A product quality inspection method, system and medium based on multiple cameras

Abstract

The present invention discloses a product quality inspection method, system and medium based on multiple camera terminals, the method comprising the following steps: setting model parameters and configuration parameters, configuring a first camera module based on the model parameters and the configuration parameters; obtaining quality inspection requirements, setting a startup quantity threshold, performing an image capture operation based on the first camera module, the quality inspection requirements and the startup quantity threshold, and obtaining a product image; obtaining a reference image package, and performing a product quality inspection judgment operation based on the product image and the reference image package; the present invention can automatically identify the features of an object to be inspected, and intelligently adapt a corresponding camera terminal according to the features, use a plurality of camera terminals to capture a three-dimensional image of a product for three-dimensional restoration and reproduction, and then perform digital quality inspection on the product based on the restored three-dimensional image, which can not only intelligently adapt to a variety of different products, but also improve the efficiency and accuracy of quality inspection, and has extremely high application value in a production workshop.

Description

Technical Field

[0001] The present invention relates to the technical field of product quality inspection, and in particular to a product quality inspection method, system and medium based on multiple camera terminals. Background Art

[0002] At present, most production workshops still use manual quality inspection of products before warehousing to control the production quality of products. This quality inspection method is inefficient, has high labor costs, and is prone to certain quality inspection omissions due to the uncertainty of manual operation. Therefore, it is necessary to design a method for product quality inspection using automated equipment, thereby improving product quality inspection efficiency and reducing quality inspection omissions such as false inspections and missed inspections. Summary of the invention

[0003] The main purpose of the present invention is to design a method for product quality inspection using automated equipment, thereby improving product quality inspection efficiency and reducing quality inspection omissions such as false inspections and missed inspections.

[0004] To achieve the above object, a technical solution adopted by the present invention is to provide a product quality inspection method based on multiple cameras, comprising the following steps:

[0005] Camera module configuration step: setting model parameters and configuration parameters, and configuring the first camera module based on the model parameters and the configuration parameters;

[0006] Product image capturing step: obtaining quality inspection requirements, setting a startup quantity threshold, performing an image capturing operation based on the first camera module, the quality inspection requirements and the startup quantity threshold, and obtaining a product image;

[0007] Product comparison and quality inspection step: obtaining a reference image package, and performing a product quality inspection and judgment operation based on the product image and the reference image package.

[0008] As an improved solution, the model parameters include: model shape information and model size information;

[0009] The configuration parameters include: a first spacing distance and a second spacing distance;

[0010] The step of configuring the first camera module based on the model parameters and the configuration parameters further includes:

[0011] Read the model parameters; construct a three-dimensional model block based on the model shape information and the model size information; identify a plurality of first corner points of the three-dimensional model block, and select a second corner point from the plurality of first corner points; establish a three-dimensional coordinate system with the second corner point as the origin; confirm a first axis of the three-dimensional model block based on the three-dimensional coordinate system; set the first corner point located on the first axis as a third corner point;

[0012] A first feature recognition unit is disposed at a position directly above the second corner point and at a distance corresponding to the first interval from the second corner point;

[0013] Counting a first number of the third corner points; configuring first camera units corresponding to the third corner points one by one according to the first number; confirming an outer edge area of ​​the three-dimensional model block; and aligning the first camera unit with the third corner point corresponding to the first camera unit in the outer edge area according to the second spacing distance.

[0014] As an improved solution, there are at least two first camera units; the first feature recognition unit and a plurality of the first camera units constitute the first camera module;

[0015] The first camera unit corresponding to the third corner point is located on an extension line of the first axis where the third corner point is located.

[0016] As an improved solution, the quality inspection requirement includes: a first requirement and a second requirement; the first requirement is that the first product to be inspected exists; the second requirement is that the first product to be inspected does not exist;

[0017] The image capturing operation includes:

[0018] Identifying the quality inspection requirements;

[0019] If the quality inspection requirement is the first requirement, calling the first feature recognition unit to obtain the first feature of the first product to be inspected; calculating the feature specification and feature shape of the first product to be inspected based on the first feature; and confirming a second camera unit among the plurality of first camera units based on the feature shape;

[0020] Counting a second number of the second camera units, determining whether the second number matches the activation number threshold; if so, setting the second camera unit as a camera unit to be activated; if not, identifying a third camera unit from a plurality of the second camera units based on the feature specification, and setting the third camera unit as a camera unit to be activated;

[0021] A first capturing step is performed based on the camera unit to be started to obtain the product image.

[0022] As an improved solution, the reference image package stores: a source image of the product to be inspected and an image of a reference inspection point;

[0023] The product quality inspection and judgment operation includes:

[0024] Reading the reference image package;

[0025] Based on the source image of the product to be detected, the product image is subjected to three-dimensional restoration processing to obtain a three-dimensional model image; based on the source image of the product to be detected, it is determined whether the three-dimensional model image has image missing;

[0026] If there is no image missing, select the point to be inspected in the three-dimensional model image based on the reference detection point image; enlarge the three-dimensional model image corresponding to the point to be inspected to obtain an enlarged model image; compare the enlarged model image with the reference detection point image to see if they match. If they match, output the first detection result; if not, output the second detection result.

[0027] As an improved solution, the product quality inspection and judgment operation further includes:

[0028] If there is an image missing, confirm the missing image based on the source image of the product to be detected and the three-dimensional model image; identify the first position information of the missing image relative to the source image of the product to be detected; identify a backup camera unit among several first camera units based on the first position information; and return to the first capturing step based on the camera unit to be started and the backup camera unit.

[0029] As an improved solution, the first capturing step performed by the to-be-started camera unit includes:

[0030] The to-be-activated camera unit is activated to capture a first image of the first product to be inspected, and the first image is set as the product image.

[0031] As an improved solution, the first test result is that the product is qualified; the second test result is that the product is unqualified;

[0032] The product quality inspection method based on multiple cameras also includes:

[0033] Configure a manual operation detection unit and manual inspection notification information, and obtain the communication address of the administrator of the first product to be inspected; when the second inspection result is output in the product quality inspection judgment operation, set the enlarged model image that does not match the reference inspection point image as a non-compliant image; identify the second position information of the non-compliant image with respect to the source image of the product to be inspected; and send the manual inspection notification information and the second position information to the communication address of the administrator.

[0034] The present invention also provides a product quality inspection system based on multiple cameras, comprising:

[0035] Module configuration module, capture control module and quality inspection control module;

[0036] The module configuration module is used to set model parameters and configuration parameters, and configure the first camera module based on the model parameters and the configuration parameters;

[0037] The capture control module is used to obtain quality inspection requirements and set a startup quantity threshold, and the capture control module performs an image capture operation based on the first camera module, the quality inspection requirements and the startup quantity threshold to obtain a product image;

[0038] The quality inspection control module is used to obtain a reference image package and perform a product quality inspection judgment operation based on the product image and the reference image package.

[0039] The present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the product quality inspection method based on multiple camera terminals are implemented.

[0040] The beneficial effects of the present invention are:

[0041] 1. The product quality inspection method based on multiple cameras described in the present invention can realize automatic identification of features of the object to be inspected, and intelligently adapt the corresponding camera according to the features, and use the three-dimensional images of the product captured by multiple cameras to perform three-dimensional restoration and reproduction, and then perform digital quality inspection on the product based on the restored three-dimensional images. It can not only intelligently adapt to a variety of different products, but also improve the efficiency and accuracy of quality inspection, which makes up for the shortcomings of the existing technology and has extremely high application value in production workshops.

[0042] 2. The product quality inspection system based on multiple cameras described in the present invention can realize automatic identification of the characteristics of the object to be inspected through the mutual cooperation of the module configuration module, the capture control module and the quality inspection control module, and intelligently adapt the corresponding camera end according to the characteristics, and use the three-dimensional images of the products captured by multiple camera ends to perform three-dimensional restoration and reproduction, and then perform digital quality inspection on the products based on the restored three-dimensional images. It can not only intelligently adapt to a variety of different products, but also improve the efficiency and accuracy of quality inspection, which makes up for the shortcomings of the existing technology and has extremely high application value in production workshops.

[0043] 3. The computer-readable storage medium described in the present invention can realize the coordination of the guidance module configuration module, the capture control module and the quality inspection control module, thereby realizing automatic identification of the features of the object to be inspected, and intelligently adapting the corresponding camera end according to the features, and using the three-dimensional images of the product captured by multiple camera ends to perform three-dimensional restoration and reproduction, and then digitally inspecting the product according to the restored three-dimensional images. It can not only intelligently adapt to a variety of different products, but also improve the efficiency and accuracy of quality inspection, make up for the shortcomings of the existing technology, and effectively improve the operability of the product quality inspection method based on multiple camera ends. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0045] Figure 1 is a flow chart of the product quality inspection method based on multiple cameras according to Embodiment 1 of the present invention;

[0046] Figure 2 It is a schematic diagram of a specific process of the product quality inspection method based on multiple cameras according to Embodiment 1 of the present invention;

[0047] Figure 3 It is a schematic diagram of the implementation effect of the three-dimensional coordinate system, the three-dimensional model block, the third corner point and the first axis in Embodiment 1 of the present invention;

[0048] Figure 4 This is an architecture diagram of a product quality inspection system based on multiple cameras as described in Example 2 of the present invention. DETAILED DESCRIPTION

[0049] The preferred embodiments of the present invention are described in detail below in conjunction with the accompanying drawings so that the advantages and features of the present invention can be more easily understood by those skilled in the art, thereby making a clearer and more definite definition of the protection scope of the present invention.

[0050] In the description of the present invention, it should be noted that the embodiments described in the present invention are only part of the embodiments of the present invention, rather than all of the embodiments; based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0051] In the description of the present invention, it should be noted that the terms “first”, “second” and “third” are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0052] Example 1

[0053] This embodiment provides a product quality inspection method based on multiple cameras. Figure 1 to Figure 3 As shown, the following steps are included:

[0054] S100, camera module configuration steps, specifically including:

[0055] S110, setting model parameters and configuration parameters, and configuring the first camera module based on the model parameters and the configuration parameters;

[0056] Specifically, the model parameters include: model shape information and model size information; in this embodiment, the model shape information is the shape of the three-dimensional model to be established, which is a regular cube in this embodiment; the model size information is the size of the volume of the three-dimensional model to be established, that is, the length, width and height; correspondingly, the model size information is set according to the maximum specification of the equipment produced in the production workshop to which the method is applied;

[0057] Specifically, the configuration parameters include: a first spacing distance and a second spacing distance; in this embodiment, the first spacing distance is a parameter related to installing a feature recognition module, and the second spacing distance is a parameter related to installing a camera module;

[0058] Specifically, the step of configuring the first camera module based on the model parameters further includes: reading the model parameters; using the actual workshop environment as the construction background, constructing a three-dimensional model block based on the model shape information and the model size information, and correspondingly, the construction environment of the three-dimensional model block is based on the quality inspection area divided in the production workshop; identifying several first corner points of the three-dimensional model block, the first corner point is the endpoint at the corner on the three-dimensional model block, and selecting a second corner point from the several first corner points; establishing a three-dimensional coordinate system with the second corner point as the origin, and correspondingly, in this embodiment, the second corner point is the bottom surface of the three-dimensional model block. The endpoint of any corner on the upper part; in this embodiment, the three-dimensional coordinate system is a coordinate system based on the x, y and z axes; based on the three-dimensional coordinate system, the first axis of the three-dimensional model block is determined, and the first axis is a line connecting two mutually distant corner points in the three-dimensional model block, and the line can divide the three-dimensional model block into two mutually symmetrical three-dimensional models; the first corner point located on the first axis is set as the third corner point; in this embodiment, due to different quality inspection requirements and different shooting ranges required for quality inspection, there can be multiple first axes, and the corresponding third corner points can also be multiple, but usually there are only two first corner points located on the same first axis;

[0059] In an actual workshop, a first feature recognition unit is set according to the relevant positional relationship of the position directly above the second corner point and corresponding to the first interval distance from the second corner point; in this embodiment, the first feature recognition unit is a processing module designed based on object feature recognition technology, which is set at a position directly above the second corner point and spaced by the first interval distance, and is actually far away from the three-dimensional model block, in order to be able to fully detect the complete features of the products in the quality inspection area, and then distinguish the products; correspondingly, in an actual production workshop, the first feature recognition unit is to establish a coordinate system and a three-dimensional model block with the same proportions as the three-dimensional coordinate system and the three-dimensional model block in the quality inspection area, and then configure and install the corresponding positions according to the above positional relationship; count the first number of the third corner points; in this embodiment, the first number is at least 2; configure the first camera unit corresponding to the third corner points one by one according to the first number; correspondingly, the first camera unit uses a 3D camera module in this embodiment, and the captured image That is, a three-dimensional information image; confirm the outer edge area of ​​the three-dimensional model block, the outer edge area is the surrounding area close to / able to wrap the three-dimensional block; in the outer edge area, the first camera unit and the third corner point corresponding to the first camera unit are aligned according to the second spacing distance; correspondingly, in this embodiment, if there is only one first axis, then the first camera unit is at least two; the first feature recognition unit and a plurality of the first camera units constitute the first camera module; in this embodiment, the alignment of the first camera unit and the third corner point is to set the first camera unit in the outer edge area and on the first axis where the corresponding third corner point is located, and at the same time, it is necessary to satisfy that the distance between the first camera unit and the third corner point satisfies the second spacing distance, thereby ensuring that the first camera unit has a certain shooting range and there is no shooting blind spot; therefore, the first camera unit corresponding to the third corner point is located on the extension line of the first axis where the third corner point is located; Figure 3 As shown, the X-axis, Y-axis and Z-axis form a three-dimensional coordinate system, and the product to be inspected corresponds to a three-dimensional model block, that is, a regular cube. Figure 3 The line connecting point B and point P is the first axis. Correspondingly, point B and point P are two third corner points located on the first axis. Figure 3 The two "triangles" in the figure are the 3D camera units corresponding to the two third corner points of point B and point P respectively. After completing the above settings, the preliminary camera module and quality inspection environment are built, and intelligent quality inspection operations can be carried out later.

[0060] S200, product image capturing step, specifically comprising:

[0061] S210, obtaining quality inspection requirements, setting a startup quantity threshold, and performing an image capture operation based on the first camera module, the quality inspection requirements, and the startup quantity threshold to obtain a product image;

[0062] Specifically, the quality inspection requirement includes: a first requirement and a second requirement; the first requirement is that the first product to be inspected exists; the second requirement is that the first product to be inspected does not exist;

[0063] Specifically, the image capture operation includes: identifying the quality inspection requirement; if the quality inspection requirement is the first requirement, it means that there is a first product to be inspected in the area corresponding to the above-mentioned camera module that needs to be inspected, so the first feature recognition unit is called to obtain the first feature of the first product to be inspected, and the first feature is not limited to the image feature or information, number feature of the first product to be inspected; in this embodiment, the first feature recognition unit is provided with an AI intelligent recognition algorithm and a corresponding feature data set, and the feature data set pre-stores product image features, information, number features and corresponding product specifications and product shapes of several different products; therefore, the first feature recognition unit will analyze and identify the product according to the AI ​​intelligent recognition algorithm and the corresponding feature data set and based on the first feature. Corresponding product information, and then calculating the characteristic specifications and characteristic shapes of the first product to be inspected; in this embodiment, the characteristic specifications and characteristic shapes are the actual volume specification shape parameters corresponding to the first product to be inspected; in this embodiment, the second camera unit is confirmed in a number of the first camera units based on the characteristic shape. Because different products have different shapes, it is not necessary to call all camera units to capture and shoot each product. Sometimes the product specifications are small, the shape is simple, and there are fewer points to be detected, so many camera units are not needed, and even only one camera unit is needed to capture the corresponding product image. Therefore, based on this, in this embodiment, more camera units are initially set, so the corresponding product shape is matched according to the position of the set camera unit. For example, Figure 3As shown, the currently configured camera units are two first camera units located on the first axis, one corresponding to point P at the top and one corresponding to point B at the bottom. If the characteristic shape is a circular sheet, only the first camera unit corresponding to point P at the top needs to be called. If the characteristic shape is a regular cube or a conical triangle, then both first camera units need to be called to capture the corresponding complete image. Then, the camera unit to be called is confirmed by this logic to be the second camera unit. In this embodiment, the startup quantity threshold is 2, and the second quantity of the second camera units is counted to determine whether the second quantity is equal to the startup quantity. threshold; if matched, it means that the number of second camera units to be started in this method has reached the minimum value and cannot be reduced any further, and the second camera unit is set as the camera unit to be started; if not matched, it means that the current number of second camera units is greater than 2, so in order to save costs and resources and achieve the maximum utilization of camera units and ensure the intelligence of this method, a third camera unit is confirmed from a number of second camera units based on the feature specifications, and the third camera unit is set as the camera unit to be started; the third camera unit is a camera unit further screened out from a number of second camera units, and is confirmed to meet the product requirements through the aforementioned steps. After selecting the camera unit of the shape, further screening can be carried out according to the specifications of the products. Under normal circumstances, if the product to be inspected is small, then starting one camera unit can also meet the image capture requirements. Therefore, in this embodiment, a corresponding specification threshold is set to determine whether the feature specification has reached the specification threshold. If so, the camera unit located above will be selected from the second camera unit as the camera unit to be started. At the same time, the position parameters of the second camera unit in the three-dimensional coordinate system can also be obtained, and compared with the position parameters according to the feature specification. For example, the feature specification is 5*5*5, and there are currently three second camera units, and their corresponding position parameters are The setting parameters are (7,0,7), (0,7,7) and (7,7,0), then correspondingly, only the second camera unit with the position parameter of (7,0,7) is called to match the second camera unit with the position parameter of (0,7,7) or (0,7,7); in this method, the technical idea reflected is to identify the camera unit that can achieve the maximum shooting range and meet the minimum number of camera units in the camera module according to the shape and specifications of the product to be inspected, so as to meet the quality inspection requirements and achieve a higher effective utilization rate and intelligence of the camera unit; finally, the first capture step is performed based on the camera unit to be started to obtain the product image.

[0064] Specifically, the first capturing step includes: starting the camera unit to be started to capture a first image of the first product to be inspected. The first image is the three-dimensional image information captured by the camera unit to be started, so the first image is set as the product image.

[0065] S300, product comparison quality inspection steps, specifically including:

[0066] S310: Acquire a reference image package, and perform a product quality inspection and judgment operation based on the product image and the reference image package.

[0067] Specifically, the reference image package stores: a source image of the product to be inspected and a reference inspection point image; correspondingly, in this embodiment, the source image of the product to be inspected is a 3D model image of a qualified product corresponding to the first product to be inspected, and the reference inspection point image is an image corresponding to a position in the 3D model image where the first product to be inspected needs to undergo quality inspection;

[0068] Specifically, the product quality inspection judgment operation includes: reading the reference image package; performing three-dimensional restoration processing on the product image based on the source image of the product to be inspected to obtain a three-dimensional model image; in this embodiment, the three-dimensional restoration processing is an overall three-dimensional restoration, including but not limited to the use of three-dimensional reproduction technology; correspondingly, based on the image comparison technology, and according to the source image of the product to be inspected, it is judged whether the three-dimensional model image has image missing, that is, it is judged whether the restored three-dimensional model image has missing parts or image information compared with the source image of the product to be inspected; if there is no image missing, it means that the camera unit to be started confirmed in the above steps is correct, so it is sufficient to perform quality inspection based on the image, so based on the reference detection point image, the point to be inspected is framed in the three-dimensional model image; the point to be inspected is framed, that is, based on the image comparison technology, an image matching the features of the reference detection point image is confirmed in the three-dimensional model image. In order to improve the accuracy of quality inspection, the three-dimensional model image corresponding to the point to be inspected is enlarged to obtain an enlarged model image. In this embodiment, the enlargement also performs the same magnification processing on the reference detection point image, which is enlarged by 1 in this embodiment.5 times; compare the enlarged model image with the reference detection point image to see if they match, that is, compare whether there are parts missing or product defects in the enlarged model image and the reference detection point image. If none of the above situations exist, that is, they match, then the first detection result is output; if any of the above situations exists, that is, they do not match, then the second detection result is output; correspondingly, if there is image missing, it means that the number of camera units to be started in the above steps is insufficient or the camera unit itself has an abnormality, so it is necessary to reselect the camera unit to recapture the image, so based on the image comparison technology, the source image of the product to be detected and the 3D model image are compared to confirm the missing parts in the 3D model image. missing image; at the same time, identifying the first position information of the missing image with respect to the source image of the product to be detected, the first position information, for example, the missing image is an image of the top center of the source image of the product to be detected; therefore, based on the first position information, a backup camera unit is identified among the plurality of first camera units, and the confirmation of the backup camera unit is to confirm whether there is a camera unit close to or corresponding to the first position information in the previously selected camera unit to be started according to the position parameters in the three-dimensional coordinate system corresponding to the previously selected camera unit to be started and the first position information, for example, according to the position parameters in the three-dimensional coordinate system corresponding to the previously selected camera unit to be started Determine the orientation of the camera unit relative to the product to be inspected, and whether it is located above or below the product to be inspected; and then further confirm based on the first orientation information, for example, the missing image is an image at the top center of the source image of the product to be inspected, and the aforementioned camera unit to be started is located below the product to be inspected, and there is no camera unit located above the product to be inspected, that is, there is no camera unit close to or corresponding to the first orientation information; on the contrary, if there is, then the camera unit close to or corresponding to the first orientation information has an abnormality, and it is necessary to reconfigure a camera unit at this position as a backup camera unit; if not, it is necessary to start the first camera. The first camera unit corresponding to the first position information in the image module is used as a backup camera unit; subsequently, based on the camera unit to be started and the backup camera unit, the first capture step is returned to, that is, the three-dimensional image is recaptured, the three-dimensional image is restored, and subsequent comparison operations are performed; in this embodiment, the number of operations of returning to the first capture step is designed to be at most two times. After more than two times, the quality inspection of the product is stopped, indicating that there is a problem with the hardware or software of the first camera module of this method, and there has been image loss, so maintenance and debugging are required, and equipment alarm maintenance prompts are issued at the same time. .

[0069] Specifically, the first test result is that the product is qualified; the second test result is that the product is unqualified; specifically, the product quality inspection method based on multiple cameras also includes: configuring a manual operation detection unit and manual detection notification information, and obtaining the communication address of the manager of the first product to be inspected; in this embodiment, the manual operation detection unit is a related quality inspection module that performs detection through manual intervention; the manual detection notification information is "the intelligent detection result is that the product does not meet the standard, please conduct manual inspection for further confirmation"; when the second test result is output in the product quality inspection judgment operation, it means that the product does not meet the standard. In order to further ensure the intelligence of the method and the accuracy of the quality inspection results, it is necessary to perform manual confirmation again when confirming that the product is unqualified, thereby improving the control quality of the product leaving the factory; therefore, based on the image comparison technology, the enlarged model image that does not match the reference detection point image is set as a substandard image; identify the second position information of the substandard image with respect to the source image of the product to be inspected, and the first The recognition principle of the second position information is the same as that of the first position information; therefore, the manual inspection notification information and the second position information are sent to the communication address of the management personnel; the communication address of the management personnel is the information receiving address of the mobile terminal or PC of the management user who is responsible for controlling the corresponding quality inspection process in the production workshop; and then, by intelligently notifying the management personnel, effective and timely reconfirmation of unqualified products is achieved; at the same time, because of the existence of the second position information, the manual inspection efficiency of the management personnel can be further improved, and the management personnel do not need to look for defects or unqualified points in the product again, and the degree of intelligence is extremely high; it can be imagined that before sending the manual inspection notification information and the second position information, the workshop location information of the corresponding quality inspection area is obtained, and when sending the manual inspection notification information and the second position information, the workshop location information is sent together, thereby helping the management personnel to quickly reach the specific quality inspection area, so that this method can be applied to production workshops with multiple quality inspection areas.

[0070] Example 2

[0071] This embodiment is based on the same inventive concept as the product quality inspection method based on multiple cameras described in Embodiment 1, and provides a product quality inspection system based on multiple cameras, such as Figure 4 As shown, it includes: module configuration module, capture control module and quality inspection control module;

[0072] In the product quality inspection system based on multiple cameras, the module configuration module is used to set model parameters and configuration parameters, and configure the first camera module based on the model parameters and the configuration parameters;

[0073] Specifically, the model parameters include: model shape information and model size information; the configuration parameters include: a first spacing distance and a second spacing distance;

[0074] Specifically, the operation of configuring the first camera module based on the model parameters by the module configuration module further includes: the module configuration module reads the model parameters; the module configuration module constructs a three-dimensional model block based on the model shape information and the model size information; the module configuration module identifies a plurality of first corner points of the three-dimensional model block, and selects a second corner point from the plurality of first corner points; the module configuration module establishes a three-dimensional coordinate system with the second corner point as the origin; the module configuration module confirms the first axis of the three-dimensional model block based on the three-dimensional coordinate system; the module configuration module sets the first corner point located on the first axis as the third corner point; the module configuration module sets a first feature recognition unit at a position directly above the second corner point and corresponding to the first spacing distance from the second corner point; the module configuration module counts the first number of the third corner points; the module configuration module configures the first camera unit corresponding to the third corner point one by one according to the first number; the module configuration module confirms the outer edge area of ​​the three-dimensional model block; within the outer edge area, the module configuration module aligns the first camera unit with the third corner point corresponding to the first camera unit according to the second spacing distance.

[0075] Specifically, in this embodiment, there are at least two first camera units; the first feature recognition unit and several first camera units constitute the first camera module; the first camera unit corresponding to the third corner point is located on the extension line of the first axis where the third corner point is located.

[0076] In the product quality inspection system based on multiple cameras, the capture control module is used to obtain quality inspection requirements and set a startup quantity threshold, and the capture control module performs an image capture operation based on the first camera module, the quality inspection requirements and the startup quantity threshold to obtain a product image;

[0077] Specifically, the quality inspection requirement includes: a first requirement and a second requirement; the first requirement is that the first product to be inspected exists; the second requirement is that the first product to be inspected does not exist;

[0078] Specifically, the image capture operation includes: the capture control module identifies the quality inspection requirement; if the quality inspection requirement is the first requirement, the capture control module calls the first feature recognition unit to obtain the first feature of the first product to be inspected; the capture control module calculates the feature specification and feature shape of the first product to be inspected based on the first feature; the capture control module confirms the second camera unit among several first camera units based on the feature shape; the capture control module counts the second number of the second camera units and determines whether the second number matches the start-up number threshold; if they match, the capture control module sets the second camera unit as the camera unit to be started; if they do not match, the capture control module confirms the third camera unit among several second camera units based on the feature specification, and the capture control module sets the third camera unit as the camera unit to be started; the capture control module executes the first capture step based on the camera unit to be started to obtain the product image.

[0079] Specifically, the first capturing step includes: the capturing control module starts the to-be-started camera unit to capture a first image of the first product to be inspected, and sets the first image as the product image.

[0080] In the product quality inspection system based on multiple cameras, the quality inspection control module is used to obtain a reference image package and perform a product quality inspection judgment operation based on the product image and the reference image package;

[0081] Specifically, the reference image package stores: a source image of the product to be inspected and an image of a reference inspection point;

[0082] Specifically, the product quality inspection judgment operation includes: the quality inspection control module reads the reference image package; the quality inspection control module performs three-dimensional restoration processing on the product image based on the source image of the product to be inspected to obtain a three-dimensional model image; the quality inspection control module determines whether there is image missing in the three-dimensional model image based on the source image of the product to be inspected; if there is no image missing, the quality inspection control module selects the point to be inspected in the three-dimensional model image based on the reference inspection point image; the quality inspection control module enlarges the three-dimensional model image corresponding to the point to be inspected to obtain an enlarged model image; the quality inspection control module compares the enlarged model image with the reference inspection point image to see if they match, and if so, the quality inspection control module outputs a first inspection result; if not, the quality inspection control module outputs a second inspection result; the first inspection result is that the product is qualified; the second inspection result is that the product is unqualified;

[0083] Specifically, the product quality inspection judgment operation also includes: if there is an image missing, the quality inspection control module confirms the missing image based on the source image of the product to be inspected and the three-dimensional model image; the quality inspection control module identifies the first position information of the missing image with respect to the source image of the product to be inspected; the quality inspection control module confirms a backup camera unit among a plurality of the first camera units based on the first position information; the quality inspection control module returns to the first capture step based on the camera unit to be started and the backup camera unit;

[0084] Specifically, the system also includes the following operations: the quality inspection control module configures a manual operation detection unit and manual inspection notification information, and the quality inspection control module obtains the communication address of the manager of the first product to be quality inspected; when the quality inspection control module outputs the second detection result in the product quality inspection judgment operation, the quality inspection control module sets the enlarged model image that does not match the reference detection point image as a substandard image; the quality inspection control module identifies the second position information of the substandard image relative to the source image of the product to be inspected; the quality inspection control module sends the manual inspection notification information and the second position information to the manager's communication address.

[0085] Example 3

[0086] This embodiment provides a computer-readable storage medium, including:

[0087] The storage medium is used to store computer software instructions used to implement the product quality inspection method based on multiple cameras described in the above-mentioned embodiment 1, which includes programs for executing the above-mentioned program set for the product quality inspection method based on multiple cameras; specifically, the executable program can be built into the product quality inspection system based on multiple cameras described in embodiment 2, so that the product quality inspection system based on multiple cameras can implement the product quality inspection method based on multiple cameras described in embodiment 1 by executing the built-in executable program.

[0088] In addition, the computer-readable storage medium of this embodiment may adopt any combination of one or more computer-readable storage media, wherein the computer-readable storage medium includes electrical, optical, electromagnetic, infrared or semiconductor systems, devices or components, or any combination thereof.

[0089] Different from the existing technology, the product quality inspection method, system and medium based on multiple camera terminals of the present application can automatically identify the characteristics of the object to be inspected, and intelligently adapt the corresponding camera terminal according to the characteristics. The three-dimensional images of the products captured by multiple camera terminals are used for three-dimensional restoration and reproduction, and then the products are digitally inspected according to the restored three-dimensional images. Not only can a variety of different products be intelligently adapted, but also the efficiency and accuracy of quality inspection can be improved. The system provides effective technical support for the method, which ultimately makes up for the shortcomings of the existing technology and has extremely high application value in production workshops.

[0090] The serial numbers of the embodiments disclosed in the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0091] A person skilled in the art will appreciate that all or part of the steps for implementing the above embodiments may be accomplished by hardware, or may be accomplished by a program to instruct the relevant hardware. The program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk, or an optical disk, etc.

[0092] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A product quality inspection method based on multiple cameras, characterized in that: The following steps are involved: Camera module configuration step: setting model parameters and configuration parameters, and configuring the first camera module based on the model parameters and the configuration parameters; Product image capturing step: obtaining quality inspection requirements, setting a startup quantity threshold, performing an image capturing operation based on the first camera module, the quality inspection requirements and the startup quantity threshold, and obtaining a product image; Product comparison quality inspection step: obtaining a reference image package, and performing a product quality inspection judgment operation based on the product image and the reference image package; The model parameters include: model shape information and model size information; The configuration parameters include: a first spacing distance and a second spacing distance; The step of configuring the first camera module based on the model parameters and the configuration parameters further includes: reading the model parameters; constructing a three-dimensional model block based on the model shape information and the model size information; identifying a plurality of first corner points of the three-dimensional model block, and selecting a second corner point from the plurality of first corner points; establishing a three-dimensional coordinate system with the second corner point as the origin; confirming the first axis of the three-dimensional model block based on the three-dimensional coordinate system; setting the first corner point located on the first axis as the third corner point; setting a first feature recognition unit at a position directly above the second corner point and corresponding to the first spacing distance from the second corner point; counting the first number of the third corner points; configuring the first camera unit corresponding to the third corner point one by one according to the first number; confirming the outer edge area of ​​the three-dimensional model block; and in the outer edge area, aligning the first camera unit with the third corner point corresponding to the first camera unit according to the second spacing distance; The quality inspection requirements include: a first requirement and a second requirement; the first requirement is that the first product to be inspected exists; the second requirement is that the first product to be inspected does not exist; The image capture operation includes: identifying the quality inspection requirement; if the quality inspection requirement is the first requirement, calling the first feature recognition unit to obtain the first feature of the first product to be inspected; calculating the feature specification and feature shape of the first product to be inspected based on the first feature; confirming a second camera unit from a number of the first camera units based on the feature shape; counting a second number of the second camera units, and determining whether the second number matches the start-up number threshold; if they match, setting the second camera unit as the camera unit to be started; if they do not match, confirming a third camera unit from a number of the second camera units based on the feature specification, and setting the third camera unit as the camera unit to be started; performing the first capture step based on the camera unit to be started to obtain the product image; The reference image package stores: a source image of the product to be inspected and an image of a reference inspection point; The product quality inspection judgment operation includes: reading the reference image package; performing three-dimensional restoration processing on the product image based on the product source image to be inspected to obtain a three-dimensional model image; judging whether there is image missing in the three-dimensional model image based on the product source image to be inspected; if there is no image missing, selecting the point to be inspected in the three-dimensional model image based on the reference inspection point image; enlarging the three-dimensional model image corresponding to the point to be inspected to obtain an enlarged model image; comparing the enlarged model image with the reference inspection point image to see if they match, and if so, outputting a first detection result; if not, outputting a second detection result.

2. The product quality inspection method based on multiple cameras according to claim 1, characterized in that: There are at least two first camera units; the first feature recognition unit and a plurality of the first camera units constitute the first camera module; The first camera unit corresponding to the third corner point is located on an extension line of the first axis where the third corner point is located.

3. The product quality inspection method based on multiple cameras according to claim 1, characterized in that: The product quality inspection and judgment operation also includes: If there is an image missing, confirm the missing image based on the source image of the product to be detected and the three-dimensional model image; identify the first position information of the missing image relative to the source image of the product to be detected; confirm a backup camera unit among several first camera units based on the first position information; and return to the first capturing step based on the camera unit to be started and the backup camera unit.

4. The product quality inspection method based on multiple cameras according to claim 1, characterized in that: The first capturing step performed based on the to-be-started camera unit includes: The to-be-activated camera unit is activated to capture a first image of the first product to be inspected, and the first image is set as the product image.

5. The product quality inspection method based on multiple cameras according to claim 1, characterized in that: The first test result is that the product is qualified; the second test result is that the product is unqualified; The product quality inspection method based on multiple cameras also includes: Configure a manual operation detection unit and manual inspection notification information, and obtain the communication address of the administrator of the first product to be inspected; when the second inspection result is output in the product quality inspection judgment operation, set the enlarged model image that does not match the reference inspection point image as a non-compliant image; identify the second position information of the non-compliant image with respect to the source image of the product to be inspected; and send the manual inspection notification information and the second position information to the communication address of the administrator.

6. A product quality inspection system based on multiple cameras based on a product quality inspection method based on multiple cameras according to any one of claims 1 to 5, characterized in that: include: Module configuration module, capture control module and quality inspection control module; The module configuration module is used to set model parameters and configuration parameters, and configure the first camera module based on the model parameters and the configuration parameters; The capture control module is used to obtain quality inspection requirements and set a startup quantity threshold, and the capture control module performs an image capture operation based on the first camera module, the quality inspection requirements and the startup quantity threshold to obtain a product image; The quality inspection control module is used to obtain a reference image package and perform a product quality inspection judgment operation based on the product image and the reference image package.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, the steps of the product quality inspection method based on multiple cameras in any one of claims 1 to 5 are implemented.

Images