A 3D information detection device

Abstract

The utility model belongs to three -dimensional sensing and machine vision technical field discloses a kind of 3D information detection equipment, including equipment shell;Composite optical projection module, composite optical projection module is integrated in equipment shell front end, and composite optical projection module includes line laser emitter and micro face array pattern projector;Synchronous imaging module, synchronous imaging module includes at least two side-by-side arrangement image sensors, and is constructed stereoscopic camera pair, and the lens of each image sensor is all provided with composite optical filter;Core processing module, core processing module is set in equipment shell, and core processing module is electrically connected with composite optical projection module and synchronous imaging module respectively.The utility model is by fusing anti-interference strong, precision high line laser scanning and point cloud density high, speed fast face array structure light, to solve the problem that traditional structure light fails under strong light, binocular vision fails in weak texture area matching, can realize full scene high adaptability detection simultaneously.

Description

Technical Field

[0001] This utility model relates to the field of three-dimensional sensing and machine vision technology, and in particular to a 3D information detection device. Background Technology

[0002] With the rapid development of intelligent manufacturing, autonomous driving, AR / VR and robot navigation, the demand for real-time and accurate detection of information such as the three-dimensional shape, position and posture of objects is becoming increasingly urgent. Existing 3D detection technologies are mainly divided into the following categories: 1) Structured light technology: By projecting coded gratings or dot matrix patterns onto the surface of an object, the depth is calculated by capturing the pattern deformation with a camera. However, it is easily affected by ambient light interference, and the effect is not good in strong light or outdoor scenes. In addition, the system calibration is complicated. 2) Binocular stereo vision: mimicking the human eye, it captures images from different perspectives using two cameras and calculates depth using parallax. Its accuracy depends on the accuracy of feature matching. It is prone to matching failure in areas with missing textures, repetitive textures, or low lighting, and requires a large amount of real-time computation.

[0003] 3) Time-of-flight method: Depth is obtained directly by measuring the round-trip time of the light pulse. It has strong anti-interference ability, but it is usually expensive, has relatively low resolution, and suffers from optical crosstalk between multiple devices.

[0004] In existing technologies, single 3D sensing solutions often present contradictions in terms of accuracy, speed, anti-interference capability, or cost. Therefore, we propose a 3D information detection device. Utility Model Content

[0005] In view of the problems existing in the above-mentioned background technology of 3D information detection, this utility model is proposed.

[0006] To solve the above-mentioned technical problems, this utility model provides the following technical solution: A 3D information detection device, including a device housing; A composite optical projection module is integrated into the front end of the device housing. The composite optical projection module includes a line laser emitter for emitting a line laser and a miniature area array pattern projector for projecting a coded area array pattern. A synchronous imaging module, comprising at least two image sensors arranged side by side to form a stereo camera pair, wherein each image sensor has an electrically switchable composite optical filter in front of its lens; The core processing module is disposed inside the device housing and is electrically connected to the composite optical projection module and the synchronous imaging module, respectively.

[0007] As a technical solution of the 3D information detection device of this utility model, the line laser emitter and the micro-area pattern projector are set parallel to each other or at a fixed small angle and share the same optical projection window.

[0008] As a technical solution of the 3D information detection device of this utility model, the composite optical filter has two working states: the first state is a narrowband filtering state, in which the central transmission wavelength matches the wavelength of the line laser emitter; the second state is a broadband light-passing state, which allows light within the spectral range of the micro-area pattern projector to pass through.

[0009] As a technical solution of the 3D information detection device of this utility model, the composite optical filter is a liquid crystal tunable filter.

[0010] As a technical solution of the 3D information detection device of this utility model, the core processing module is configured to control the line laser emitter and the micro area array pattern projector to work alternately in a time-sharing manner, and synchronously control the composite optical filter to switch between the two working states.

[0011] As a technical solution of the 3D information detection device of this utility model, wherein: when the line laser emitter is working, the composite optical filter is in the narrow band filtering state; When the micro-area pattern projector is working, the composite optical filter is in the broadband light-passing state, and at the same time, it processes the laser line image acquired when the line laser emitter is working and the coded pattern image acquired when the micro-area pattern projector is working, and generates the first three-dimensional point cloud data and the second three-dimensional point cloud data respectively.

[0012] As a technical solution of the 3D information detection device of this utility model, it further includes an ambient light sensor installed on the housing of the device. The ambient light sensor is electrically connected to the core processing module. The core processing module can adaptively adjust the working time ratio or output power of the line laser emitter and the micro area array pattern projector according to the signal of the ambient light sensor.

[0013] Compared with the prior art, the present invention has at least the following beneficial effects: 1. This utility model solves the problems of traditional structured light failing under strong light and binocular vision failing to match in weak texture areas by integrating line laser scanning with strong anti-interference and high precision with area array structured light with high point cloud density and high speed. At the same time, it can achieve high adaptability detection in all scenes.

[0014] 2. This utility model avoids spectral interference by controlling the synchronous switching of the projector and filter states in a time-division manner. Combined with adaptive adjustment of ambient light, it can dynamically allocate the working ratio of laser and area array modes, achieving a balance between energy efficiency and performance.

[0015] 2. This utility model, through the integrated design of a shared optical window, composite optical filter and core processing module, can reduce hardware complexity and manufacturing cost, and at the same time, through dual point cloud data fusion, it provides accuracy and detail reproduction capabilities that surpass those of a single solution. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of this utility model, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this utility model. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Among them: Figure 1 This is a schematic diagram of the main structure of this utility model.

[0017] Figure 2 This is a schematic diagram of the separate structure of the miniature area array pattern projector and the line laser emitter of this utility model.

[0018] Figure 3 This is a schematic diagram of the control flow of the core processing module of this utility model.

[0019] Explanation of reference numerals in the attached figures: In the diagram: 1. Equipment casing; 2. Composite optical projection module; 201. Line laser emitter; 202. Miniature area array pattern projector; 3. Synchronous imaging module; 301. Image sensor; 302. Composite optical filter; 4. Core processing module; 5. Ambient light sensor. Detailed Implementation

[0020] To make the above-mentioned objectives, features and advantages of this utility model more apparent and understandable, the specific embodiments of this utility model will be described in detail below with reference to the accompanying drawings.

[0021] Reference Figures 1-3 A 3D information detection device is provided, which includes a device housing 1; The composite optical projection module 2 is integrated at the front end of the device housing 1. The composite optical projection module 2 includes a line laser emitter 201 (785nm wavelength) for emitting a line laser and a micro area array pattern projector 202 (DLP technology, 850nm infrared light) for projecting a coded area array pattern. The synchronous imaging module 3 includes at least two image sensors 301 arranged side by side (such as dual CMOS image sensors 301, global shutter, resolution 1280×1024) and form a stereo camera pair. Each image sensor 301 has an electrically controllable and switchable composite optical filter 302 (such as LCTF) in front of its lens. The core processing module 4 is located inside the device housing 1 and is electrically connected to the composite optical projection module 2 and the synchronous imaging module 3. In application, by integrating the line laser emitter 201 and the micro area array pattern projector 202 into the same front end of the device, hardware miniaturization is achieved, reducing system complexity. The advantages of line laser emitter 201 scanning (high precision) and area array structured light (high density point cloud) are combined to adapt to different detection scenario requirements. At the same time, stereo vision is constructed through dual image sensors 301 to improve the accuracy of depth calculation.

[0022] Reference Figure 1 and Figure 2 The optical axes of the line laser emitter 201 and the micro-area pattern projector 202 are set parallel or at a fixed small angle (e.g., the optical axis angle ≤ 0.5°), and they share the same optical projection window (e.g., a quartz-coated optical window). In application, the parallel or small angle optical axis design ensures the consistency of the field of view of the two projection modes, reduces the difficulty of subsequent data fusion calibration, and the shared optical window simplifies the mechanical structure and avoids light loss and aberration problems introduced by the beam splitter.

[0023] Reference Figures 1-3 The composite optical filter 302 has two operating states. The first state is a narrowband filtering state, in which the center transmission wavelength matches the wavelength of the line laser emitter 201. The second state is a broadband light-passing state, which allows light within the spectral range of the micro-area pattern projector 202 to pass through. The composite optical filter 302 is a liquid crystal tunable filter. For example, in the narrowband state, the center wavelength is 785nm and the bandwidth is ±5nm (suppressing ambient light interference). In the broadband state, the light transmission range is 800-900nm (compatible with the area pattern spectrum). In applications, the narrowband filtering state (matching the laser wavelength) can suppress ambient light interference and improve the laser mode signal-to-noise ratio. The broadband light-passing state ensures complete imaging of the area pattern. At the same time, the liquid crystal tunable filter has a fast electronic switching speed (millisecond level), no mechanical wear, long life and precise response.

[0024] Reference Figures 1-3 The core processing module 4 is configured to control the line laser emitter 201 and the micro array pattern projector 202 to work alternately in a time-sharing manner, and synchronously control the composite optical filter 302 to switch between two working states. When the line laser emitter 201 is working, the composite optical filter 302 is in a narrowband filtering state. When the miniature area array pattern projector 202 is working, the composite optical filter 302 is in a broadband light-passing state. Simultaneously, it processes the laser line images acquired when the online laser emitter 201 is working and the coded pattern images acquired when the miniature area array pattern projector 202 is working, respectively generating first and second 3D point cloud data. For example, the core processing module 4 is equipped with a dual-processor FPGA and ARM. The FPGA is responsible for real-time control of the projector / filter switching (timing accuracy ≤1ms), and the ARM runs 3D reconstruction algorithms (such as line laser triangulation and area array structured light phase calculation). Laser mode processing involves extracting the sub-pixel center points of the laser lines, and the stereo camera matches corresponding points. Finally, triangulation generates a high-precision point cloud (accuracy ±0.05mm); the area array mode is processed by decoding Gray code patterns, calculating the phase difference between the left and right images, and finally generating a high-density point cloud (point spacing 0.1mm); point cloud fusion is based on ICP algorithm registration of dual point clouds, weighted fusion (laser point cloud weight 0.7, area array point cloud weight 0.3), and finally outputting a complete 3D model. In application, the time-division control strategy layout can avoid spectral crosstalk between laser and area array patterns, ensuring that the two modes operate independently and efficiently. Combined with dual point cloud fusion, it can generate high-precision linear laser point clouds (suitable for edges and contours) and high-density area array point clouds (suitable for curved surface details), which can complement each other to improve the quality of 3D reconstruction.

[0025] Reference Figures 1-3 It also includes an ambient light sensor 5 installed on the device housing 1. The ambient light sensor 5 is electrically connected to the core processing module 4. The core processing module 4 can adaptively adjust the working time ratio or output power of the line laser emitter 201 and the micro area array pattern projector 202 according to the signal of the ambient light sensor 5. For example, if the ambient light sensor 5 has a monitoring range of 0-100klux, the dynamic adjustment strategy is to increase the proportion of laser mode to 70% in strong light environment (>50klux) and the proportion of area array mode to 80% in weak light environment (<10lux). In application, the working mode can be dynamically adjusted according to the ambient light intensity (such as prioritizing laser mode under strong light), which can significantly improve the robustness of outdoor / complex lighting scenarios.

[0026] The working principle of this utility model is as follows: Equipment initialization stage: First, power on and start, the core processing module 4 performs self-test (such as the status of the miniature area array pattern projector 202, sensor image sensor 301, and composite optical filter 302), then load preset calibration parameters (such as camera intrinsic parameters, extrinsic parameters, and distortion coefficient), and finally calibrate the ambient light sensor 5 (aligning with a standard whiteboard for light measurement). Work mode selection stage: Automatic mode (default): Ambient light sensor 5 provides real-time feedback on light intensity. At this time, the core processing module 4 dynamically adjusts the working time ratio of laser / area array mode (e.g., laser accounts for 70% under strong light). Manual mode: Users can set a fixed mode ratio via the host computer (e.g., enable only laser scanning). Data acquisition and reconstruction phase: First, the device is aligned with the target object (working distance 0.5-3m), and then the acquisition command is triggered. At this time, laser / area array patterns are projected alternately in sequence, and the imaging module 3 captures the image synchronously. During this period, dual point cloud data is generated and fused in real time, and a three-dimensional mesh model is output (e.g., format: .obj or .ply). Results optimization and application stage: such as industrial inspection, comparing point clouds with CAD models, and marking dimensional deviations (such as hole position errors in sheet metal parts). For example, in robot navigation, the outlines of obstacles are extracted from point clouds to generate obstacle avoidance paths; In AR / VR, point clouds are transmitted to the rendering engine in real time to drive virtual object interaction.

[0027] It should be noted that the above embodiments are only used to illustrate the technical solution of this utility model and are not intended to limit it. Although this utility model has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solution of this utility model without departing from the spirit and scope of the technical solution of this utility model, and all such modifications or substitutions should be covered within the scope of the claims of this utility model.

Claims

1. A 3D information detection device, characterized in that: include: Equipment casing (1); The composite optical projection module (2) is integrated at the front end of the device housing (1). The composite optical projection module (2) includes a line laser emitter (201) for emitting a line laser and a micro array pattern projector (202) for projecting a coded array pattern. Synchronous imaging module (3), the synchronous imaging module (3) includes at least two image sensors (301) arranged side by side, forming a stereo camera pair, and each image sensor (301) has an electrically controllable and switchable composite optical filter (302) in front of its lens. The core processing module (4) is disposed inside the device housing (1) and is electrically connected to the composite optical projection module (2) and the synchronous imaging module (3) respectively.

2. The 3D information detection device according to claim 1, characterized in that: The line laser emitter (201) and the micro-area pattern projector (202) are set parallel to each other or at a fixed small angle, and share the same optical projection window.

3. The 3D information detection device according to claim 1, characterized in that: The composite optical filter (302) has two working states. The first state is a narrowband filter state, in which the central transmission wavelength matches the wavelength of the line laser emitter (201). The second state is a broadband light-passing state, which allows light within the spectral range of the micro-area pattern projector (202) to pass through.

4. The 3D information detection device according to claim 3, characterized in that: The composite optical filter (302) is a liquid crystal tunable filter.

5. The 3D information detection device according to claim 3, characterized in that: The core processing module (4) is configured to control the line laser emitter (201) and the micro-area pattern projector (202) to work alternately in a time-sharing manner, and to synchronously control the composite optical filter (302) to switch between the two working states.

6. The 3D information detection device according to claim 5, characterized in that: When the line laser emitter (201) is working, the composite optical filter (302) is in the narrowband filtering state; When the micro-area pattern projector (202) is working, the composite optical filter (302) is in the broadband light-passing state, and at the same time, it processes the laser line image acquired when the line laser emitter (201) is working and the coded pattern image acquired when the micro-area pattern projector (202) is working, and generates the first three-dimensional point cloud data and the second three-dimensional point cloud data respectively.

7. The 3D information detection device according to any one of claims 1-6, characterized in that: It also includes an ambient light sensor (5) installed on the housing (1) of the device. The ambient light sensor (5) is electrically connected to the core processing module (4). The core processing module (4) can adaptively adjust the working time ratio or output power of the line laser emitter (201) and the micro area array pattern projector (202) according to the signal of the ambient light sensor (5).

Images