Design method of active and passive fusion camera for space target measurement

Abstract

The invention discloses an active and passive fusion camera design method for space target measurement, which comprises the steps of active and passive fusion measurement camera model selection, active and passive fusion camera combined structure design, active and passive fusion camera calibration design, binocular camera selection of a passive measurement sensor, and ToF module selection of an active measurement sensor. Through fusion of the binocular camera and the ToF module, the system has the comprehensive advantages of complex illumination, dynamic target, depth information acquisition, clear and stable imaging, low power consumption and simple structure, the active and passive fusion camera is designed to be a far / near field active and passive fusion measurement system, the far field is responsible for 25m-9m target measurement, the near field is responsible for 11m-0.5 m target measurement, and the near field is responsible for 25m-9m target measurement. The design result shows that the designed active and passive fusion measurement camera can be used for space target pose measurement within the range of 25m-2m.

Description

Technical Field

[0001] The invention belongs to the field of space target measurement, and in particular relates to a design method of an active and passive fusion camera for space target measurement. Background Art

[0002] When space on-orbit service and other control tasks are carried out, in order to accurately capture the target, it is necessary to accurately measure the space target. According to whether active fill light is required during the measurement process, it is divided into two modes: passive imaging and active imaging. Binocular camera measurement is a passive measurement. The detector of this technical approach is relatively mature and has a high resolution. However, it is limited by the field of view angle, measurement baseline, sunlight incidence angle and observation angle. It has the disadvantages of limited measurement range, poor robustness, strong dependence on space ambient light and poor real-time algorithm. ToF camera measurement is an active measurement. Based on the time of flight method, the distance information of the target can be directly obtained. Point cloud data can be further obtained through internal reference calibration. Compared with binocular stereo vision, its measurement range is wider, but the pixel size of the detector system is larger and the resolution is lower.

[0003] In summary, this patent proposes a technical approach to integrate a ToF module with a binocular camera to overcome the shortcomings of traditional visual measurement methods and existing TOF measurement technologies. In order to make the measurement range reach 25m, two sets of active and passive fusion measurement systems for far / near fields of view will be designed, of which the far field of view is responsible for 25m~9m and the near field is responsible for 11m~0.5m. Summary of the invention

[0004] The object of the present invention is to provide a design method of an active-passive fusion camera for space target measurement.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A design method of an active-passive fusion camera for space target measurement is characterized by comprising the following steps:

[0007] Step S1, selection of active and passive fusion measurement camera;

[0008] Step S2, active and passive fusion camera combination structure design;

[0009] Step S3: active and passive fusion camera calibration design.

[0010] Preferably, the step S1 comprises the following steps:

[0011] S1.1, the active and passive fusion measurement camera is composed of a visible light module, a TOF module, a power module, an information processing module, and a connector;

[0012] S1.2, the visible light module includes an optical lens, a light shield, and an interface circuit;

[0013] S1.3, the TOF module includes an optical lens, a light shield, a laser, and an interface circuit;

[0014] S1.4, the power module outputs two voltages, 5V and 24V, of which the visible light module is powered by 5V and the ToF module is powered by 24V;

[0015] S1.5, the information processing module includes an external interface module and embedded development;

[0016] S1.6, the near-field ToF module within the range of 11m to 0.5m is selected as the blaze-101-TOF camera, with a power supply voltage of 24V; the near-field monocular camera within the range of 11m to 0.5m is selected as the a2A2840-48umBAS camera, with a power supply voltage of 5V;

[0017] S1.7, the far-field ToF module in the range of 25m to 9m is selected as the XT-S240Pro-TOF camera, and the power supply voltage is designed to be 24V. The far-field monocular camera in the range of 25m to 9m is selected as the a2A2448-75umBAS camera, and the power supply voltage is designed to be 5V.

[0018] S1.8, information processing module design, the CPU selects quad-core Cortex-A76@2.42Hz + quad-core Cortex-A55@1.8Hz, the GPU selects Mail-G610MP4, the computing power selects 6TOPS, the RAM selects 8GBLPDDR4, the ROM selects 64GBeMMC, and the external interfaces include 3 USB3.1 interfaces, 2 USB2.0 interfaces, 4 PCIe3.0 interfaces, 3 PCIe2.0 interfaces, 2 Gigabit Ethernet ports, 1 HDMI input interface, 2 HDMI output interfaces, and 2 MIPI interfaces.

[0019] Preferably, the step S2 comprises the following steps:

[0020] S2.1, the near-field ToF module and the far-field ToF module are connected to the development board through the network port, and the two monocular cameras are connected to the development board through the USB3.1 port;

[0021] S2.2, the near-field ToF module and the far-field TOF module are used to acquire depth information of the target respectively. The far-field TOF module acquires the depth of the target in the range of 25m to 9m, and the near-field TOF module acquires the depth of the target in the range of 11m to 0.5m;

[0022] S2.3, the field of view of the far-field TOF camera is designed to be 72°×58°, the resolution is 320×240 pixels, the volume is designed to be 65mm×65mm*×41mm, 4 LEDs are evenly arranged around the lens, and the light wavelength of the LED is selected to be 940nm;

[0023] S2.4, the field of view of the far-field TOF camera is designed to be 67°×51°, the resolution is 640×480 pixels, the volume is designed to be 99.6mm×80.6mm×63.1mm, and the four LEDs are evenly arranged around the lens. The wavelength of the LED light is selected to be 940nm;

[0024] S2.5, the long-field monocular camera lens is designed as ComputarLensM1614-MP2F1.4f16mm2 / 3", the resolution is 2448×2048pixel, the pixel size is 2.74μm×2.74μm, the frame rate is 75Fps, the interface is USB3.1, and the volume is 42.8mm×29mm×29mm;

[0025] S2.6, the near-field monocular camera lens is designed as ComputarLensM1614-MP2F1.4f8mm2 / 3", the resolution is 2448×2048pixel, the pixel size is 2.74μm×2.74μm, the frame rate is 48Fps, the interface is USB3.1, and the volume is 42.8mm×29mm×29mm;

[0026] S2.6, when the far-field monocular camera is at 25m, the field of view is designed to be 10.5m×8.4m, and when the far-field monocular camera is at 10m, the field of view is designed to be 4.1m×3.5m;

[0027] S2.7, when the near-field monocular camera is at 10m, the field of view is designed to be 9.7m×9.7m, and when the near-field monocular camera is at 0.5m, the field of view is designed to be 0.48m×0.48m;

[0028] S2.8. At 10m where the far / near field of view monocular cameras are switched, both monocular cameras can image a target of 1m×1m×0.2m.

[0029] Preferably, the step S3 comprises the following steps:

[0030] S3.1. Calibration of near-field monocular camera and near-field TOF module. Place the checkerboard calibration plate within the field of view of the near-field TOF module and the near-field monocular camera, and its image is clear. Use the near-field monocular camera and the near-field TOF module to collect 100 images of different angles and positions, respectively, to ensure that the calibration plate has good coverage in the image. For each image, use the ORB corner point extraction algorithm to detect the corner points on the calibration plate. The corner points are defined as the intersection points of the checkerboard. For the near-field monocular camera and the near-field TOF module, use the image and the corresponding corner points to calculate the internal parameters K of the near-field monocular camera and the near-field TOF module. 近-单 With K 近-ToF (such as focal length, principal point position) and external parameters (such as the rotation and translation matrix of the camera), where Represents the external parameters from the near-field monocular camera to the near-field TOF module;

[0031] S3.2. Calibration of the far-field monocular camera and the far-field TOF module. Place the checkerboard calibration plate within the field of view of the far-field TOF module and the far-field monocular camera, and its image is clear. Use the far-field monocular camera and the far-field TOF module to collect 200 images of different angles and positions respectively to ensure that the calibration plate has good coverage in the image. For each image, use the ORB corner point extraction algorithm to detect the corner points on the calibration plate. The corner points are defined as the intersection points of the checkerboard. For the near-field monocular camera and the far-field TOF module, use the image and the corresponding corner points to calculate the internal parameters K of the near-field monocular camera and the near-field TOF module. 远-单 With K 远-ToF (such as focal length, principal point position) and external parameters (such as the rotation and translation matrix of the camera), where Represents the external parameters from the far-field monocular camera to the far-field TOF module;

[0032] S3.3. Calibration of near-field monocular camera and far-field monocular camera. Place the checkerboard calibration plate within the field of view of the near-field monocular camera and the far-field monocular camera, and its image is clear. Use the near-field monocular camera and the far-field monocular camera to collect 25 images of different angles and positions respectively to ensure that the calibration plate has good coverage in the image. For each image, use the ORB corner point extraction algorithm to detect the corner points on the calibration plate. The corner points are defined as the intersection points of the checkerboard. For the near-field monocular camera and the far-field monocular camera, use the image and the corresponding corner points to calculate the camera's internal parameters K 近-单 With K 远-单 (such as focal length, principal point position) and external parameters (such as the rotation and translation matrix of the camera), where Represents the external parameters from the near-field monocular camera to the far-field monocular camera;

[0033] S3.4, Calibration of near-field TOF module and far-field TOF module, make a disk with a diameter of 300mm, the disk is made of white reflective material, and the center point of the disk is used to locate the near-field TOF simulation and the far-field TOF module. Find the disk within the public field of view, use the center of the disk as the common point for calibration calculation, and then perform rigid transformation to obtain the calibration parameter matrix between the two cameras Represents the external parameters from the near-field ToF module to the far-field ToF module;

[0034] S3.5. Calculate the external parameters of the near-field monocular camera and the far-field TOF module Calculate the external parameters of the far-field monocular camera and the near-field TOF module in

[0035] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0036] 1. The active and passive fusion of ToF and binocular cameras can overcome the shortcomings of traditional binocular cameras and existing TOF measurements, and has the comprehensive advantages of complex lighting, long-distance imaging, and depth information acquisition. The far / near field of view switching method can adapt to various types of targets in space, ensure that the target is always in the field of view, and alleviate the problem of limited field of view angle. The connection method of binocular cameras, TOF modules and embedded development boards has the advantages of simple structure and easy replacement. The different field of view angle settings of far / near field of view TOF modules can capture targets in the field of view faster, and the distortion of the image edge is less.

[0037] 2. The TOF module and the embedded development board are connected by the network port, and the data transmission is stable. The three calibrations of two monocular cameras and two TOF modules have the advantages of simple calibration, robust results, and easy implementation, which reduces the complexity of the traditional rail calibration scheme. The switching distance design of the two monocular cameras and two TOF modules can cope with the image tailing problem caused by the rolling of space targets. The layout design of the two monocular cameras and two TOF modules has the advantages of compact structure and non-interference. The two monocular cameras communicate through USB3.1, and the two TOF modules communicate through the network port, which has the advantage of self-adaptation, reduces the constraints on the communication bandwidth, and improves the efficiency of on-orbit applications. The combination of two monocular cameras and two TOF modules can form a variety of redundant schemes, such as far-field monocular + far-field ToF, near-field monocular + near-field ToF, far-field monocular + near-field monocular, far-field ToF + near-field ToF, which can effectively deal with the failure of a certain type of camera and obtain multi-depth and multi-depth data. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0039] Figure 1 A flow chart of a design method of an active-passive fusion camera for space target measurement provided by the present invention;

[0040] Figure 2 A connection diagram between the active and passive fusion camera and the development board provided by the present invention;

[0041] Figure 3 The installation relationship of the active and passive fusion camera provided by the present invention. DETAILED DESCRIPTION

[0042] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] Example 1

[0044] See also Figure 1-Figure 3 As shown, the present invention provides a design method of an active-passive fusion camera for space target measurement, comprising the following steps:

[0045] Step S1, selection of active and passive fusion measurement camera;

[0046] Step S2, active and passive fusion camera combination structure design;

[0047] Step S3: active and passive fusion camera calibration design.

[0048] The step S1 comprises the following steps:

[0049] S1.1, the active and passive fusion measurement camera is composed of a visible light module, a TOF module, a power supply and information processing module, connectors, etc.;

[0050] S1.2, the visible light module includes an optical lens, a light shield, an interface circuit, and an FPGA module;

[0051] S1.3, TOF module includes optical lens, light shield, laser, interface circuit, ToF module embedded development board;

[0052] S1.4, power supply and information processing module includes power supply module, external interface module and core processor module;

[0053] S1.5, the near-field ToF module in the range of 11m to 0.5m is the blaze-101-TOF camera, with a power supply voltage of 24V, and the near-field monocular camera is the a2A2840-48umBAS camera, with a power supply voltage of 5V;

[0054] S1.6, The far-field ToF module in the range of 25m to 9m is selected as the XT-S240Pro-TOF camera, with a power supply voltage of 24V, and the far-field monocular camera is selected as the a2A2448-75umBAS camera, with a power supply voltage of 5V;

[0055] S1.7, camera, light source and development board can be powered by 5V or 24V;

[0056] S1.8, ToF module embedded development board parameter selection, CPU selects quad-core Cortex-A76@2.42Hz+quad-core Cortex-A55@1.8Hz, GPU selects Mail-G610MP4, computing power selects 6TOPS, RAM selects 8GB LPDDR4, ROM selects 64GB eMMC, development board interface, including 3 USB3.1 interfaces, 2 USB2.0 interfaces, 4 PCIe3.0 interfaces, 3 PCIe2.0 interfaces, 2 Gigabit Ethernet ports, 1 HDMI input interface, 2 HDMI output interfaces, and 2 MIPI interfaces;

[0057] S1.9, FPGA module parameter design, FPGA model selected as XC7Z035-2FFG676l, embedded development board selected as OK3588-C.

[0058] The step S2 comprises the following steps:

[0059] S2.1, the near-field ToF module and the far-field ToF module are connected to the embedded development board through the network port, and the far-field monocular camera and the near-field monocular camera are connected to the development board through the USB3.1 interface;

[0060] S2.2, the near-field ToF module and the far-field TOF module are used to acquire depth information of the target. The far-field TOF module acquires the depth of the target within the range of 25m to 9m, and the near-field TOF module acquires the depth of the target within the range of 11m to 0.5m;

[0061] S2.3, the field of view of the far-field TOF camera is designed to be 72°×58°, the resolution is 320×240 pixels, the volume is designed to be 65mm×65mm×41mm, and the four LEDs are evenly arranged around the lens. The wavelength of the LED light is selected to be 940nm;

[0062] S2.4, the field of view of the far-field TOF camera is designed to be 67°×51°, the resolution is 640×480 pixels, the volume is designed to be 99.6mm×80.6mm×63.1mm, and the four LEDs are evenly arranged around the lens. The wavelength of the LED light is selected to be 940nm;

[0063] S2.5, the long-field monocular camera lens is designed as ComputarLensM1614-MP2F1.4f16mm2 / 3", the resolution is 2448×2048pixel, the pixel size is 2.74μm×2.74μm, the frame rate is 75Fps, the interface is USB3.1, and the volume is 42.8mm×29mm×29mm;

[0064] S2.6, the near-field monocular camera lens is designed as ComputarLensM1614-MP2F1.4f8mm2 / 3", the resolution is 2448×2048pixel, the pixel size is 2.74μm×2.74μm, the frame rate is 48Fps, the interface is USB3.1, and the volume is 42.8mm×29mm×29mm;

[0065] S2.6, when the far-field monocular camera is at 25m, the field of view is designed to be 10.5m×8.4m, and when the far-field monocular camera is at 10m, the field of view is designed to be 4.1m×3.5m;

[0066] S2.7, When the near-field monocular camera is at 10m, the field of view is designed to be 9.7m×9.7m. When the near-field monocular camera is at 0.5m, the field of view is designed to be 0.48m×0.48m;

[0067] S2.8. At 10m where the far / near field of view monocular cameras are switched, both monocular cameras can image a target of 1m×1m×0.2m.

[0068] The step S3 comprises the following steps:

[0069] S3.1. Calibration of near-field monocular camera and near-field TOF module. Place the checkerboard calibration plate within the field of view of the near-field TOF module and the near-field monocular camera, and its image is clear. Use the near-field monocular camera and the near-field TOF module to collect 100 images of different angles and positions respectively to ensure that the calibration plate has good coverage in the image. For each image, use the ORB corner point extraction algorithm to detect the corner points on the calibration plate. The corner points are defined as the intersection points of the checkerboard. For the near-field monocular camera and the near-field TOF module, use the image and the corresponding corner points to calculate the internal parameters K of the near-field monocular camera and the near-field TOF module. 近-单 With K 近-ToF(such as focal length, principal point position) and external parameters (such as the rotation and translation matrix of the camera), where Represents the external parameters from the near-field monocular camera to the near-field TOF module;

[0070] S3.2. Calibration of the far-field monocular camera and the far-field TOF module. Place the checkerboard calibration plate within the field of view of the far-field TOF module and the far-field monocular camera, and its image is clear. Use the far-field monocular camera and the far-field TOF module to collect 200 images of different angles and positions respectively to ensure that the calibration plate has good coverage in the image. For each image, use the ORB corner point extraction algorithm to detect the corner points on the calibration plate. The corner points are defined as the intersection points of the checkerboard. For the near-field monocular camera and the far-field TOF module, use the image and the corresponding corner points to calculate the internal parameters K of the near-field monocular camera and the near-field TOF module. 远-单 With K 远-ToF (such as focal length, principal point position) and external parameters (such as the rotation and translation matrix of the camera), where Represents the external parameters from the far-field monocular camera to the far-field TOF module;

[0071] S3.3. Calibration of near-field monocular camera and far-field monocular camera. Place the checkerboard calibration plate within the field of view of the near-field monocular camera and the far-field monocular camera, and its image is clear. Use the near-field monocular camera and the far-field monocular camera to collect 25 images of different angles and positions respectively to ensure that the calibration plate has good coverage in the image. For each image, use the ORB corner point extraction algorithm to detect the corner points on the calibration plate. The corner points are defined as the intersection points of the checkerboard. For the near-field monocular camera and the far-field monocular camera, use the image and the corresponding corner points to calculate the camera's internal parameters K 近-单 With K 远-单 (such as focal length, principal point position) and external parameters (such as the rotation and translation matrix of the camera). Represents the external parameters from the near-field monocular camera to the far-field monocular camera;

[0072] S3.4. Calibration of near-field TOF module and far-field TOF module. Make a disc with a diameter of 300 mm. The disc is made of white reflective material. Use the center point of the disc to locate the near-field TOF simulation and the far-field TOF module. Find the disc within the public field of view, use the center of the disc as the common point for calibration calculation, and then perform rigid transformation to obtain the calibration parameter matrix between the two cameras. Represents the external parameters from the near-field ToF module to the far-field ToF module;

[0073] S3.5. Calculate the external parameters of the near-field monocular camera and the far-field TOF module Calculate the external parameters of the far-field monocular camera and the near-field TOF module in

[0074] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A design method for an active-passive fusion camera for space target measurement, characterized in that: The following steps are involved: Step S1, selection of active and passive fusion measurement camera; Step S2, active and passive fusion camera combination structure design; Step S3: active and passive fusion camera calibration design.

2. The active-passive fusion camera design method for space target measurement according to claim 1, characterized in that: The step S1 comprises the following steps: S1.1, the active and passive fusion measurement camera is composed of a visible light module, a TOF module, a power module, an information processing module, and a connector; S1.2, the visible light module includes an optical lens, a light shield, and an interface circuit; S1.3, the TOF module includes an optical lens, a light shield, a laser, and an interface circuit; S1.4, the power module outputs two voltages, 5V and 24V, of which the visible light module is powered by 5V and the ToF module is powered by 24V; S1.5, the information processing module includes an external interface module and embedded development; S1.6, the near-field ToF module within the range of 11m to 0.5m is selected as the blaze-101-TOF camera, with a power supply voltage of 24V; the near-field monocular camera within the range of 11m to 0.5m is selected as the a2A2840-48umBAS camera, with a power supply voltage of 5V; S1.7, the far-field ToF module in the range of 25m to 9m is selected as the XT-S240Pro-TOF camera, and the power supply voltage is designed to be 24V. The far-field monocular camera in the range of 25m to 9m is selected as the a2A2448-75umBAS camera, and the power supply voltage is designed to be 5V. S1.8, information processing module design, CPU chooses quad-core Cortex-A76@2.42Hz+quad-core Cortex-A55@1.8Hz, GPU chooses Mail-G610MP4, computing power chooses 6TOPS, RAM chooses 8GB LPDDR4, and ROM chooses 64GB eMMC. External interfaces include 3 USB3.1 interfaces, 2 USB2.0 interfaces, 4 PCIe3.0 interfaces, 3 PCIe2.0 interfaces, 2 Gigabit Ethernet ports, 1 HDMI input interface, 2 HDMI output interfaces, and 2 MIPI interfaces.

3. The active-passive fusion camera design method for space target measurement according to claim 1, characterized in that: The step S2 comprises the following steps: S2.1, the near-field ToF module and the far-field ToF module are connected to the embedded development board through the network port, and the far-field monocular camera and the near-field monocular camera are connected to the development board through the USB3.1 interface; S2.2, the near-field ToF module and the far-field TOF module are used to acquire depth information of the target. The far-field TOF module acquires the depth of the target within the range of 25m to 9m, and the near-field TOF module acquires the depth of the target within the range of 11m to 0.5m; S2.3, the field of view of the far-field TOF camera is designed to be 72°×58°, the resolution is 320×240 pixels, the volume is designed to be 65mm×65mm×41mm, and the four LEDs are evenly arranged around the lens. The wavelength of the LED light is selected to be 940nm; S2.4, the field of view of the far-field TOF camera is designed to be 67°×51°, the resolution is 640×480 pixels, the volume is designed to be 99.6mm×80.6mm×63.1mm, and the four LEDs are evenly arranged around the lens. The wavelength of the LED light is selected to be 940nm; S2.5, the long-field monocular camera lens is designed as ComputarLensM1614-MP2F1.4f16mm2 / 3", the resolution is 2448×2048pixel, the pixel size is 2.74μm×2.74μm, the frame rate is 75Fps, the interface is USB3.1, and the volume is 42.8mm×29mm×29mm; S2.6, the near-field monocular camera lens is designed as ComputarLensM1614-MP2F1.4f8mm2 / 3", the resolution is 2448×2048pixel, the pixel size is 2.74μm×2.74μm, the frame rate is 48Fps, the interface is USB3.1, and the volume is 42.8mm×29mm×29mm; S2.7, when the far-field monocular camera is at 25m, the field of view is designed to be 10.5m×8.4m, and when the far-field monocular camera is at 10m, the field of view is designed to be 4.1m×3.5m; S2.8, when the near-field monocular camera is at 10m, the field of view is designed to be 9.7m×9.7m, and when the near-field monocular camera is at 0.5m, the field of view is designed to be 0.48m×0.48m; S2.

9. At 10m where the far / near field of view monocular cameras are switched, both monocular cameras can image a target of 1m×1m×0.2m.

4. The active-passive fusion camera design method for space target measurement according to claim 3, characterized in that: The step S3 comprises the following steps: S3.

1. Calibration of near-field monocular camera and near-field TOF module. Place the checkerboard calibration plate within the field of view of the near-field TOF module and the near-field monocular camera, and its image is clear. Use the near-field monocular camera and the near-field TOF module to collect 100 images of different angles and positions, respectively, to ensure that the calibration plate has good coverage in the image. For each image, use the ORB corner point extraction algorithm to detect the corner points on the calibration plate. The corner points are defined as the intersection points of the checkerboard. For the near-field monocular camera and the near-field TOF module, use the image and the corresponding corner points to calculate the internal parameters K of the near-field monocular camera and the near-field TOF module. 近-单 With K 近-ToF (such as focal length, principal point position) and external parameters (such as the rotation and translation matrix of the camera). Represents the external parameters from the near-field monocular camera to the near-field TOF module; S3.

2. Calibration of the far-field monocular camera and the far-field TOF module. Place the checkerboard calibration plate within the field of view of the far-field TOF module and the far-field monocular camera, and its image is clear. Use the far-field monocular camera and the far-field TOF module to collect 200 images of different angles and positions respectively to ensure that the calibration plate has good coverage in the image. For each image, use the ORB corner point extraction algorithm to detect the corner points on the calibration plate. The corner points are defined as the intersection points of the checkerboard. For the meta-field monocular camera and the far-field TOF module, use the image and the corresponding corner points to calculate the internal parameters K of the near-field monocular camera and the near-field TOF module. 远-单 With K 远-ToF (such as focal length, principal point position) and external parameters (such as the rotation and translation matrix of the camera). Represents the external parameters from the far-field monocular camera to the far-field TOF module; S3.

3. Calibration of near-field monocular camera and far-field monocular camera. Place the checkerboard calibration plate within the field of view of the near-field monocular camera and the far-field monocular camera, and its image is clear. Use the near-field monocular camera and the far-field monocular camera to collect 25 images of different angles and positions respectively to ensure that the calibration plate has good coverage in the image. For each image, use the ORB corner point extraction algorithm to detect the corner points on the calibration plate. The corner points are defined as the intersection points of the checkerboard. For the near-field monocular camera and the far-field monocular camera, use the image and the corresponding corner points to calculate the camera's internal parameters K 近-单 With K 远-单 (such as focal length, principal point position) and external parameters (such as the rotation and translation matrix of the camera), where Represents the external parameters from the near-field monocular camera to the far-field monocular camera; S3.

4. Calibration of near-field TOF module and far-field TOF module. Make a disc with a diameter of 300 mm. The disc is made of white reflective material. Use the center point of the disc to locate the near-field TOF simulation and the far-field TOF module. Find the disc within the public field of view, use the center of the disc as the common point for calibration calculation, and then perform rigid transformation to obtain the calibration parameter matrix between the two cameras. Represents the external parameters from the near-field ToF module to the far-field ToF module; S3.

5. Calculate the external parameters of the near-field monocular camera and the far-field TOF module Calculate the external parameters of the far-field monocular camera and the near-field TOF module in

Images