge 22 and the vertical plate 24 are in the same plane. The connecting part between the gauge 22 and the vertical plate 24 is a connecting plate 223, a limiting cylinder 222 is arranged on the gauge 22 on the left side of the connecting plate 223, and a sleeve frame 224 is arranged on the vertical plate 24 on the right side of the connecting plate 223.

[0040] In some embodiments, the bottom of the limiting cylinder 222 is coaxially connected with a connecting shaft 212, and a pressure sensor 211 is installed below the connecting shaft 212. The nozzle 4 comprises a sleeve plate 43, a spray head 41 is installed on the bottom of the sleeve plate 43, and a connecting pipe 42 is arranged in the middle of the nozzle 4.

[0041] Working principle: the measuring mechanism includes a gauge spacing positioning surface 21, a gauge 22, a vertical plate 24, a base plate 23 and the like. The gauge spacing positioning surface 21 provides a standard reference surface for calibrating and fixing the gauge; the gauge 22 is the main standard measuring component, which forms a vertical support with the vertical plate 24 to ensure the calibration accuracy; the sleeve slot 25 at the bottom of the vertical plate 24 is used to stably install the measuring unit on the base structure; the gauge 22 and the vertical plate 24 are connected through the connecting plate 223, the limiting cylinder 222 on the left side of the connecting plate 223 can be slidably or sleeved outside the gauge, and the sleeve frame 224 on the right side is sleeved outside the vertical plate, so that the adjustable and stable connection is realized; the limiting cylinder 222 is coaxially connected with the connecting shaft 212 at the bottom, and the pressure sensor 211 is integrated below the connecting shaft 212, which can sense the measuring force applied by the gauge, realize automatic judgment of the contact state and the compression degree, and prevent misreading.

[0042] The adjustable table 3 includes a horizontal plate 34, a slot 33, a hinge 32 and a supporting leg 31; the horizontal plate 34 is used for overall supporting the measuring module and the vision system; the slot 33 can flexibly adjust the relative position of the measuring mechanism on the platform; the hinge 32 is connected with the supporting leg 31 to form a foldable structure, which is convenient for on-site carrying and deployment; the adjustable table has good horizontal adjustment capability, which ensures the stability and flatness of the whole platform.

[0043] The vision detection unit 1 is composed of a bracket 13, an LED group and an industrial camera 11; the bracket 13 provides stable fixation for the vision system; the LED group provides uniform illumination to improve the image clarity; the industrial camera 11 collects the gauge scale image and transmits it to the background through an image acquisition card.

[0044] The inertial measurement unit IMU senses the attitude change (pitch, yaw and roll) of the platform in real time; and the data is corrected in combination with the control module 7.

[0045] The control module 7 outputs control instructions to the gas distribution tool 5 and the image processing unit.

[0046] The gas distribution tool is connected with a plurality of nozzles 4 and is supplied with gas through a gas pump 6; and is used for automatically cleaning the surface of the gauge or performing soft contact assisted positioning.

[0047] The spray head 41 is used for directional gas spraying to remove foreign matters; the connecting pipe 42 connects the gas distribution tool and controls the flow distribution; and the sleeve plate 43 is used for stable installation to improve the stability of the gas spraying direction.

[0048] The above merely describes a preferred specific implementation manner of the present application; however, the protection scope of the present application is not limited to this. Any skilled person in the art, according to the technical scheme and the improved conception of the present application, can make equivalent replacement or change within the technical range disclosed by the present application, which should be covered in the protection scope of the present application.

Claims

1. Gauge block standard based on machine vision and deep learning technology, comprising a measuring mechanism (2), characterized in that: The measuring mechanism (2) is installed on the adjustable table (3), the right side of the measuring mechanism (2) is provided with a visual detection unit (1) and is installed on the adjustable table (3), the adjustable table (3) is provided with an inertial measurement unit (IMU) (8), the front side of the inertial measurement unit (IMU) (8) is provided with a control module (7), the bottom of the adjustable table (3) and the rear side of the control module (7) are provided with a gas distribution tool (5), the gas distribution tool (5) is controlled by a gas pump (6), and the nozzles (4) are installed on the bottom of the two sides of the adjustable table (3).

2. The gauge standard based on machine vision and deep learning techniques according to claim 1, wherein: The adjustable table (3) comprises a horizontal plate (34), the horizontal plate (34) is provided with a slot (33), and the horizontal plate (34) is connected with a supporting leg (31) through a hinge (32) below the horizontal plate (34). 3.The gauge standard based on machine vision and deep learning techniques according to claim 1, wherein: The visual detection unit (1) comprises a support (13), the support (13) is provided with an LED group (12), and the LED group (12) is provided with an industrial camera (11).

4. The gauge standard based on machine vision and deep learning techniques according to claim 1, wherein: The measuring mechanism (2) comprises a gauge (22) installed below a gauge positioning surface (21), a vertical plate (24) is installed on the right side of the gauge (22), the bottom of the vertical plate (24) is provided with a sleeving groove (25), a bottom plate (23) is arranged below the gauge (22), and the gauge (22) and the vertical plate (24) are in the same plane.

5. The gauge standard based on machine vision and deep learning techniques according to claim 4, wherein: The connecting part between the gauge (22) and the vertical plate (24) is a connecting plate (223), the left side of the connecting plate (223) is provided with a limiting cylinder (222) sleeved on the gauge (22), and the right side of the connecting plate (223) is provided with a sleeve frame (224) sleeved on the vertical plate (24). 6.The gauge standard based on machine vision and deep learning techniques of claim 5, wherein: The bottom of the limiting cylinder (222) is coaxially connected with a connecting shaft (212), and the connecting shaft (212) is installed below a pressure sensor (211). 7.The gauge standard based on machine vision and deep learning techniques of claim 1, wherein: The nozzle (4) comprises a sleeve plate (43), the bottom of the sleeve plate (43) is provided with a spray head (41), and the nozzle (4) is provided with a connecting pipe (42) in the middle.

Images