Image acquisition robot for machine learning

Abstract

The utility model discloses an image acquisition robot for machine learning, which comprises a base, a walking assembly is mounted at the bottom of the base, a battery box is fixed at the top of the base, a bearing plate is welded on the outer wall of the top of the battery box, an image acquisition part is mounted at the top of the bearing plate, and a top plate is fixed at the top of the image acquisition part through a connecting piece. A second electric telescopic rod is fixed to the outer wall of the top of the bearing plate, a connecting base is fixed to the output end of the second electric telescopic rod through a pin, and a storage cylinder is fixed to the outer wall of the connecting base. Through the storage cylinder, the first rubber ring at the top of the storage cylinder and the second rubber ring at the bottom of the storage cylinder, the robot can completely store the image collecting part in a non-working state, damage of moisture and dust to equipment is effectively prevented, the second electric telescopic rod drives the storage cylinder to move downwards in the working process, the stabilizing assembly is attached to the ground, and the working efficiency is improved. It is ensured that the robot keeps stable in the image collection process, and collected image blurring caused by vibration or movement is avoided.

Description

TECHNICAL FIELD

[0001] The utility model relates to image acquisition robot technical field especially relates to a kind of image acquisition robots for machine learning. BACKGROUND

[0002] With the rapid development of artificial intelligence (AI) technology, machine learning is increasingly widely applied in various fields, especially in computer vision, autonomous driving, intelligent monitoring and industrial automation, the training and optimization of machine learning model rely on a large amount of high-quality data, especially image data, so efficient and accurate image acquisition equipment has become an indispensable key component in machine learning application.

[0003] According to the search, the patent with Chinese patent application number 202420078428.7 discloses an image acquisition robot for machine learning, which comprises a remote control vehicle body, a hydraulic cylinder, an acquisition mechanism, an embedded slot, a miniature camera and a central control module.

[0004] The image acquisition robot for machine learning in the above patent has the following disadvantages: the robot cannot protect the identification camera when it is in a non-working state, and when the environment is dusty and humid, the humidity and dust will damage the identification camera and affect its service life. UTILITY MODEL CONTENT

[0005] The utility model aims at solving the shortcomings in the prior art and provides an image acquisition robot for machine learning.

[0006] To achieve the above-mentioned purpose, the utility model adopts the following technical scheme:

[0007] An image acquisition robot for machine learning comprises a base, a walking assembly is installed at the bottom of the base, a battery box is fixed at the top of the base, a bearing plate is welded to the top outer wall of the battery box, an image acquisition part is installed at the top of the bearing plate, a top plate is fixed to the top of the image acquisition part through a connecting piece, an electric telescopic rod II is fixed to the top outer wall of the bearing plate, a connecting seat is fixed to the output end of the electric telescopic rod II through a pin, a storage cylinder is fixed to the outer wall of the connecting seat, a rubber ring I for cooperation with the top plate is bonded to the top of the storage cylinder, a pressure plate is integrally formed at the bottom of the storage cylinder, a rubber ring II for cooperation with the bearing plate is bonded to the top of the pressure plate, and a stabilizing assembly is installed on the side wall of the storage cylinder.

[0008] As a further scheme of the utility model: the walking assembly comprises a Meckheim wheel, a driving motor I and a mounting plate, the mounting plate is fixed to the outer wall at the bottom of the base, the driving motor I is fixed to the outer wall on one side of the mounting plate, and the Meckheim wheel is connected to the output end of the driving motor I through a coupling.

[0009] As a further scheme of the utility model: the stable assembly includes connecting plate and support seat, the connecting plate is welded to the lateral wall of the receiving cylinder, and the support seat is fixed to the bottom of the connecting plate.

[0010] As a further scheme of the utility model: the image acquisition part includes electric telescopic rod, rotary motor and fixing frame, the electric telescopic rod is fixed to the top outer wall of the bearing plate, and the fixing frame is fixed to the output end of the electric telescopic rod through a pin, and the rotary motor is fixed to the top inner wall of the fixing frame.

[0011] As a further scheme of the utility model: the rotary motor output end is fixed with rotary seat through a pin, and the rotary seat top is welded with containing plate, and the rotating shaft is movably connected between the two containing plates, and the rotating shaft circumferential outer wall is welded with placing plate, and the image acquisition camera is fixed to the top of the placing plate.

[0012] As a further scheme of the utility model: the containing plate one side outer wall is fixed with drive motor two, and the drive motor two output end is connected to the rotating shaft one end through a shaft coupling.

[0013] As a further scheme of the utility model: the connecting piece is fixed rod, and the fixed rod bottom is fixed to the top outer wall of the rotary seat, and the top plate is welded to the top outer wall of the fixed rod.

[0014] Compared with the prior art, the utility model provides an image acquisition robot for machine learning, which has the following beneficial effects:

[0015] 1. By adopting the Mecanum wheel as the walking assembly, the robot can realize multiple movement modes such as advancing, retreating, left-right translation, oblique movement and in-place rotation in a complex environment, and is particularly suitable for flexible movement in a narrow space.

[0016] 2. By adopting the receiving cylinder and the rubber ring one at the top and the rubber ring two at the bottom, the robot can completely store the image acquisition part in the non-working state, effectively prevents the damage of moisture and dust to the equipment, and drives the receiving cylinder to move down when working, so that the stable assembly is attached to the ground, and the robot can keep stable during image acquisition, thereby avoiding the blurred image caused by vibration or movement.

[0017] 3. By combining the electric telescopic rod, the rotary motor and the drive motor two, the robot can adjust the height and shooting direction of the image acquisition camera, reduce the shooting dead angle, and significantly improve the image acquisition efficiency.

[0018] The parts not involved in the device are the same as or can be realized by the prior art, and the utility model has the advantages of simple structure and convenient operation. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 The utility model provides a kind of overall structure schematic view of image acquisition robot for machine learning;

[0020] Figure 2 The utility model provides a kind of side direction structure schematic view of image acquisition robot for machine learning;

[0021] Figure 3 The utility model provides a kind of protective assembly installation structure schematic view of image acquisition robot for machine learning;

[0022] Figure 4 The utility model provides a kind of protective assembly main body structure schematic view of image acquisition robot for machine learning;

[0023] Figure 5 The utility model provides a kind of image acquisition part structure schematic view of image acquisition robot for machine learning.

[0024] In the drawing: base 1;Accommodation cylinder 2;Mackheim wheel 3;Battery box 4;Driving motor one 5;Connecting plate 6;Supporting seat 7;Mounting plate 8;Top plate 9;Electric telescopic rod 10;Bearing plate 11;Electric telescopic rod two 12;Linking seat 13;Rubber ring one 14;Rubber ring two 15;Driving motor two 16;Image acquisition camera 17;Rotating shaft 18;Fixed rod 19;Accommodation plate 20;Rotary motor 21;Fixing frame 22;Placing plate 23;Rotary seat 24;Pressing plate 25. DETAILED DESCRIPTION

[0025] 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, not all the embodiments.

[0026] An image acquisition robot for machine learning, as shown in Figures 1 to 5 The utility model discloses an image acquisition robot for machine learning, which comprises a base 1, a walking assembly is installed at the bottom of the base 1, a battery box 4 is fixed at the top of the base 1 through bolts, a bearing plate 11 is welded to the outer wall at the top of the battery box 4, an image acquisition part is installed at the top of the bearing plate 11, a top plate 9 is fixed to the top of the image acquisition part through a connecting piece, an electric telescopic rod two 12 is fixed to the outer wall at the top of the bearing plate 11 through bolts, a linking seat 13 is fixed to the output end of the electric telescopic rod two 12 through a pin, an accommodation cylinder 2 is fixed to the outer wall of the linking seat 13 through bolts, a rubber ring one 14 that is used in cooperation with the top plate 9 is bonded to the top of the accommodation cylinder 2, a pressing plate 25 is integrally formed at the bottom of the accommodation cylinder 2, a rubber ring two 15 that is used in cooperation with the bearing plate 11 is bonded to the top of the pressing plate 25, and a stabilizing assembly is installed on the side wall of the accommodation cylinder 2;

[0027] The image acquisition part installed on the bearing plate 11 is used to collect images, and the battery box 4 provides power. In the initial state, the top of the rubber ring 14 and the bottom of the top plate 9 are attached to each other, and the rubber ring 15 and the outer wall of the bottom of the bearing plate 11 are attached to each other. In the non-working state, the image acquisition part is accommodated in the storage cylinder 2, realizing the moisture-proof and dust-proof protection of the image acquisition part. When image acquisition is needed, the lower part of the storage cylinder 2 is driven by the second electric telescopic rod 12 to move downward, and the image acquisition part is exposed during the downward movement of the storage cylinder 2, so that the image acquisition part is used for image acquisition. At the same time, the stable assembly installed on the side wall of the storage cylinder 2 moves downward, and when the stable assembly moves downward to be attached to the ground, the stable assembly plays a role in stabilizing the whole robot.

[0028] The walking assembly comprises a McKibben wheel 3, a driving motor 5 and a mounting plate 8. The mounting plate 8 is fixed to the outer wall of the bottom of the base 1 by bolts, and the driving motor 5 is fixed to the outer wall of one side of the mounting plate 8 by bolts. The McKibben wheel 3 is connected to the output end of the driving motor 5 through a shaft coupling.

[0029] The driving motor 5 can drive the McKibben wheel 3 to rotate. The McKibben wheel 3 can realize forward movement, backward movement, left-right translation, oblique movement and rotation in place without changing its own direction. This feature enables the robot to move flexibly in complex environments.

[0030] The stable assembly comprises a connecting plate 6 and a support seat 7. The connecting plate 6 is welded to the side wall of the storage cylinder 2, and the support seat 7 is fixed to the bottom of the connecting plate 6 by bolts.

[0031] When the storage cylinder 2 is driven to move downward by the second electric telescopic rod 12, the support seat 7 moves downward synchronously and is attached to the ground. The support seat 7 plays a role in stabilizing the whole robot.

[0032] The image acquisition part comprises an electric telescopic rod 10, a rotating motor 21 and a fixing frame 22. The electric telescopic rod 10 is fixed to the outer wall of the top of the bearing plate 11 by bolts. The fixing frame 22 is fixed to the output end of the electric telescopic rod 10 by a pin. The rotating motor 21 is fixed to the inner wall of the top of the fixing frame 22 by bolts. The output end of the rotating motor 21 is fixed with a rotating seat 24 by a pin. The top of the rotating seat 24 is welded with a containing plate 20. A rotating shaft 18 is rotatably connected between two containing plates 20. The circumferential outer wall of the rotating shaft 18 is welded with a placing plate 23. The top of the placing plate 23 is fixed with an image acquisition camera 17 by bolts. The outer wall of one side of the containing plate 20 is fixed with a driving motor 2 by bolts. The output end of the driving motor 2 is connected to one end of the rotating shaft 18 through a shaft coupling.

[0033] The height of the image acquisition camera 17 can be adjusted by the electric telescopic rod 10, the rotating motor 21 can indirectly drive the image acquisition camera 17 to rotate, thereby adjusting the shooting direction of the image acquisition camera 17 in the transverse direction, the driving motor two 16 can drive the rotating shaft 18 and the placement plate 23 to rotate, thereby adjusting the shooting direction of the image acquisition camera 17 in the longitudinal direction, so as to reduce the shooting dead angle of the image acquisition camera 17 and improve the image acquisition efficiency;

[0034] When the image acquisition camera 17 is used for machine learning in the embodiment, the image acquisition camera 17 is preferably battery-powered. The specific principle is as follows: the image acquisition camera 17 collects image data for training a machine learning model, such as object recognition, scene classification, target detection and the like. The collected data can be used to generate a training data set for supervised learning by manual or automatic labeling tools. An edge computing device (such as a GPU or an AI chip) can be carried in the image acquisition camera 17. When collecting images, real-time processing is performed, a pre-trained deep learning model (such as a CNN) is used for object recognition, face recognition or scene analysis, and the behavior is adjusted (such as obstacle avoidance and path planning) according to the recognition result. Through image acquisition and environmental information, combined with SLAM algorithm, different objects and environmental types are identified through image classification and scene understanding, and the decision-making system is assisted to adapt to different working environments.

[0035] The connecting piece is a fixed rod 19, the bottom of the fixed rod 19 is fixed to the top outer wall of the rotating seat 24 by bolts, and the top plate 9 is welded to the top outer wall of the fixed rod 19;

[0036] The fixed rod 19 can be used to fix the top plate 9, so that after the image acquisition is completed, the top plate 9 cooperates with the storage cylinder 2 and the bearing plate 11 to protect the image acquisition part from dust and moisture.

[0037] Working principle: when the robot is in a static state, the image acquisition part is stored inside the storage cylinder 2, the top of the rubber ring 14 is tightly attached to the bottom of the top plate 9, and the rubber ring 15 is tightly attached to the outer wall of the bottom of the bearing plate 11, effectively preventing moisture and dust from entering the protection image acquisition part, when the robot needs to be moved, the driving motor 5 is started, the coupling drives the micromechanical wheel 3 to rotate, the micromechanical wheel 3 realizes forward, backward, left and right translation, oblique movement and rotation in place and other movement modes, and adapts to complex environment, when reaching the specified position and needing to collect images, the electric telescopic rod 12 is started, the driving adapter 13 and the storage cylinder 2 are driven to move downward, the image acquisition part gradually exposes and prepares to collect images, at the same time, the stable assembly moves downward with the storage cylinder 2, the support seat 7 contacts the ground and provides stable support, ensuring that the robot remains stable during image acquisition, the electric telescopic rod 10 is started to adjust the height of the image acquisition camera 17 according to the requirement, to obtain image information at different heights, the rotary motor 21 is started to indirectly drive the image acquisition camera 17 to rotate in the transverse direction through the rotating seat 24 to adjust the shooting direction, the driving motor 16 is started to drive the rotating shaft 18 and the placing plate 23 to rotate, further adjusting the shooting direction of the image acquisition camera 17 in the longitudinal direction, through the adjustment of the above three dimensions, the image acquisition camera 17 can flexibly capture the image of the target area, reducing the dead angle and improving the image acquisition efficiency;

[0038] In the embodiment, the robot performs various operations by using a remote control mode, such as using a remote controller, a computer terminal, etc., for sending control instructions to the robot, using a transmission channel to realize the communication connection between the remote operation device and the robot, the transmission channel can be wireless (such as Wi-Fi, mobile communication network), using the controller (not shown) on the robot to receive the control instructions from the remote operation device and convert them into executable action instructions, such as the movement of the walking assembly, the image acquisition of the image acquisition camera 17 and the adjustment of the image acquisition dimension, and the movement of the storage cylinder 2 driven by the electric telescopic rod 12, the remote control belongs to the prior art, which will not be described in detail here.

[0039] The above is only a preferred specific embodiment of the present application, but the protection scope of the present application is not limited thereto, any skilled person in the art can make equivalent replacement or change according to the technical scheme and the inventive concept of the present application within the technical range disclosed by the present application, which should be covered within the protection scope of the present application.

Claims

1. An image acquisition robot for machine learning comprising a base (1), characterized in that, The bottom of the base (1) is provided with a walking assembly, the top of the base (1) is fixedly provided with a battery box (4), the top outer wall of the battery box (4) is welded with a bearing plate (11), the top of the bearing plate (11) is provided with an image acquisition part, the top of the image acquisition part is fixedly provided with a top plate (9) through a connecting piece, the top outer wall of the bearing plate (11) is fixedly provided with an electric telescopic rod (12), the output end of the electric telescopic rod (12) is fixedly provided with a connecting seat (13) through a pin, the outer wall of the connecting seat (13) is fixedly provided with a receiving cylinder (2), the top of the receiving cylinder (2) is bonded with a rubber ring (14) used in cooperation with the top plate (9), the bottom of the receiving cylinder (2) is integrally formed with a pressure plate (25), the top of the pressure plate (25) is bonded with a rubber ring (15) used in cooperation with the bearing plate (11), and a stabilizing assembly is arranged on the side wall of the receiving cylinder (2).

2. The image acquisition robot for machine learning according to claim 1, wherein, The walking assembly comprises a Meckheim wheel (3), a driving motor (5) and a mounting plate (8), the mounting plate (8) is fixed to the bottom outer wall of the base (1), the driving motor (5) is fixed to the outer wall of one side of the mounting plate (8), and the Meckheim wheel (3) is connected to the output end of the driving motor (5) through a shaft coupling.

3. The image acquisition robot for machine learning of claim 2, wherein, The stabilizing assembly comprises a connecting plate (6) and a supporting seat (7), the connecting plate (6) is welded to the side wall of the receiving cylinder (2), and the supporting seat (7) is fixed to the bottom of the connecting plate (6).

4. The image acquisition robot for machine learning of claim 3, wherein, The image acquisition part comprises an electric telescopic rod (10), a rotary motor (21) and a fixing frame (22), the electric telescopic rod (10) is fixed to the top outer wall of the bearing plate (11), the fixing frame (22) is fixed to the output end of the electric telescopic rod (10) through a pin, and the rotary motor (21) is fixed to the top inner wall of the fixing frame (22).

5. The image acquisition robot for machine learning of claim 4, wherein, The output end of the rotary motor (21) is fixedly provided with a rotating seat (24), the top of the rotating seat (24) is welded with a containing plate (20), two containing plates (20) are movably connected with a rotating shaft (18) between them, the circumferential outer wall of the rotating shaft (18) is welded with a placing plate (23), and the top of the placing plate (23) is fixedly provided with an image acquisition camera (17).

6. The image acquisition robot for machine learning of claim 5, wherein, The outer wall of one side of the containing plate (20) is fixedly provided with a driving motor (16), and the output end of the driving motor (16) is connected to one end of the rotating shaft (18) through a shaft coupling.

7. The image acquisition robot for machine learning of claim 6, wherein, The connecting piece is a fixed rod (19), the bottom of the fixed rod (19) is fixed to the top outer wall of the rotating seat (24), and the top plate (9) is welded to the top outer wall of the fixed rod (19).

Images