Automated photographing method and application

Abstract

This application relates to the field of intelligent image capture technology, providing an automated capture method and its application. The automated capture method includes: automated capture; outputting an image quality report based on the capture results; if the image quality report is unqualified, automatically adjusting parameters, re-capturing, and re-outputting the image quality report; if the image quality report is qualified, outputting a qualified image quality report. This application ensures the quality of the final capture report by automating capture and tracking capture results in real time, and can automatically adjust parameters and re-capture based on unqualified reports, thus providing preliminary support for subsequent application scenarios.

Description

Technical Field

[0001] This application belongs to the field of intelligent image capture technology, specifically involving automated capture methods and applications. Background Technology

[0002] The value of small, precision items such as wristwatches heavily relies on their appearance details. These items typically feature polished metal surfaces, plating, and cases with continuously varying curvature, resulting in high surface reflectivity and significant variations in highlight distribution with viewing angle. Current image acquisition methods often employ manual handheld shooting or single-shot shooting from a fixed position. Firstly, unidirectional light sources illuminating curved surfaces create mirror-like highlights, causing overexposure in localized areas. Key features such as hands, markers, and movement details are obscured by these highlights, making it impossible to recover the relevant image information through post-processing. Secondly, due to limitations in imaging depth of field, a single frame image cannot simultaneously and clearly present both the overall outline and microstructure of the watch case, requiring repeated focus adjustments and batch shooting to capture detailed features. Thirdly, the shooting angle, object distance, and lighting conditions depend on the operator's experience; images of the same item acquired at different times and in different batches are unlikely to maintain consistency in posture and scale, resulting in a lack of reproducible coordinate benchmarks for image-based comparisons. Therefore, there is an urgent need to propose an automated shooting method to address the aforementioned challenges in existing technologies. Summary of the Invention

[0003] In view of the shortcomings or deficiencies of the prior art, the technical problem to be solved by this application is an automated shooting method and its application.

[0004] To solve the above-mentioned technical problems, this application provides the following technical solution: This application proposes an automated shooting method, including: Automated shooting; Based on the shooting results, an image quality report is generated; If the image quality report is unsatisfactory, the parameters will be automatically adjusted, the image will be retaken, and the image quality report will be output again. If the imaging quality report is satisfactory, a satisfactory imaging quality report will be output.

[0005] Further, optionally, it also includes: if the image quality report is unqualified, the parameters are automatically adjusted and the image is reshot. If the condition is still not met after exceeding the maximum number of retries, the image quality report is manually reviewed and a qualified image quality report is re-output.

[0006] Further, optionally, it also includes: retaining the reasons for failure in automated shooting.

[0007] Further, optionally, in automated shooting, the following are included: Photograph the object to be tested; Get shooting templates; Panoramic positioning and attitude estimation; Template coordinate mapping and path planning; Based on the imaging module, focusing and shooting are collaboratively acquired; Output an image quality report.

[0008] Further optionally, it also includes receiving tasks: reading order, product, equipment and shooting group information; and / or, selecting a template according to the model of the object to be tested, and if the model is unknown, first entering a general template or AI model recognition process.

[0009] Further optionally, it also includes equipment self-testing and calibration: at least checking the imaging module, panoramic platform, light source, network, and storage; and updating pixel ratio, axis origin, panoramic platform center, and hysteresis compensation through reference markers.

[0010] Further optionally, the above-mentioned shooting of the target location of the object to be tested includes: obtaining the point distribution of the object to be tested, wherein the point distribution is determined by at least some of the following elements: the specific object to be tested, the shooting location, the shooting method, the degree of image capture, and the algorithm used, rather than just saving example images.

[0011] Further optionally, the above-mentioned acquisition of the shooting template includes: converting the normalized points of the template into physical coordinates of the device, preferably satisfying... ,in For template points, For scale / perspective transformation, For the angle compensation of the panoramic shooting platform, This refers to the equipment origin and clamping deviation.

[0012] Optionally, the above-mentioned panoramic positioning and attitude estimation includes: acquiring a panoramic image of the object under test under low magnification and uniform illumination to obtain the actual attitude and scale. Specifically, taking a wristwatch as an example, a panoramic image of the wristwatch is acquired under low magnification and uniform illumination, and the center of the dial, the outline of the case, the direction of the crown, the lugs, the center line of the strap, the clasp, and the outline of the case back are detected to obtain the actual attitude and scale.

[0013] Further optionally, it also includes accessibility and collision checks: the adjustment stroke of the imaging module, the shape of the imaging module, the tooling boundary and the height of the object under test are verified to ensure that the target point is accessible; for unreachable points, the angle of the panoramic imaging platform, the template position or manual processing is used.

[0014] Further optionally, the template coordinate mapping and path planning mentioned above includes: rearranging the points according to the combined costs of movement distance, panoramic platform angle switching, imaging module magnification switching and illumination switching, while retaining the sequential order of the shooting groups.

[0015] Further optionally, in the above-mentioned collaborative acquisition of focusing and shooting based on the imaging module, the following is included: when the panoramic imaging platform used to support and rotate the object under test reaches the target angle, the adjustment mechanism (such as XY-axis adjustment or XYZ-axis adjustment) used to drive the position adjustment of the imaging module reaches the target coordinate, and the magnification, focus and / or illumination and exposure of the controllable illumination module are set; the position feedback meets the error threshold and the imaging module is triggered after the vibration stabilization time.

[0016] Further optional features include multi-parameter acquisition: panoramic and macro images, multi-exposure images, multi-directional illumination images, or multiple focal plane images can be acquired at the same point on the object under test; larger components can be acquired and stitched together in blocks according to XY or XYZ grids.

[0017] Further optionally, it also includes online quality judgment, calculating the image capture sharpness, exposure, highlight saturation, occlusion, component coverage, and posture deviation of the object under test; if it fails, adjust the focus, exposure, lighting, magnification, or micro-shift according to the failure type and then retake the image.

[0018] Further optional steps include packaging and routing: sending the imaging quality report to the next process according to the task tag, such as AI quality inspection, AI identification, or product details output.

[0019] This application also proposes an application of an automated shooting method, which is used at least in scenarios involving product detail images, product condition images, or product authentication images.

[0020] Compared with the prior art, this application has the following technical effects: This application enables automated shooting of the subject matter. For substandard image quality reports, it automatically adjusts parameters, reshoots, and outputs a new image quality report; if the image quality report is satisfactory, it outputs a satisfactory image quality report. By automating shooting and tracking the shooting results in real time, and automatically adjusting parameters and reshooting based on substandard reports, it ensures the quality of the final image report, providing preliminary support for subsequent application scenarios (such as AI quality inspection, AI authentication, product detail images, product condition images, or product authentication images). Attached Figure Description

[0021] Other features, objects, and advantages of this application will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 This is a flowchart illustrating the automated shooting method according to an embodiment of this application; Figure 2 This is a schematic diagram of the automated shooting point location group in the embodiments of this application; Figure 3 This is a flowchart illustrating the coordinate system calibration and the mapping of the object's location template in the embodiments of this application. Figure 4 This is a schematic diagram of the imaging module in an embodiment of this application; Among them, 100-frame, 101-platform base, 110-bearing and rotating mechanism, 120-adjustment mechanism, 121-X-axis guide rail, 122-Y-axis slide, 123-Z-axis lifting mechanism, 130-imaging module, 140-controllable lighting module, 141-ring light source, 142-four-directional light source, 150-motion and acquisition controller, and 160-object under test. Detailed Implementation

[0022] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0023] like Figure 1 As shown, this embodiment provides an automated imaging method. This method achieves closed-loop control of the entire process from image acquisition to result output by constructing a two-way feedback mechanism between physical acquisition and digital analysis. Specifically, this method is not a single linear execution process, but a dynamic iterative system encompassing automated imaging, image quality evaluation, intelligent analysis, and targeted supplementary evidence, used to solve the problem of unreliable results caused by the disconnect between acquisition and analysis in existing technologies. The details are as follows: In step S100, automated shooting is performed. This automated shooting is not merely triggering the shutter, but a complex sequential process integrating device self-checking, panoramic positioning, template registration, path planning, and multi-axis collaborative acquisition. For example, the system first reads the task information and loads the corresponding shooting template based on the model of the object under test 160. Then, it obtains the actual posture of the object under test 160 through panoramic positioning and maps the normalized points in the template to the device's physical coordinates. During acquisition, the motion and acquisition controller coordinates the panoramic shooting platform, adjustment mechanism 120, imaging module 130, and controllable illumination module 140 to perform multi-scale, multi-light, or multi-focal-plane collaborative acquisition of the target points according to a preset path and parameter sequence. It should be understood that although this embodiment uses wristwatch testing as an example, this automated shooting process is also applicable to other objects under test 160 with complex surface features, such as jewelry and precision parts, as long as the acquisition points can be defined using a preset template. The aforementioned adjustment mechanism can adopt an XY two-dimensional relative motion mechanism or an XYZ three-dimensional relative motion structure; the specific structure adopts existing technology.

[0024] In step S200, an imaging quality report is output based on the shooting results. This step constitutes the first line of defense in the physical acquisition layer. Specifically, the online image quality module calculates multiple indicators in real time before the image leaves the device, including sharpness, exposure level, proportion of high-saturation areas, target occlusion rate, component coverage, and pose deviation. The imaging quality report not only includes a binary judgment of "qualified" or "unqualified," but also records the specific values ​​of each indicator and the reason code for failure, providing a precise basis for subsequent automatic adjustments. This online evaluation mechanism ensures the baseline quality of the data entering the algorithm layer, avoiding invalid analysis caused by image blurring, overexposure, or missing key parts.

[0025] In step S300, if the image quality report is unqualified, the system automatically adjusts the parameters, re-shoots, and outputs another image quality report. This is a rule-based physical layer self-healing closed loop. For example, when image defocus is detected, the system automatically adjusts the focus position of the imaging module or re-executes the autofocus process; when local overexposure is detected, the system reduces the exposure time or switches to a low-brightness light source channel; when insufficient target coverage is detected, the system fine-tunes the XY-axis coordinates or XYZ-axis coordinates or adjusts the angle of the panoramic imaging platform to correct the field of view. Only when the re-shot image passes the quality gate after parameter adjustment will it be marked as valid data and proceed to the next stage. If the maximum number of retries still fails to meet the quality requirements, the system retains the reason for the failure and transfers it to the manual review process, thus ensuring automation efficiency while providing a safety net for extreme abnormal situations.

[0026] In step S400, if the imaging quality report is qualified, a qualified imaging quality report is output.

[0027] The specific implementation process and parameterized control logic of automated shooting are further refined. Automated shooting includes: shooting the object under test (160°); acquiring a shooting template; panoramic positioning and attitude estimation; template coordinate mapping and path planning; collaborative acquisition of focusing and shooting based on the imaging module; and outputting an imaging quality report. This series of steps constitutes a complete time-series chain from task analysis to physical execution and data verification, ensuring the standardization and repeatability of the acquisition process.

[0028] Specifically, the process begins with step S101, which involves receiving the task: reading order, product, equipment, and shooting group information; and / or selecting a template based on the model of the object under test 160. If the model is unknown, the system first enters the general template or AI model recognition process. In this step, the system not only reads basic business order information but also loads the corresponding digital shooting template based on the specific model of the object under test 160. This embodiment uses a wristwatch location distribution map as an example for illustration. See [link to documentation]. Figure 2As shown. It should be understood that the template here is not a simple collection of example images, but a structured data file containing all elements such as target component definition, normalized point coordinates, recommended magnification, focus strategy, lighting formula, exposure parameters, panoramic shooting platform angle, quality threshold, and algorithm task tags. This parameterized template definition method allows the same hardware equipment to be adapted to different categories and models of test objects 160 through software configuration without frequent tooling changes or re-teaching. When encountering a test object 160 with an unknown model, the system can automatically switch to a general template for exploratory shooting, or trigger an AI model recognition process to identify the model before loading a dedicated template, thereby ensuring the continuity of the business process.

[0029] In step S102, the device performs self-testing and calibration: at least the imaging module, panoramic imaging platform, light source, network, and storage are checked; and the pixel ratio, axis origin, panoramic imaging platform center, and hysteresis compensation are updated through reference markers. This is the physical basis for ensuring the accuracy of subsequent coordinate mapping. For example, before each power-on or task start, the system drives the imaging module to photograph a preset calibration board or reference marker, and uses image processing algorithms to calculate in real time the current pixel physical ratio (mm / pixel), XY axis mechanical origin, panoramic imaging platform rotation center offset, and transmission mechanism hysteresis compensation value. These calibration parameters are written into the device status object for subsequent steps to call. Through this dynamic calibration mechanism, systematic errors caused by device thermal deformation, mechanical wear, or re-clamping can be effectively eliminated, ensuring consistent acquisition over long-term operation.

[0030] In step S103, panoramic positioning and attitude estimation are performed: a panoramic image of the test object 160 is acquired under low magnification and uniform illumination to obtain its actual attitude and scale. Specifically, taking the test object 160 as a wristwatch as an example, the system acquires a panoramic image containing the entire watch under low magnification and a large field of view conditions with uniform illumination. Subsequently, a visual algorithm is used to detect key geometric features such as the dial center, case outline, crown direction, lug position, strap centerline, clasp, and case back outline. Based on the relative positional relationships of these features, the system can calculate the actual placement attitude of the watch on the current fixture (such as rotation angle and tilt angle) and the scale factor relative to the standard template. This step achieves preliminary alignment from the "ideal template space" to the "actual physical space," providing the necessary transformation parameters for subsequent precise coordinate mapping.

[0031] In step S104, template coordinate mapping is performed: the normalized points of the template are converted into physical coordinates of the device. See [link to relevant documentation]. Figure 3 As shown. This conversion process preferably satisfies the following formula:

[0032] in, These are the normalized point coordinates defined in the template, and their value range is usually normalized to the [0,1] interval or a specific logical coordinate system. The scale / perspective transformation matrix is ​​calculated based on the posture and scale information obtained in step S103 and is used to correct the projection distortion of the object under test 160 caused by changes in placement angle or distance. This is the angle compensation matrix for the panoramic imaging platform, used to map two-dimensional planar coordinates to a three-dimensional execution space that takes into account the rotation angle; The equipment origin and clamping deviation vector, determined by the calibration results in step S102, are used to eliminate the inherent offset between the mechanical zero point and the visual zero point. Through this formula, the system can accurately convert abstract, equipment-independent template points into physical coordinate commands executable by the adjustment mechanism and the panoramic imaging platform, thereby achieving adaptive and precise positioning across batches and individuals.

[0033] Before executing physical movements, the system performs accessibility and collision checks: verifying whether the imaging module's adjustment stroke, shape, tooling boundaries, and the 160° height of the object under test support reaching the target point; for inaccessible points, adjusting the angle of the panoramic imaging platform, the template position, or manually handling the issue. This is a safety defense mechanism to prevent equipment damage due to extreme clamping postures or irregular structures. For example, when it detects that the crown is too high and may touch the lens, the system will automatically attempt to rotate the panoramic imaging platform to avoid the interference zone, or mark the point as "requiring manual assistance," rather than forcibly executing movement commands that may cause malfunctions.

[0034] In step S106, path planning is performed: the points are rearranged according to the combined costs of movement distance, panoramic platform switching, imaging module magnification switching, and illumination switching, while preserving the sequential order of shooting groups with dependencies. To improve acquisition efficiency, the system does not simply execute according to the ID order defined in the template, but uses the Traveling Salesman Problem (TSP) or similar optimization algorithms to comprehensively consider the action time of each actuator. For example, due to the large rotational inertia and long settling time of the panoramic platform, the system will prioritize executing all points with the same angle consecutively before switching angles; similarly, magnification and illumination switching are also included in the cost function for global optimization. However, for shooting groups with logical dependencies (such as requiring a panoramic positioning image to be captured before a macro detail image can be captured), the system will strictly adhere to the sequential constraints to ensure the logical correctness of data acquisition.

[0035] In steps S107 and S108, collaborative acquisition based on focusing and shooting using the imaging module 130 is performed. Specifically, this includes: when the panoramic imaging platform used to support and rotate the object 160 reaches the target angle, and the adjustment mechanism used to drive the imaging module's position adjustment reaches the target coordinates, setting the magnification, focus, and / or illumination and exposure of the controllable illumination module 140; and triggering the imaging module after the position feedback meets the error threshold and the vibration stabilization time has elapsed. Here, the two key points of "collaboration" and "stabilization" are emphasized. In multi-axis linkage scenarios, a simple position signal does not necessarily mean that the imaging conditions are ready. After mechanical movement stops, there is often slight residual vibration; if the shutter is triggered immediately, the image will be blurred. Therefore, this embodiment introduces a "vibration stabilization time" as a hard waiting window, combined with a position feedback error threshold for double confirmation. Only when the deviation between the actual position fed back by the encoder and the target position is consistently less than the set threshold, and the stabilization timer times out, will the system issue a trigger signal. Furthermore, multi-parameter acquisition can be performed at the same location: panoramic and macro images, multi-exposure images, multi-directional illumination images, or multiple focal plane images can be acquired at the same location on the object under test (160°); larger components can be acquired and stitched together in blocks according to XY or XYZ grids. This multi-dimensional data acquisition strategy provides rich information redundancy for subsequent AI quality inspection and identification, effectively solving the problem of feature loss under a single viewpoint or single illumination.

[0036] In step S109, online quality assessment is performed: the image sharpness, exposure, highlight saturation, occlusion, component coverage, and attitude deviation of the test object 160 are calculated; if the image is unqualified, the focus, exposure, lighting, magnification, or micro-shift is adjusted according to the failure type before reshooting. This is the core defense depth of the physical acquisition layer. The system has built-in multiple targeted self-healing strategies: for example, when insufficient sharpness is detected, refocusing or fine-tuning the focal plane is automatically triggered; when local highlight overexposure is detected, the exposure time is automatically reduced or the diffuse light source channel is switched; when the target coverage is detected as low, the XY coordinates are automatically fine-tuned or the magnification is adjusted to correct the composition. If the imaging quality report is unqualified, the parameters are automatically adjusted and the image is reshot. If the condition is still not met after exceeding the maximum number of retries, the image is manually reviewed and a qualified imaging quality report is re-output. At the same time, the system also includes: retaining the reasons for failure in automated shooting. These failure logs (such as "3 consecutive focus failures" or "constant overexposure at a specific angle") are not only the basis for handling the abnormality of this task, but also important data assets for subsequent optimization of template parameters, improvement of hardware design, or training of quality evaluation models. Through this closed-loop mechanism of "detection-adjustment-recording", the system can maximize the availability of the output data even when unattended.

[0037] Finally, in step S110, packaging and routing are performed: the imaging quality report is sent to the next process according to the task tag, such as to AI quality inspection, AI authentication, or product detail output scenarios. Qualified image data, along with its corresponding point ID, actual acquisition parameters (such as actual exposure value and actual coordinates), and quality score, are packaged into standardized data packets. Based on the tags specified during task creation, these data streams are routed to different downstream processing modules. For example, image groups labeled "quality inspection" enter the AI ​​quality inspection module for defect classification, image groups labeled "authentication" enter the AI ​​authentication module for authenticity comparison, and image groups labeled "product detail" are directly used to generate product display materials. This tag-based routing mechanism allows the same acquisition system to flexibly support multiple business scenarios, maximizing the utilization of hardware resources.

[0038] This embodiment provides an application of an automated shooting method, which can be used at least for scenarios involving product detail images, product condition images, or product authentication images. It should be understood that this embodiment is merely an exemplary application of the technical solution of this application in a specific field and is not intended to limit the scope of protection of this application.

[0039] Reference Figure 4 As shown, this embodiment also proposes an automated imaging system, which is the physical carrier for implementing the aforementioned automated imaging method. Specifically, the system includes an imaging module 130 and an imaging quality evaluation module. The imaging module 130 is used for automated imaging of the object 160 under test; the imaging quality evaluation module is used to output an imaging quality report based on the imaging results, and automatically adjusts parameters to re-encode the image and outputs another imaging quality report if the report is unqualified. It should be understood that although this embodiment describes each function as an independent module, in actual engineering implementation, these modules can be deployed as independent software units on a local processor or cloud server, or integrated into the same computing unit. Even some lightweight evaluation logic can be embedded in the firmware of the imaging module 130, as long as it can achieve the aforementioned functional closed loop.

[0040] To support the operation of the aforementioned functional modules, the system of this application employs a sophisticated mechatronics-optical integrated architecture at the hardware level. For example... Figure 4As shown, the system includes a frame 100, which defines an imaging space. An imaging module 130 and a support and rotation mechanism 110 are also installed within this imaging space. The frame 100 not only supports and fixes the components but, more importantly, creates at least a semi-enclosed environment, effectively shielding against external ambient light interference and ensuring the stability and repeatability of the illumination conditions of the controllable illumination module 140 during each image capture. The imaging module 130 is positioned opposite the support and rotation mechanism 110 and is used to image the object under test 160. In terms of specific layout, the optical axis of the imaging module 130 is typically perpendicular to the support surface of the object under test 160, or it can be set at an angle according to specific detection requirements. The support and rotation mechanism 110 includes a support portion for supporting the object under test 160 and a circular imaging platform for driving the support portion to rotate 360°. The support portion is mounted on the circular imaging platform, which is mounted on the frame 100. The carrier unit can use replaceable special fixtures or universal adaptive tooling to adapt to watches or other objects to be tested of different sizes and shapes. The panoramic imaging platform has a built-in high-precision angle encoder and servo motor, which can achieve accurate positioning and speed control at any angle, providing a physical basis for panoramic acquisition and multi-angle feature extraction.

[0041] To achieve precise coverage of any point on the surface of the object 160 under test, the system also includes an adjustment mechanism 120. This adjustment mechanism 120 is used to adjust the distance between the imaging module 130 and the bearing surface of the carrying and rotating mechanism 110. The carrying and rotating mechanism 110 includes a panoramic imaging platform driven by a drive mechanism. The panoramic imaging platform is used to support and place the object 160 under test. The carrying and rotating mechanism 110 can be mounted on the platform base 101. Specifically, in conjunction with... Figure 4 As shown, the adjustment mechanism 120 can specifically be embodied as a combination of an X-axis guide rail 121, a Y-axis slide 122, and a Z-axis lifting mechanism 123, which is only an example. In another embodiment, the adjustment mechanism 120 includes two linear motion units (X-axis guide rail 121 and Y-axis slide 122) in intersecting directions, a driver, and a position feedback component, enabling the imaging module 130 to reach any target point within a two-dimensional area parallel to the bearing surface; more preferably, the working distance is dynamically adjusted through adjustment in the Z-axis direction or the automatic focusing mechanism of the imaging module 130 itself. This multi-degree-of-freedom motion capability is the physical prerequisite for executing the "directional reshooting scheme" in the aforementioned embodiments. For example, when the algorithm determines that macro details on the side of the crown need to be reshot, the adjustment mechanism 120 can accurately move the imaging module 130 to a preset three-dimensional coordinate and adjust it to a suitable working distance, thereby transforming the abstract reshooting command into a real physical acquisition action. Figure 4 The dashed lines in the diagram represent control connections, and the blue arrows represent controllable degrees of freedom.

[0042] Illumination conditions are a key factor determining image quality and feature extraction effectiveness. Therefore, the system also includes a controllable illumination module 140 disposed within the imaging space. This controllable illumination module 140 includes at least two light source units with different incident directions or angles. This embodiment illustrates the possible configuration of a ring light source 141 and a four-way light source 142. The ring light source 141 and the four-way light source 142 can each have independent driving channels. These source units are individually controlled in terms of on / off states and brightness under the control of the motion and acquisition controller. Figure 4 As shown, in a preferred embodiment, the controllable lighting module 140 may include four surface light sources with an incident angle of approximately 45 degrees and a high-angle ring light source with an incident angle of approximately 60 degrees, each light source being equipped with an independent constant current drive circuit. This branch-independent control structure allows the system to flexibly combine lighting formulas according to the material characteristics of the surface of the object under test 160 (such as metallic reflection, glass transmission, and leather diffuse reflection). For example, when detecting scratches on a dial, only low-angle grazing light can be turned on to enhance texture contrast; when performing OCR character recognition, uniform diffuse light can be turned on to eliminate shadows; when identifying micro-processes, different light source sequences can be switched to obtain multi-dimensional surface morphology information. It should be understood that the number, angle, and type of light sources are not limited to the examples above, as long as at least two differentiated lighting conditions can be provided and independent control is supported, they are all within the scope of protection of this application.

[0043] Above the aforementioned hardware architecture, the system also includes a motion and acquisition controller 150, a calibration module, a template and task engine, an online image quality module, a closed-loop and result module, and a communication module. The motion and acquisition controller 150, as the central hub for hardware scheduling, coordinates the timing of the actions of the panoramic imaging platform, adjustment mechanism 120, imaging module 130, and controllable illumination module 140, ensuring that the shutter is triggered only after mechanical vibration has stabilized, and records the actual acquisition parameters corresponding to each image. The calibration module, using a calibration board or reference marker, provides the system with key geometric parameters such as pixel ratio, axis origin, panoramic imaging center, and hysteresis compensation, serving as the physical reference for establishing the mapping relationship between template coordinates and equipment physical coordinates. The template and task engine, online image quality module, and closed-loop and result module constitute the system's intelligent decision-making layer. They can be deployed on a local industrial control computer or a cloud server and maintain real-time data interaction with the hardware actuators through the communication module. It is through this deep coupling of software and hardware that the system of this application is able to transform the "perception-decision-execution" closed loop in the aforementioned method embodiment from a logical concept into a workable engineering entity, thereby achieving full-process automation and intelligence while ensuring detection accuracy.

[0044] The above embodiments are only used to illustrate the technical solutions of this application and are not intended to limit it. The preferred embodiments have been described in detail. Those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of this application without departing from the spirit and scope of the technical solutions of this application, and all such modifications and substitutions should be covered within the scope of the claims of this application.

Claims

1. An automated shooting method, characterized in that, include: Automated shooting; Based on the shooting results, an image quality report is generated; If the image quality report is unsatisfactory, the parameters will be automatically adjusted, the image will be retaken, and the image quality report will be output again. If the imaging quality report is satisfactory, a satisfactory imaging quality report will be output.

2. The automated shooting method according to claim 1, characterized in that, The automated shooting process described above also includes: if the image quality report is unqualified, the parameters are automatically adjusted and the image is reshot. If the maximum number of retries is exceeded and the conditions are still not met, the process is transferred to manual review and a qualified image quality report is re-output.

3. The automated shooting method according to claim 2, characterized in that, Also includes: Reasons for failures in automated shooting.

4. The automated shooting method according to claim 1, characterized in that, In automated shooting, the following are included: Photograph the object to be tested; Get shooting templates; Panoramic positioning and attitude estimation; Template coordinate mapping and path planning; Based on the imaging module, focusing and shooting are collaboratively acquired; Output an image quality report.

5. The automated shooting method according to claim 4, characterized in that, It also includes receiving tasks: reading order, product, equipment and shooting group information; and / or, selecting a template according to the model of the object to be tested, and if the model is unknown, first entering the general template or AI model recognition process.

6. The automated shooting method according to claim 4, characterized in that, It also includes equipment self-testing and calibration: at least checking the imaging module, panoramic imaging platform, light source, network, and storage; and updating the pixel ratio, axis origin, panoramic imaging platform center, and hysteresis compensation through reference markers.

7. The automated shooting method according to claim 4, characterized in that, The above-mentioned shooting of the test object includes: obtaining the point distribution of the test object, wherein the point distribution is determined by at least some of the following elements: the specific test object, the shooting location, the shooting method, the degree of image capture, and the algorithm used, etc.

8. The automated shooting method according to claim 4, characterized in that, The above-mentioned acquisition of the shooting template includes: converting the normalized points of the template into physical coordinates of the device, preferably satisfying the following: ,in For template points, For scale / perspective transformation, For the angle compensation of the panoramic shooting platform, This refers to the equipment origin and clamping deviation.

9. The automated shooting method according to claim 4, characterized in that, The aforementioned panoramic positioning and attitude estimation includes: acquiring a panoramic image of the object under test under low magnification and uniform illumination to obtain its actual attitude and dimensions. Specifically, taking a wristwatch as an example, a panoramic image of the wristwatch is acquired under low magnification and uniform illumination, and the center of the dial, the outline of the case, the direction of the crown, the lugs, the center line of the strap, the clasp, and the outline of the case back are detected to obtain the actual attitude and dimensions.

10. The automated shooting method according to claim 4, characterized in that, It also includes accessibility and collision checks: verifying the reachability of the target point by adjusting the travel of the imaging module, the shape of the imaging module, the tooling boundary, and the height of the object under test; for unreachable points, adjusting the angle of the panoramic imaging platform, the template position, or handling them manually.

11. The automated shooting method according to claim 4, characterized in that, The template coordinate mapping and path planning mentioned above include rearranging the points according to the combined costs of movement distance, pan-shooting platform switching, imaging module magnification switching and illumination switching, while retaining the order of shooting groups that depend on each other.

12. The automated shooting method according to claim 4, characterized in that, In the above-mentioned collaborative acquisition based on focusing and shooting of the imaging module, the following steps are included: when the panoramic imaging platform used to support and rotate the object under test reaches the target angle, the adjustment mechanism used to drive the position adjustment of the imaging module reaches the target coordinates, and the magnification, focus and / or illumination and exposure of the controllable illumination module are set; the position feedback meets the error threshold and the imaging module is triggered after the vibration stabilization time has elapsed.

13. The automated shooting method according to claim 4, characterized in that, It also includes multi-parameter acquisition: panoramic and macro images, multi-exposure, multi-directional illumination or multiple focal plane images can be acquired at the same point of the object under test; larger parts can be acquired and stitched together according to XY or XYZ grid blocks.

14. The automated shooting method according to claim 4, characterized in that, It also includes online quality assessment, calculating the image sharpness, exposure, highlight saturation, occlusion, component coverage, and posture deviation of the object under test; if it fails, it adjusts the focus, exposure, lighting, magnification, or micro-shift according to the failure type and then retakes the image.

15. The automated shooting method according to any one of claims 1 to 14, characterized in that, It also includes packaging and routing: sending the imaging quality report to the next process according to the task tag.

16. The application of the automated shooting method as described in any one of claims 1 to 15, characterized in that, It should be used at least for product detail images, product condition images, or product authentication images.

Images