Automated imaging devices and imaging methods

Abstract

This invention provides an automated imaging device for image construction of an object under test. The automated imaging system includes a detection mechanism, a synchronization device, and a main control computer. The detection mechanism is located within the detection area of ​​an automated guide rail mechanism that carries and transports the object under test. The detection mechanism includes a mechanical device and a radar device. The mechanical device, which follows a fixed execution path, is triggered by the synchronization device to perform multiple detections by the radar device, generating a plurality of high-speed point data. The data processing module of the main control computer then binds the high-speed point data detected by the radar device with the position information detected by the radar device triggered by the synchronization device. The image reconstruction modeling group of the main control computer receives this information and accurately synthesizes a two-dimensional / three-dimensional internal perspective image, achieving rapid and low-cost automated detection.

Description

Technical Field

[0001] This invention relates to an imaging device, particularly an automated imaging device and imaging method for inspection on automated mechanical devices. Background Technology

[0002] In the current technological landscape of Industry 4.0, non-destructive testing (NDT) is undergoing a profound transformation. Traditional testing methods, such as X-ray or ultrasound inspection, often face challenges related to safety, accessibility limitations, or insufficient resolution when dealing with new composite materials, advanced semiconductor packaging, and precision electronic components.

[0003] The sub-THz and millimeter-wave (mmWave) frequency bands, typically defined as the electromagnetic spectrum between 30 GHz and 1 THz, have become the focus of next-generation detection technologies due to their non-ionizing properties (safety), ability to penetrate non-metallic materials, and high spatial resolution comparable to optical systems. However, the transition of sub-THz detection technology from the laboratory to the production line faces the dual obstacles of imaging efficiency and stability.

[0004] Early systems relied heavily on handheld scanning devices. While these devices offered high operational flexibility, in practical applications, the tremors and irregular movement speeds of the human operator led to extreme spatial non-uniformity in the distribution of sampled points. This non-uniformity is fatal to Synthetic Aperture Radar (SAR) algorithms, as SAR depends on precise phase compensation to simulate large-aperture antennas; any minute displacement error will translate into severe image blurring and phase noise. Summary of the Invention

[0005] To overcome the aforementioned problems, the present invention aims to combine a low-cost single-transmitter-single-receiver (1T1R) radar device with a mechanical structure (such as a robotic arm, a rotary table, or a conveyor belt) and introduce a displacement encoder for hardware-level synchronization, thereby improving the imaging quality and its commercial value.

[0006] The present invention provides an automated imaging device for constructing an image of a test object. The automated imaging system includes: a detection mechanism, a synchronization device, and a main control computer.

[0007] The detection mechanism is located within a detection area of ​​an automated guide rail mechanism that carries and transports the object under test. The detection mechanism includes a mechanical device and a radar device. The mechanical device has a fixture and a driver. The fixture fixes the object under test in a position, and the driver drives the mechanical device to perform a displacement movement, which is a programmable and precise displacement. The radar device is located at a distance from the mechanical device, or is situated at one end of the mechanical device. The radar device operates in the Asia-Pacific Hertz or millimeter-wave band. The radar device emits a signal towards the object under test and receives the signal, converting it into sensing information, which is a single-channel waveform image (A-Scan).

[0008] The synchronization device includes a motion controller and a displacement encoder. The motion controller is electrically connected to the driver and controls the displacement motion and its speed planning according to an execution path. The displacement encoder is electrically connected to the driver and the radar device. After detecting that the displacement motion meets a trigger condition, the displacement encoder sends a pulse signal to the radar device to detect the object under test.

[0009] The main control computer includes: a parameter setting module, a data processing module, and an image remodeling module. The parameter setting module is electrically connected to the motion controller and the radar module, sets multiple parameters, and calculates the execution path. The data processing module is electrically connected to the synchronization device and the radar module. The data binding module acquires the sensing information and a position message that meets the trigger condition, and binds them into trigger data, ensuring that each piece of sensing information carries a corresponding and accurate position message. The image remodeling module is electrically connected to the data binding module. The image remodeling module receives multiple trigger data, performs a synthetic aperture radar (SAR) calculation, and a coordinate conversion operation. The SAR calculation outputs a visual pattern, and the coordinate conversion operation outputs a planar image or a three-dimensional volumetric image.

[0010] The displacement encoder achieves hardware synchronization between the mechanical device and the radar device, generating a plurality of trigger data. The synthetic aperture radar (SAR) calculation utilizes these trigger data during the movement to simulate a large-aperture antenna, focusing the ambiguous signal into a clear point. The coordinate conversion process transforms the position information and the sensed information into one of the following: a two-dimensional longitudinal section image (B-Scan), a horizontal three-dimensional section image (C-Scan), or a 3D point cloud map or a 3D voxel.

[0011] Furthermore, the radar device is a transmit-receive (1T1R) antenna architecture or a transmit-receive (1T1R) antenna architecture arranged in a small array.

[0012] Furthermore, the mechanical device is one of a rotary table, a robotic arm, and a conveyor belt.

[0013] Furthermore, the plurality of these parameters are the set frequency band, step size, depth of focus, type of scanning path for the corresponding mechanical device, and trigger condition.

[0014] Furthermore, the triggering condition is one of a location trigger, a time trigger, and a mixed trigger. The location trigger is triggered by moving a distance, the time trigger is triggered at a time interval, and the mixed trigger is triggered primarily by the location trigger and secondarily by the time trigger.

[0015] In one embodiment, the automated mechanical device is further defined as the rotary table, the displacement encoder is a rotary encoder, and the parameter also includes polar coordinates (r, The setting of , z).

[0016] In one embodiment, the automated mechanical device is further defined as the robotic arm, the displacement encoder is an integrated multi-axis joint encoder, and the parameters also include settings for six degrees of freedom (X, Y, Z, Rx, Ry, Rz).

[0017] In one embodiment, the automated mechanical device is further defined as the conveyor belt, the displacement encoder is a linear encoder, and the parameters also include the setting of Cartesian coordinates (X, Y, Z).

[0018] In one embodiment, the synthetic aperture radar calculation further utilizes the superposition of multiple sensing information sets to simulate a huge virtual antenna array with a lateral resolution Δ The formula is as follows:

[0019] in For wavelength, This is the perspective for synthesizing the aperture.

[0020] In one embodiment, the image remodeling group further includes a preprocessing module that performs preprocessing on the combined reconstructed pattern, the preprocessing including background subtraction, noise suppression, phase correction and geometric correction.

[0021] In one embodiment, the image remodeling group further extracts the peak energy of a plurality of the reconstructed patterns at specific depth levels to form a perspective plane (C-Scan) of the object and obtains the reconstructed pattern of the entire space of the object under test. It then uses a volume reconstruction algorithm to fill the pattern into a three-dimensional matrix (Voxel Grid) to realize a rotatable and sliceable digital twin internal model.

[0022] In one embodiment, the image remodeling group further includes a display and interpretation unit, which is electrically connected to the image remodeling group. The display and interpretation unit analyzes the 2D image, maximum density projection (MIP), isosurface, and volume rendering generated by the image remodeling group, and performs defect segmentation and measurement.

[0023] In one embodiment, the image remodeling group further includes a quality judgment unit, which is electrically connected to the display and interpretation unit. The quality judgment unit interprets the defect segmentation and the measurement, and generates an evaluation value based on the signal-to-noise ratio (SNR), point spread function (PSF) width, contrast, and coverage. When the evaluation value is less than a preset value, the parameter setting is corrected and the trigger line process is repeated.

[0024] In one embodiment, the automated imaging device is further used in conjunction with an adaptive scanning device for detection. The adaptive scanning device uses an optical positioning module to locate a suspected defect area, and then the automated imaging device performs high-resolution and precise detection of the suspected defect area, thereby improving detection efficiency and accuracy.

[0025] Furthermore, the present invention also provides an automated imaging method, performed by an automated imaging apparatus as described in claim 1, comprising the following steps: S1: A testing mechanism, a synchronization device, and a main control computer are provided. The testing mechanism is electrically connected to the synchronization device and the main control computer, and the main control computer is electrically connected to the synchronization device. The mechanical device has a fixture and a driver.

[0026] S2: Set multiple parameters and a trigger condition from a setting module of the main control computer, and calculate an execution path.

[0027] S3: An action controller of the synchronization device receives the execution path and the triggering condition.

[0028] S4: The motion controller moves the mechanical device along the execution path via the drive.

[0029] S5: A displacement encoder of the synchronization device detects whether the movement has completed the triggering condition. If yes, it reports a position message and executes S6; otherwise, it continues to execute S4.

[0030] S6: The displacement encoder sends a pulse signal.

[0031] S7: The radar module receives the pulse signal, transmits a beam to detect a target object, and generates sensing information.

[0032] S8: A data binding module of the main control computer acquires the sensing information and the location information and binds them as trigger data.

[0033] S9: The image remodeling group of the main control computer receives the trigger data.

[0034] S10: The motion controller determines whether the position message is a start position or an end position of the parameter. If yes, execute S11; otherwise, execute S4.

[0035] S11: The image remodeling group performs a synthetic aperture calculation on multiple trigger data and performs a preprocessing to generate a processed reconstructed pattern.

[0036] S12: A display and interpretation unit of the image remodeling group receives the processed reconstructed pattern and performs a defect segmentation and a measurement.

[0037] S13: A quality judgment unit of the image remodeling group receives the defect segmentation and the measurement, and generates an evaluation value.

[0038] S14: The quality judgment unit determines whether the evaluation value is less than a preset evaluation value of the parameter. If not, it corrects multiple parameters and executes S2; if so, it executes S15.

[0039] S15: The image remodeling group performs a coordinate conversion operation on multiple reconstructed patterns after processing, synthesizes the images, and outputs a combined reconstructed pattern.

[0040] Based on the above, the advantages of the present invention are as follows: The automated imaging device of the present invention utilizes known and repeatable displacement motion, and the data processing module binds each single-channel waveform image (A-Scan) acquired by the radar device with a position message detected by the radar device triggered by the synchronization device, into a single piece of information, and accurately synthesizes it in the spatiotemporal dimension. Through the real-time feedback of the position message from the displacement encoder, the main control computer can ensure that each radar echo is tightly bound to the position message, thereby achieving high-quality C-Scan (planar image) and 3D volumetric image reconstruction. This shift from conventional "manual scanning" to "automated aperture synthesis" not only reduces the dependence on expensive large-area array radars, but also provides an economical and efficient technical means for full inspection in industrial production lines. Attached Figure Description

[0041] Figure 1 This is a simplified structural appearance diagram of the automated imaging device of the present invention.

[0042] Figure 2 This is a schematic diagram of the architecture of the automated imaging device of the present invention.

[0043] Figure 3 This is a schematic diagram illustrating the steps of the automated imaging method of the present invention.

[0044] Figure label: 1: Automated imaging device; 11: Adaptive scanning device; 111: Optical positioning module; 2: Testing mechanism; 21: Mechanical device; 211: Fixture; 212: Driver; 213: Location information; 22: Radar device; 221: Signal; 222: Sensor information; 3: Synchronization device; 31: Motion controller; 32: Displacement encoder; 4: Main control computer; 41: Setting parameter module; 411: Parameters; 42: Data processing module; 421: Trigger data; 43: Image remodeling module; 431: Pre-processing module; 44: Display and interpretation unit; 45: Quality judgment unit; 451: Evaluation value; 5: Automated guide rail mechanism; 51: Detection area 71: Synthetic Aperture Radar Calculation; 711: Visual Pattern; 72: Coordinate-to-image conversion; 721: Planar image; 722: 3D volumetric image; 8: Displacement motion; 81: Execution path; 82: Triggering condition; 9: Test object; 91: Suspected defect area; S1~S15: Steps. Detailed Implementation

[0045] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of the present invention. However, the present invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0046] It should be noted that when an element is referred to as being "fixed to" or "set on" another element, it can be directly on the other element or there may be an intervening element. When an element is considered to be "connected to" another element, it can be directly connected to the other element or there may be an intervening element. The terms "vertical," "horizontal," "upper," "lower," "left," "right," and similar expressions used in this specification are for illustrative purposes only and do not represent the only possible embodiments.

[0047] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" or "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0048] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can mean that the first feature is in direct contact with the second feature, or that the first feature and the second feature are in indirect contact through an intermediate medium. Furthermore, "above," "over," or "on top" the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher level than the second feature. "Above," "over," or "on top" the second feature can also mean that the first feature is directly below or diagonally below the second feature, or simply indicates that the first feature is at a lower level than the second feature.

[0049] Unless otherwise defined, all technical and scientific terms used in this specification have the same meaning as commonly known to one of ordinary knowledge in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used in this specification includes any and all combinations of one or more of the associated listed items.

[0050] It should be noted that in the automated imaging device of the present invention, the mechanical device performs displacement movement along the same execution path. Using a synchronization device, when the mechanical device meets the triggering condition, the radar device is simultaneously triggered to perform detection, thereby generating high-speed point data and its position information at the time of triggering. The main control computer then binds the detected high-speed point data and its position information into information. The main control computer receives the information and accurately synthesizes two-dimensional / three-dimensional internal perspective images, thereby achieving fast and low-cost automated detection.

[0051] Please refer to the attached document. Figure 1 as well as Figure 2 As shown, the present invention provides an automated imaging device 1 for constructing an image of a test object 9. The automated imaging device 1 includes: a detection mechanism 2, a synchronization device 3, and a main control computer 4.

[0052] The detection mechanism 2 is located in a detection area 51 of an automated guide rail mechanism 5 that carries and transports the object to be tested 9. The detection mechanism 2 includes a mechanical device 21 and a radar device 22.

[0053] The mechanical device 21 has a fixture 211 and a driver 212. The fixture 211 fixes the object to be tested 9 in a position, and the driver 212 drives the mechanical device 21 to perform a displacement movement 8, which has a programmable precision displacement.

[0054] The radar device 22 is located at a distance from the mechanical device 21, or the radar device 22 is located at one end of the mechanical device 21. The radar device 22 operates in the Asia-Pacific Hertz or millimeter-wave band. The radar device 22 emits a signal 221 towards the object under test 9 and receives the signal, converting it into sensing information 222. The sensing information 222 is a single-channel waveform image (A-Scan). In this invention, the radar device 22 is a transmit-receive (1T1R) antenna architecture or a transmit-receive (1T1R) antenna architecture arranged in a small array.

[0055] The synchronization device 3 includes: an action controller 31 and a displacement encoder 32.

[0056] The motion controller 31 is electrically connected to the driver 212, and the motion controller 31 controls the displacement motion 8 and its speed planning according to an execution path 81.

[0057] The displacement encoder 32 is electrically connected to the driver 212 and the radar device 22. After detecting that the displacement movement 8 meets a trigger condition 82, the displacement encoder 32 sends a pulse signal to the radar device 22 to detect the object under test 9. The trigger condition 82 is one of a position trigger, a time trigger, and a mixed trigger. The position trigger is triggered by moving a distance, the time trigger is triggered at time intervals, and the mixed trigger is triggered primarily by the position trigger and secondarily by the time trigger.

[0058] The main control computer 4 includes: a parameter setting module 41, a data processing module 42, and an image remodeling module 43.

[0059] The parameter setting module 41 is electrically connected to the motion controller 31 and the radar device 22, sets a plurality of parameters 411, and calculates the execution path. In a preferred embodiment of the present invention, the plurality of parameters 411 may be a set frequency band, step size, focusing depth, execution path 81 corresponding to the type of mechanical device, and trigger condition 82, but are not limited thereto.

[0060] The data processing module 42 is electrically connected to the synchronization device 3 and the radar device 22. The data processing module 42 acquires the sensing information 222 and the position information 213 of the mechanical device 21 that meets the trigger condition 82, and binds them into a trigger data 421 to ensure that each sensing information 222 carries the corresponding and accurate position information 213.

[0061] The image remodeling group 43 is electrically connected to the data processing module 42. The image remodeling group 43 receives a plurality of trigger data 421 and performs a synthetic aperture radar operation 71 and a coordinate conversion operation 72. The synthetic aperture radar operation 71 outputs a visualization pattern 711, and the coordinate conversion operation 72 outputs a planar image 721 or a three-dimensional volumetric image 722.

[0062] The displacement encoder 32 achieves hardware synchronization between the mechanical device 21 and the radar device 22, generating a plurality of trigger data 421.

[0063] The synthetic aperture radar (SAR) operation 71 utilizes a plurality of trigger data 421 during the displacement motion 8 to simulate a large-aperture antenna, focusing the ambiguous signal into a clear point. The SAR operation 71 uses the superposition of multiple sensing information 222 to simulate a huge virtual antenna array. However, since the radar device 22 uses a 1T1R or small array sensor, its single-sample physical aperture is extremely small, resulting in limited lateral resolution. Therefore, a lateral resolution Δ is used. The formula is as follows:

[0064] in, For wavelength, This is the perspective for synthesizing the aperture. Through the precise motion planning of this mechanical device 21, the system can easily obtain a huge aperture. This enables sub-millimeter (Sub-mm) level lateral resolution in frequency bands above 100 GHz, which is crucial for finding microcracks in composite materials or identifying open circuits inside encapsulations.

[0065] In the above, the coordinate conversion process 72 converts the position information 213 and the sensing information 222 into one of the following: a two-dimensional longitudinal section image (B-Scan), a horizontal three-dimensional section image (C-Scan), a 3D point cloud map, or a 3D voxel. The data acquisition follows a hierarchical evolution process, aiming to synthesize spatially continuous images from discrete point information.

[0066] In a preferred embodiment of the present invention, the mechanical device 21 is one of a rotary table, a robotic arm, and a conveyor belt, but is not limited thereto.

[0067] In one embodiment, the mechanical device 21 is the rotary table, the displacement encoder 32 is a rotary encoder, and the parameter 411 further includes polar coordinates (r, The setting of z) For cylindrical or centrosymmetric objects (such as lithium battery cells, medicine containers, or precision ceramic parts), this rotary stage provides the most stable imaging solution. In the automated imaging device of this invention, the radar device is typically fixed on a linear guide rail with vertical movement capability, while the object under test is placed on the rotary stage driven by a high-precision actuator. As the rotary stage rotates, a rotary encoder mounted on the actuator sends pulse signals according to the rotation angle. These signals are used not only for motor speed control but also directly as the trigger source for radar sampling. The advantage of this architecture is that its polar coordinate system data acquisition method is highly matched with the geometric features of the object under test, enabling full-circumference image acquisition within one rotation cycle. Subsequently, through a polar coordinate to cassette coordinate conversion algorithm, a three-dimensional cross-sectional image of the object's interior can be reconstructed.

[0068] In the rotational scanning embodiment, the raw data is stored in a polar coordinate system ( , , In order to perform subsequent dimensional measurements and geometric analysis, it is necessary to accurately convert to Cartesian coordinates. , , ): = cos

[0069] = sin

[0070] =

[0071] During the conversion process, non-ideal factors of the mechanical device 21 must be considered, such as the eccentricity of the rotation center or the tilt of the vertical axis. This technical solution specifically defines a "coordinate transformation and imaging" process, which corrects mechanical errors through pre-calibration technology to ensure the geometric fidelity of the 3D point cloud.

[0072] An improved wavenumber domain algorithm (RMA) or polar format algorithm (PFA) is employed. Phase curvature errors caused by circular motion are corrected through stolt interpolation in the wavenumber domain. This enables the system to maintain full-field focus during 360-degree rotation, generating artifact-free, high-contrast images.

[0073] In one embodiment, the mechanical device 21 is a six-axis robotic arm, the displacement encoder 32 is an integrated multi-axis joint encoder, and the parameter 411 also includes the setting of the six-axis robotic arm's degree of freedom parameters, with a total of six degrees of freedom (X, Y, Z, Rx, Ry, Rz) settings.

[0074] For objects with complex curved surface geometries, such as car bumpers, motorcycle helmets, or aircraft composite wings, traditional linear or rotational scanning methods are insufficient. In this embodiment, a radar device 22 is mounted on the end effector of a six-axis industrial robotic arm. The robotic arm plans a scanning path that conforms to the surface of the object under test based on a pre-imported CAD model or optical 3D scan result. The technology behind this approach lies in hardware synchronization under multi-axis linkage. The system must read data from the encoders of each joint of the arm in real time and calculate the precise coordinates of the sensor in space through forward kinematics. Furthermore, to maintain the optimal imaging focal length, the sensor must always remain perpendicular to the surface normal and maintain a constant standoff distance. This highly flexible system enables "Arbitrary Surface SAR," making it the optimal tool for detecting internal defects in asymmetric objects.

[0075] Additionally, traditional radar devices typically use an internal clock trigger, meaning they transmit a signal at fixed intervals. However, in industrial machinery, the starting, stopping, vibration, and load changes of motors all cause fluctuations in the actual moving speed V(t). If time triggering is relied upon, the sampling interval Δ in space... = Δ The aperture will no longer be constant. This will lead to uneven point cloud distribution after imaging, severely affecting the focusing effect of synthetic aperture calculation.

[0076] In contrast, the position triggering mechanism can be directly connected to the pulse output of the displacement encoder. For example, when the linear encoder detects a movement of 10 micrometers ( When the vehicle's speed changes, a voltage transition pulse is generated, which directly drives the front end of the radar device 22 to transmit a beam. This ensures that the distribution of sampling points on the spatial trajectory remains absolutely uniform regardless of the vehicle's speed. For the Asia-Pacific Hertz system, which requires nanosecond-level synchronization accuracy, this is the most direct means of eliminating motion artifacts.

[0077] In one embodiment, the mechanical device 21 is the conveyor belt, the displacement encoder 32 is a linear encoder, and the parameter 411 also includes the setting of Cartesian coordinates (X, Y, Z).

[0078] In scenarios requiring continuous, large-scale full inspection (such as PCB assembly lines, food packaging, or production of wound electrode materials), the conveyor belt is the only feasible solution. In this architecture, the radar device 22 is typically mounted on an X-axis rail spanning the conveyor belt for lateral reciprocating movement, while the conveyor belt provides continuous longitudinal (Y-axis) feed. The synchronization device 3 is particularly important in this scenario. To avoid image stretching and distortion caused by conveyor belt slippage or uneven drive rotation speed, a displacement encoder 32 must be installed at the end of the conveyor belt. Each pulse from the radar device 22 is triggered by a physical increment in the conveyor belt's movement, rather than a fixed time interval. This "distance domain" sampling method ensures that the generated images possess absolute geometric accuracy, facilitating subsequent precise dimensional measurements and defect location.

[0079] In addition, in one embodiment, the image remodeling group 43 further includes a preprocessing module 431, which performs preprocessing on the visualization pattern 711, including background subtraction, noise suppression, phase correction and geometric correction.

[0080] The image remodeling group 43 extracts the peak energy of a plurality of the visualization patterns 711 at specific depth levels to form a perspective plane (C-Scan) of the object, and obtains the visualization pattern 711 in the full space of the object under test. It then uses a volume reconstruction algorithm to fill it into a three-dimensional matrix (Voxel Grid) to realize a rotatable and sliceable digital twin internal model.

[0081] The main control computer 4 also includes a display and interpretation unit 44, which is electrically connected to the image remodeling group 43. The display and interpretation unit 44 analyzes the 2D image, maximum density projection (MIP), isosurface or volume rendering generated by the image remodeling group 43, and performs a defect segmentation and a measurement.

[0082] The main control computer 4 also includes a quality judgment unit 45, which is electrically connected to the display and interpretation unit 44. The quality judgment unit 45 interprets the defect segmentation and measurement, and generates an evaluation value 451 based on the signal-to-noise ratio (SNR), point spread function (PSF) width, contrast and coverage. When the evaluation value 451 is less than a preset evaluation value of the parameter 411, the parameter setting is corrected and the test object 9 is re-tested.

[0083] Additionally, the automated imaging device 1 is used in conjunction with an adaptive scanning device 11 for detection. The adaptive scanning device 11 uses an optical positioning module 111 to locate a suspected defect area 91, and then the automated imaging device 1 performs high-resolution and precise detection on the suspected defect area 91, thereby improving detection efficiency and accuracy.

[0084] Please also refer to the attached document. Figure 3 The present invention also provides an automated imaging method, performed by an automated imaging apparatus as described in claim 1, comprising the following steps: S1: A testing mechanism, a synchronization device, and a main control computer are provided. The testing mechanism is electrically connected to the synchronization device and the main control computer respectively. The main control computer is electrically connected to the synchronization device. The mechanical device has a fixture and a driver. S2: Set a plurality of parameters and a trigger condition from a setting module of the main control computer, and calculate an execution path; S3: An action controller of the synchronization device receives the execution path and the triggering condition; S4: The motion controller moves the mechanical device along the execution path via the driver; S5: A displacement encoder of the synchronization device detects whether the moving motion has completed the triggering condition. If yes, it reports a position message and executes S6; otherwise, it continues to execute S4. S6: The displacement encoder sends a pulse signal; S7: The radar module receives the pulse signal, transmits a beam to detect a target object, and generates sensing information. S8: A data binding module of the main control computer acquires the sensing information and the location information and binds them as trigger data; S9: The image remodeling group of the main control computer receives the trigger data; S10: The motion controller determines whether the position message is a start position or an end position of the parameter. If yes, execute S11; otherwise, execute S4. S11: The image remodeling group performs a synthetic aperture calculation on multiple trigger data and performs a preprocessing to generate a post-processed reconstructed pattern. S12: A display and interpretation unit of the image remodeling group receives the processed reconstructed pattern and performs a defect segmentation and a measurement. S13: A quality judgment unit of the image remodeling group receives the defect segmentation and the measurement, and generates an evaluation value; S14: The quality judgment unit determines whether the evaluation value is less than a preset evaluation value of the parameter. If not, it executes S2 and corrects multiple parameters; if so, it executes S15. S15: The image remodeling group performs a coordinate conversion operation on multiple reconstructed patterns after processing, synthesizes the images, and outputs a combined reconstructed pattern.

[0085] In summary, the core technology of this invention has been described in detail. The advantages of this invention lie in its use of known and repeatable displacement motion. The data processing module binds each single-channel waveform image (A-Scan) acquired by the radar device with a positional message detected by the radar device triggered by the synchronization device, creating a single piece of information that is precisely synthesized in the spatiotemporal dimension. Through real-time feedback of this positional message from the displacement encoder, the main control computer ensures that each radar echo is tightly bound to the positional message, thereby achieving high-quality C-Scan (planar image) and 3D volumetric image reconstruction. This shift from conventional "manual scanning" to "automated aperture synthesis" not only reduces reliance on expensive, large-area array radars but also provides an economical and efficient technical means for full inspection in industrial production lines.

[0086] This image remodeling group can flatten or reconstruct a 3D internal structure diagram of a Cartesian coordinate system from radar echo data in a rotating coordinate system through phase compensation.

[0087] This automated imaging device defines a specific scanning path (e.g., a helical scan or a raster scan) so that the sampling point distribution formed by the 1T1R sensor in space is equivalent to a fully arranged 2D array. In other words, it has a "specific formula for the combination of scanning path and sampling frequency".

[0088] The "surface echo" signal from the radar device is fed back to a controller of the robotic arm in real time, allowing the robotic arm to dynamically fine-tune its height (Z-axis) to maintain the optimal imaging focal length (Auto-focus by Hardware movement).

[0089] The image construction process of the image remodeling group 43 in this invention is based on "position synchronization binding," which gradually expands the measurement data of the radar device 22 from one-dimensional signals to three-dimensional volumetric images. The radar device 22 acquires a set of single-channel waveform images (A-Scan) each time it is triggered, which are essentially the reflection intensity distribution in the range direction, corresponding to a depth axis (z-axis). Unlike conventional methods, this invention binds each A-Scan to the spatial position information provided by the displacement encoder 32 in real time, so that each measurement data corresponds to a unique spatial coordinate position.

[0090] In other words, when the mechanical device 21 is displaced in a single direction, multiple A-scans are arranged according to their corresponding spatial positions to form a two-dimensional longitudinal cross-sectional image (B-scan), with the horizontal axis representing the position in the scanning direction and the vertical axis representing depth information. Furthermore, when the scan covers a complete planar area, the reflection intensity at a specific depth layer can be captured to form a horizontal cross-sectional image (C-scan), which is used to represent the planar structure distribution at that depth.

[0091] In addition, for 3D reconstruction, the system fills all coordinate-corresponding measurement data into a 3D spatial grid, so that each voxel corresponds to a set of spatial coordinates (x, y, z) and its reflection intensity value. Through this position-driven data stacking method, the consistency and continuity of data in spatial distribution can be ensured, thereby establishing a 3D volumetric image with geometric meaning, rather than discrete data stacked only in temporal order.

[0092] Furthermore, the coordinate conversion operation 72 of this invention is used to uniformly convert the raw measurement data obtained from different scanning mechanisms into a spatial coordinate system for imaging and perform image reconstruction. In practical applications, non-ideal factors of the mechanical structure, such as rotation center offset, structural tilt, or trigger delay, must be considered, and a coordinate correction model must be established through pre-calibration to ensure the geometric correctness of the converted spatial coordinates. After the coordinate conversion is completed, each measurement data will be mapped to the corresponding image grid or voxel according to its spatial position, and the sampling gap can be filled through interpolation or weighting to make the image continuous. In addition, when performing the synthetic aperture radar operation 71 for imaging, it is also necessary to use algorithms such as phase compensation and frequency domain resampling to eliminate the phase error introduced by scanning motion and improve the image focusing quality. For example: First, the main control computer 4 sets the parameters 411 and the trigger condition 82 during scanning and plans the execution path 81 of the mechanical device. Once the mechanical device 21 begins to move, the displacement encoder 32 generates a trigger signal based on the actual displacement, driving the radar device 22 to perform measurements. Each A-scan acquired from the measurement is bound to the current spatial position information, forming raw data with coordinate information. As the scan continues, the system arranges the data according to their spatial position to form a B-scan image, and can extract planar information at a specific depth to generate a C-scan. After the overall scan is completed, all data is filled into the voxel space according to its three-dimensional coordinates, forming a preliminary three-dimensional structure. Finally, the image remodeling group 43 performs preprocessing and imaging operations on the voxel space data, including noise suppression, phase correction, and geometric correction, outputting the final two-dimensional cross-sectional image or three-dimensional volumetric image for subsequent display and defect analysis.

[0093] The above detailed description is a specific description of feasible embodiments of the present invention. However, the foregoing embodiments are not intended to limit the patent scope of the present invention. Any equivalent implementation or modification without departing from the spirit of the present invention should be included in the patent scope of this case.

Claims

1. An automated imaging device for constructing an image of an object under test, characterized in that, The automated imaging device includes: A detection mechanism is set within a detection area of ​​an automated guide rail mechanism that carries and transports the object under test. The detection mechanism includes: a mechanical device having a fixture and a driver; the fixture fixes the object under test in a position, and the driver drives the mechanical device to perform a displacement movement, the displacement movement having a programmable precision displacement; and a radar device located at a distance from the mechanical device, or located at one end of the mechanical device, the radar device operating in the Asia-Pacific Hertz or millimeter wave band, the radar device emitting a signal towards the object under test, and receiving the signal and converting it into sensing information, the sensing information being a single-channel waveform image (A-Scan). A synchronization device includes: a motion controller, which is electrically connected to the driver, and the motion controller controls the displacement motion and its speed planning according to an execution path; and a displacement encoder, which is electrically connected to the driver and the radar device, and the displacement encoder sends a pulse signal to the radar device to detect the object under test after detecting that the displacement motion meets a trigger condition. A main control computer includes: a parameter setting module, which is electrically connected to the motion controller and the radar device, sets a plurality of parameters, and calculates the execution path; A data processing module is electrically connected to the synchronization device and the radar device. The data processing module acquires the sensing information and a position message that meets the trigger condition, and binds them into trigger data to ensure that each piece of sensing information carries a corresponding and accurate position message. An image remodeling group is electrically connected to the data processing module. The image remodeling group receives a plurality of trigger data and performs a synthetic aperture radar calculation and a coordinate conversion operation. The synthetic aperture radar calculation outputs a visualization pattern, and the coordinate conversion operation outputs a planar image or a three-dimensional volumetric image. The displacement encoder achieves hardware synchronization between the mechanical device and the radar device, generating multiple trigger data. The synthetic aperture radar calculation utilizes multiple trigger data during the displacement process to simulate a large aperture antenna, focusing the ambiguous signal into a clear point. The coordinate conversion process converts the location information and the sensing information into one of the following: a two-dimensional longitudinal section image (B-Scan), a horizontal three-dimensional section image (C-Scan), a 3D point cloud map, or a 3D voxel.

2. The automated imaging apparatus as claimed in claim 1, characterized in that, The radar device is a transmit-receive (1T1R) antenna architecture or a transmit-receive (1T1R) antenna architecture arranged in a small array.

3. The automated imaging apparatus as claimed in claim 1, characterized in that, The mechanical device consists of one of the following: a rotary table, a robotic arm, and a conveyor belt.

4. The automated imaging apparatus as claimed in claim 1, characterized in that, The multiple parameters are the set frequency band, step size, depth of focus, type of scanning path for the corresponding mechanical device, and trigger condition.

5. The automated imaging apparatus as claimed in claim 1, characterized in that, The triggering condition is one of three: a location trigger, a time trigger, or a mixed trigger. The location trigger is triggered by moving a certain distance, the time trigger is triggered at a certain time interval, and the mixed trigger is triggered primarily by the location trigger and secondarily by the time trigger.

6. The automated imaging apparatus as claimed in claim 3, characterized in that, The automated mechanical device is the rotary table, and the displacement encoder is a rotary encoder. The parameters also include polar coordinates (r, The setting of , z).

7. The automated imaging apparatus as claimed in claim 3, characterized in that, The automated mechanical device is the robotic arm, and the displacement encoder is a multi-axis joint encoder integrated with parameters that also include settings for six degrees of freedom (X, Y, Z, Rx, Ry, Rz).

8. The automated imaging apparatus as claimed in claim 3, characterized in that, The automated mechanical device is the conveyor belt, the displacement encoder is a linear encoder, and the parameters also include the setting of Cartesian coordinates (X, Y, Z).

9. The automated imaging apparatus as claimed in claim 1, characterized in that, This synthetic aperture radar calculation utilizes the superposition of multiple sensed data to simulate a huge virtual antenna array with a lateral resolution Δ. The formula is as follows: in, :wavelength; : Indicates the viewing angle of the synthetic aperture.

10. The automated imaging apparatus as claimed in claim 1, characterized in that, The image remodeling group also includes a preprocessing module that performs preprocessing on the visualization pattern, including background subtraction, noise suppression, phase correction, and geometric correction.

11. The automated imaging apparatus as claimed in claim 1, characterized in that, The image remodeling team extracts the peak energy of a plurality of the visualization patterns at specific depth levels to form a perspective plane (C-Scan) of the object, and obtains the visualization pattern of the entire space of the object under test. A volume reconstruction algorithm is used to fill it into a three-dimensional matrix (Voxel Grid) to realize a rotatable and sliceable digital twin internal model.

12. The automated imaging apparatus as claimed in claim 1, characterized in that, The main control computer also includes a display and interpretation unit, which is electrically connected to the image remodeling group. The display and interpretation unit analyzes the 2D image, maximum density projection (MIP), isosurface or volume rendering generated by the image remodeling group, and performs a defect segmentation and a measurement.

13. The automated imaging apparatus as claimed in claim 12, characterized in that, The main control computer also includes a quality judgment unit, which is electrically connected to the display and interpretation unit. The quality judgment unit interprets the defect segmentation and measurement, and generates an evaluation value based on the signal-to-noise ratio (SNR), point spread function (PSF) width, contrast, and coverage. When the evaluation value is less than a preset value, the parameter settings are corrected and the test object is re-tested.

14. The automated imaging apparatus as claimed in claim 1, characterized in that, The automated imaging device is used in conjunction with an adaptive scanning device for detection. The adaptive scanning device uses an optical positioning module to locate a suspected defect area, and then the automated imaging device performs high-resolution and precise detection of the suspected defect area, thereby improving detection efficiency and accuracy.

15. An automated imaging method, performed by an automated imaging apparatus as described in claim 1, characterized in that, It includes the following steps: S1: A testing mechanism, a synchronization device, and a main control computer are provided. The testing mechanism is electrically connected to the synchronization device and the main control computer respectively. The main control computer is electrically connected to the synchronization device. The mechanical device has a fixture and a driver. S2: Set a plurality of parameters and a trigger condition from a setting module of the main control computer, and calculate an execution path; S3: An action controller of the synchronization device receives the execution path and the triggering condition; S4: The motion controller moves the mechanical device along the execution path via the driver; S5: A displacement encoder of the synchronization device detects whether the moving motion has completed the triggering condition. If yes, it reports a position message and executes S6; otherwise, it continues to execute S4. S6: The displacement encoder sends a pulse signal; S7: The radar device receives the pulse signal, transmits a beam to detect a target object, and generates sensing information; S8: A data binding module of the main control computer acquires the sensing information and the location information and binds them as trigger data; S9: The image remodeling group of the main control computer receives the trigger data; S10: The motion controller determines whether the position message is a start position or an end position of the parameter. If so, execute S11. If not, proceed to S4; S11: The image remodeling group performs a synthetic aperture calculation on multiple trigger data and performs a preprocessing to generate a post-processed reconstructed pattern. S12: A display and interpretation unit of the image remodeling group receives the processed reconstructed pattern and performs a defect segmentation and a measurement. S13: A quality judgment unit of the image remodeling group receives the defect segmentation and the measurement, and generates an evaluation value; S14: The quality judgment unit determines whether the evaluation value is less than a preset evaluation value of the parameter. If not, it corrects multiple parameters and executes S2; if yes, it executes S15. S15: The image remodeling group performs a coordinate-to-image conversion operation on multiple reconstructed patterns after processing, synthesizes the images, and outputs a combined reconstructed pattern.

Images