Vision inspection based assembly apparatus, method and program product

Abstract

Embodiments of the present disclosure disclose an assembly device, method and program product based on visual detection. A specific embodiment of the device comprises a control module, a feeding module, a conveying mechanism, a grabbing mechanism, a visual detection module and a pressing mechanism, wherein the control module is in communication connection with the feeding module, the feeding module comprises an upper cover feeding module, a lower cover feeding module and a reagent strip feeding module. The grabbing mechanism comprises an upper cover grabbing mechanism, a lower cover grabbing mechanism and a reagent strip grabbing mechanism, and the pressing mechanism is used for pressing the upper cover and the lower cover of the material transported on the conveying mechanism to realize the assembly of the upper cover and the lower cover of the material. The control module is configured to respond to the defect detection result to characterize that the assembled upper cover and lower cover of the material do not have defects, and control the conveying mechanism to convey the assembled upper cover and lower cover of the material to a storage tank. The embodiment can reduce the damage rate of the reagent strip and improve the assembly efficiency and the qualified rate of the assembled reagent strip.

Description

TECHNICAL FIELD

[0001] Embodiments of the present disclosure relate to the technical field of visual detection, and in particular, to an assembly device, method and program product based on visual detection. BACKGROUND

[0002] A reagent strip is a paper strip with dry chemical reagents, which can be used to detect chemical components in a liquid sample. Due to its convenience, reliability, low cost and other advantages, it is widely used in various technical fields such as biology and medicine. Due to the chemical properties of the reagent strip, the reagent strip cannot be in contact with air for a long time, so it needs to be assembled through a packaging box. Most of the assembly methods are manual reagent strip picking and manual defect detection.

[0003] However, the inventors have found that when the reagent strip is assembled by using the above assembly method, the following technical problems often exist:

[0004] First, for some special reagent strips, manual picking can easily cause the reagent strip to be contaminated, thereby causing the reagent strip to be damaged. For small-sized reagent strips, manual picking is not convenient, thereby easily leading to low assembly efficiency. Moreover, for small-sized reagent strips and packaging boxes, workers are easy to ignore some details of defects, thereby leading to low qualified rate of the assembled reagent strips.

[0005] Second, it is difficult to accurately distinguish small stains, thereby further leading to low qualified rate of the assembled reagent strips.

[0006] Third, it is difficult to accurately determine whether the assembled reagent strip is skewed or offset, thereby further leading to low qualified rate of the assembled reagent strips. SUMMARY

[0007] The summary of the present disclosure is used to introduce the concepts in a brief form, which will be described in detail in the specific embodiments part. The summary of the present disclosure is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions. Some embodiments of the present disclosure propose an assembly device, method and computer program product based on visual detection to solve one or more of the technical problems mentioned in the background part.

[0008] In a first aspect, some embodiments of the present disclosure provide an assembly device based on visual detection, comprising: a control module, a feeding module, a conveying mechanism, a grabbing mechanism, a visual detection module and a pressing mechanism, wherein the control module is in communication connection with the feeding module, the feeding module comprises an upper cover feeding module, a lower cover feeding module and a reagent strip feeding module; the conveying mechanism is in communication connection with the control module, the conveying mechanism is provided with at least one tray, and each tray in the at least one tray is used for carrying a material lower cover; the grabbing mechanism is in communication connection with the control module, and the grabbing mechanism is located above the conveying mechanism, the grabbing mechanism comprises an upper cover grabbing mechanism, a lower cover grabbing mechanism and a reagent strip grabbing mechanism, wherein the reagent strip grabbing mechanism is located between the upper cover grabbing mechanism and the lower cover grabbing mechanism; the control module is configured to control the lower cover grabbing mechanism to grab the material lower cover transported on the lower cover feeding module into the tray provided on the conveying mechanism, control the reagent strip grabbing mechanism to grab the reagent strip transported on the reagent strip feeding module into the material lower cover transported by the conveying mechanism, and control the upper cover grabbing mechanism to grab the material upper cover transported on the upper cover feeding module above the material lower cover carrying the reagent strip; the pressing mechanism is in communication connection with the control module, the pressing mechanism is located on one side of the upper cover grabbing mechanism, and the pressing mechanism is used for pressing the material upper cover and the material lower cover transported on the conveying mechanism to realize the assembly of the material upper cover and the material lower cover; the visual detection module is in communication connection with the control module, and the visual detection module comprises an assembly detection module, wherein the control module is further configured to detect defects of the assembled material upper cover and the material lower cover by the assembly detection module to obtain a defect detection result; and the control module is further configured to control the conveying mechanism to convey the assembled material upper cover and the material lower cover to a storage box in response to the defect detection result representing that the assembled material upper cover and the material lower cover do not have defects.

[0009] Optionally, the assembly detection module comprises an assembly camera and a contrast assembly camera, the contrast assembly camera is adjacent to the assembly camera, and a signal-to-noise ratio corresponding to the contrast assembly camera is different from a signal-to-noise ratio corresponding to the assembly camera; the control module is further configured to: capture the assembled material upper cover and the material lower cover by using the assembly camera to obtain an assembly box image; identify a stain area in the assembly box image to obtain at least one stain area; cut the assembly box image according to the at least one stain area to obtain at least one stain image; determine an image feature corresponding to each stain image in the at least one stain image to obtain a stain image feature; capture the assembled material upper cover and the material lower cover by using the contrast assembly camera to obtain a contrast assembly box image; identify a stain area in the contrast assembly box image to obtain at least one contrast stain area; cut the contrast assembly box image according to the at least one contrast stain area to obtain at least one contrast stain image; determine an image feature corresponding to each contrast stain image in the at least one contrast stain image to obtain a contrast stain image feature; determine a stain image, in which a similarity between a stain image feature corresponding to the stain image and a contrast stain image feature is greater than a preset similarity threshold, as a target stain image; determine a sum of areas of the target stain images as a target stain area; and generate a defect detection result according to the target stain area.

[0010] Optionally, the assembly detection module comprises a light sensor and a light source. The light sensor and the light source are both arranged on the assembly detection module. The control module is further configured to: determine an ambient light intensity by using the light sensor; determine a distance between the light source and a tray located below the light source; determine a target light intensity corresponding to the light source according to the ambient light intensity and the distance; and adjust a light intensity corresponding to the light source according to the target light intensity.

[0011] In a second aspect, some embodiments of the present disclosure provide an assembly method based on visual detection, applied to the assembly device based on visual detection as described in the first aspect, wherein the assembly device comprises a control module, a feeding module, a conveying mechanism, a grabbing mechanism, a visual detection module and a pressing mechanism, the feeding module comprises an upper cover feeding module, a lower cover feeding module and a reagent strip feeding module, the grabbing mechanism comprises an upper cover grabbing mechanism, a lower cover grabbing mechanism and a reagent strip grabbing mechanism, and the method comprises: grabbing the material lower cover conveyed on the lower cover feeding module by the lower cover grabbing mechanism to the inside of the tray arranged on the conveying mechanism; grabbing the reagent strip conveyed on the reagent strip feeding module by the reagent strip grabbing mechanism to the inside of the material lower cover conveyed by the conveying mechanism; grabbing the material upper cover conveyed on the upper cover feeding module by the upper cover grabbing mechanism to the upper side of the material lower cover carrying the reagent strip; pressing the material upper cover and the material lower cover conveyed on the conveying mechanism by the pressing mechanism to realize the assembly of the material upper cover and the material lower cover; defect detecting the assembled material upper cover and the material lower cover by the assembly detection module to obtain a defect detection result; and in response to the defect detection result representing that the assembled material upper cover and the material lower cover have no defects, controlling the conveying mechanism to convey the assembled material upper cover and the material lower cover to a storage box.

[0012] In a third aspect, some embodiments of the present disclosure provide a computer program product comprising a computer program which, when executed by a processor, implements the method described in any implementation manner of the second aspect.

[0013] The various embodiments of the present disclosure have the following beneficial effects: the assembly device based on visual detection of some embodiments of the present disclosure can reduce the damage rate of reagent strips and improve the assembly efficiency and the qualified rate of the assembled reagent strips. Specifically, the reason why the related assembly method causes the reagent strip damage and the low assembly efficiency and qualified rate of the assembled reagent strips is that: for some special reagent strips, manual picking is easy to cause the reagent strip pollution, and for the reagent strips with small size, manual picking is inconvenient, and the workers are easy to ignore some details of the pollution. Based on this, the assembly device based on visual detection of some embodiments of the present disclosure comprises a control module, a feeding module, a conveying mechanism, a grabbing mechanism, a visual detection module and a pressing mechanism, wherein the control module is in communication connection with the feeding module, and the feeding module comprises an upper cover feeding module, a lower cover feeding module and a reagent strip feeding module. The conveying mechanism is in communication connection with the control module, and at least one tray is arranged on the conveying mechanism, and each tray in the at least one tray is used for carrying a material lower cover. The grabbing mechanism is in communication connection with the control module, and the grabbing mechanism is located above the conveying mechanism, and the grabbing mechanism comprises an upper cover grabbing mechanism, a lower cover grabbing mechanism and a reagent strip grabbing mechanism, wherein the reagent strip grabbing mechanism is located between the upper cover grabbing mechanism and the lower cover grabbing mechanism. The control module is configured to control the lower cover grabbing mechanism to grab the material lower cover transported on the lower cover feeding module into the tray arranged on the conveying mechanism, control the reagent strip grabbing mechanism to grab the reagent strip transported on the reagent strip feeding module into the material lower cover transported by the conveying mechanism, and control the upper cover grabbing mechanism to grab the material upper cover transported on the upper cover feeding module above the material lower cover carrying the reagent strip. The pressing mechanism is in communication connection with the control module, and the pressing mechanism is located on one side of the upper cover grabbing mechanism, and the pressing mechanism is used for pressing the material upper cover and the material lower cover transported on the conveying mechanism to realize the assembly of the material upper cover and the material lower cover. The visual detection module is in communication connection with the control module, and the visual detection module comprises an assembly detection module, wherein the control module is further configured to perform defect detection on the assembled material upper cover and the material lower cover by the assembly detection module to obtain a defect detection result. The control module is further configured to control the conveying mechanism to convey the assembled material upper cover and the material lower cover to a storage box in response to the defect detection result representing that the assembled material upper cover and the material lower cover do not have defects. Because the assembly device grabs the reagent strip through the reagent strip grabbing mechanism, manual contact with the reagent strip can be avoided, and the damage rate of the reagent strip is reduced. Also, the assembled reagent strip can be subjected to defect detection through the assembly detection module, and only the assembled reagent strip without defects is transmitted, so that the qualified rate of the assembled reagent strip can be improved.Therefore, the assembly device based on visual detection provided by some embodiments of the present disclosure can reduce the damage rate of reagent strips and improve the assembly efficiency and the qualified rate of the assembled reagent strips. BRIEF DESCRIPTION OF DRAWINGS

[0014] The above and other features, advantages and aspects of embodiments of the present disclosure will become more apparent upon reading the following detailed description in conjunction with the accompanying drawings, in which like references refer to like elements. Throughout the drawings, the same or similar elements are denoted by the same or similar reference signs. It should be understood that the drawings are schematic and elements and features are not necessarily to scale.

[0015] Figure 1 is a structural schematic diagram of some embodiments of the assembly device based on visual detection according to the present disclosure;

[0016] Figure 2 is a flowchart of some embodiments of the assembly method based on visual detection according to the present disclosure. DETAILED DESCRIPTION

[0017] Embodiments of the present disclosure will be described in detail below with reference to the drawings. Although some embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms, and should not be interpreted as being limited to the embodiments set forth herein. On the contrary, these embodiments are provided to make the present disclosure more thorough and complete. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes, and are not intended to limit the scope of protection of the present disclosure.

[0018] It should also be noted that, for the convenience of description, only the parts related to the present disclosure are shown in the drawings. The embodiments in the present disclosure and the features in the embodiments can be combined with each other without conflict.

[0019] It should be noted that the terms “first”, “second”, and the like mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not intended to limit the order or interdependence of the functions performed by these devices, modules or units.

[0020] It should be noted that the terms “one”, “multiple” mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise explicitly stated in the context, it should be understood as “one or more”.

[0021] The names of the messages or information exchanged between the devices in the embodiments of the present disclosure are only for illustrative purposes, and are not intended to limit the scope of the messages or information.

[0022] The present disclosure will be described in detail below with reference to the drawings and in conjunction with the embodiments.

[0023] The present disclosure will be described in detail below with reference to the drawings and in conjunction with the embodiments.Figure 1 These are schematic diagrams of some embodiments of the vision-based inspection assembly equipment disclosed herein. Figure 1 It includes an upper cover feeding module 101, a reagent strip feeding module 102, a lower cover feeding module 103, an upper cover gripping mechanism 201, a reagent strip gripping mechanism 202, a lower cover gripping mechanism 203, a conveying mechanism 3, a vision inspection module 4, and a pressing mechanism 5.

[0024] In some embodiments, the vision-based assembly equipment described above may include a control module, a feeding module, a conveying mechanism 3, a gripping mechanism, a vision inspection module 4, and a pressing mechanism 5. The control module may be communicatively connected to the feeding module. The feeding module may include an upper cover feeding module 101, a lower cover feeding module 103, and a reagent strip feeding module 102. The control module may be a microchip that controls other mechanisms and modules included in the assembly equipment. For example, the control module may include, but is not limited to, at least one of the following: PLC (Programmable Logic Controller), SoC (System on Chip), MCU (Microcontroller Unit), or DSP (Digital Signal Processor). The feeding module may be a conveyor belt for feeding the upper cover, lower cover, and reagent strips. The upper cover may be the upper cover of a packaging box for packaging the reagent strips. The lower cover may be the lower cover of the packaging box for packaging the reagent strips. The conveying module may be a chain conveyor belt for conveying the upper cover, lower cover, and reagent strips. The aforementioned gripping mechanism can be used to grip the material top cover, material bottom cover, and reagent strip. For example, the gripping mechanism can be a robotic arm. The aforementioned vision inspection module 4 can be used to inspect the surface of the material. The aforementioned pressing mechanism 5 can be used to press the material top cover and material bottom cover together. For example, the aforementioned pressing mechanism 5 can include a power component and a pressing plate. The pressing plate can be disposed at the lower end of the aforementioned power component. The aforementioned pressing plate can be provided with a groove that fits into the material top cover. The aforementioned power component can be at least one of the following: an electric cylinder, an electromagnetic push rod. The aforementioned power component can drive the aforementioned pressing plate to press the material top cover and material bottom cover together. The aforementioned top cover feeding module 101 can be a conveyor belt for feeding the material top cover. The aforementioned bottom cover feeding module 103 can be a conveyor belt for feeding the material bottom cover. The aforementioned reagent strip feeding module 102 can be a conveyor belt for feeding the reagent strip.

[0025] In some embodiments, the conveying mechanism 3 can be in communication with the control module. The conveying mechanism 3 can be provided with at least one tray. Each of the at least one tray can be used to carry the lower cover of the material. The tray can be a jig for carrying the lower cover of the material. The tray can be provided with at least one clamping groove. The clamping groove is used to carry the lower cover of the material.

[0026] In some embodiments, the gripping mechanism can be in communication with the control module. The gripping mechanism can be located above the conveying mechanism 3. The gripping mechanism can include an upper cover gripping mechanism 201, a lower cover gripping mechanism 203, and a reagent strip gripping mechanism 202. The reagent strip gripping mechanism 202 can be located between the upper cover gripping mechanism 201 and the lower cover gripping mechanism 203. The upper cover gripping mechanism 201 can be a mechanism for gripping the upper cover of the material. The lower cover gripping mechanism 203 can be a mechanism for gripping the lower cover of the material. The reagent strip gripping mechanism 202 can be a mechanism for gripping the reagent strip. For example, the upper cover gripping mechanism 201, the lower cover gripping mechanism 203, and the reagent strip gripping mechanism 202 can all be mechanical hands.

[0027] In some embodiments, the control module can be configured to control the lower cover gripping mechanism 203 to grip the lower cover of the material transported on the lower cover feeding module 103 to the inside of the tray provided on the conveying mechanism 3. The control module can also be configured to control the reagent strip gripping mechanism 202 to grip the reagent strip transported on the reagent strip feeding module 102 to the inside of the lower cover of the material transported by the conveying mechanism 3. The control module can also be configured to control the upper cover gripping mechanism 201 to grip the upper cover of the material transported on the upper cover feeding module 101 above the lower cover of the material carrying the reagent strip. In practice, the control module can control the lower cover gripping mechanism 203 to grip the lower cover of the material transported on the lower cover feeding module 103 to the inside of the tray provided on the conveying mechanism 3 according to the pre-set time and distance of movement of the lower cover gripping mechanism 203. The control module can also control the reagent strip gripping mechanism 202 to grip the reagent strip transported on the reagent strip feeding module 102 to the inside of the lower cover of the material transported by the conveying mechanism 3 according to the pre-set time and distance of movement of the reagent strip gripping mechanism 202. The control module can also control the upper cover gripping mechanism 201 to grip the upper cover of the material transported on the upper cover feeding module 101 above the lower cover of the material carrying the reagent strip according to the pre-set time and distance of movement of the upper cover gripping mechanism 201.

[0028] In some embodiments, the pressing mechanism 5 can be communicatively connected to the control module, and the pressing mechanism 5 can be located on one side of the upper cover gripping mechanism 201. Specifically, the pressing mechanism 5 can be located on the side of the upper cover gripping mechanism 201 that is not adjacent to the reagent strip gripping mechanism 202. The pressing mechanism 5 can be used to press the upper cover and lower cover of the material transported on the conveying mechanism 3 together to achieve the assembly of the upper cover and lower cover.

[0029] In some embodiments, the vision inspection module 4 can be communicatively connected to the control module, and the vision inspection module 4 may include an assembly inspection module. The assembly inspection module can be used to detect whether there are defects in the assembled material top cover and material bottom cover. The control module can also be configured to perform defect detection on the assembled material top cover and material bottom cover through the assembly inspection module, obtaining defect detection results. The defect detection results can characterize whether there are defects in the assembled material top cover and material bottom cover. The defect can be a stain on the assembled material top cover and material bottom cover with an area greater than or equal to a preset area threshold. In practice, the control module can utilize various methods to perform defect detection on the assembled material top cover and material bottom cover through the assembly inspection module, obtaining defect detection results.

[0030] In some embodiments, the control module may also be configured to, in response to the defect detection result indicating that the assembled material top cover and material bottom cover are free of defects, control the conveying mechanism 3 to convey the assembled material top cover and material bottom cover to a storage bin. The storage bin may be a box for holding the assembled material top cover and material bottom cover free of defects. Further, in response to the defect detection result indicating that the assembled material top cover and material bottom cover have defects, the control module may control the bottom cover gripping mechanism 203 to grip the assembled material top cover and material bottom cover to a waste bin. The waste bin may be a box for holding the assembled material top cover and material bottom cover with defects.

[0031] Optionally, the upper cover grabbing mechanism 201, the lower cover grabbing mechanism 203 and the reagent strip grabbing mechanism 202 can each include a mechanical hand, a cylinder and a pneumatic nozzle assembly. The cylinder can be an electric cylinder for providing negative pressure. The pneumatic nozzle can be a suction nozzle for adsorbing materials. The pneumatic nozzle assembly can include at least one pneumatic nozzle. The mechanical hand can be detachably connected to the corresponding pneumatic nozzle assembly. The cylinder can be used to provide negative pressure for the corresponding pneumatic nozzle assembly. Thus, the pneumatic nozzle assembly can be moved by the mechanical hand, the material upper cover, the material lower cover and the reagent strip can be adsorbed by the pneumatic nozzle assembly, and multiple materials can be adsorbed at the same time by providing multiple pneumatic nozzles.

[0032] Optionally, the visual detection module 4 can further include a reagent strip position detection module. The reagent strip position detection module can be a module for detecting the position information of the reagent strip carried on the material lower cover. The control module can be further configured to detect the pose of the reagent strip carried on the material lower cover through the reagent strip position detection module to obtain a pose detection result. The pose detection result can be a result indicating whether the reagent strip skew angle and the offset of the reagent strip carried on the material lower cover meet a preset pose condition. For example, when the reagent strip skew angle and the offset of the reagent strip carried on the material lower cover meet the preset pose condition, the pose detection result can be “reagent strip position qualified”. When the reagent strip skew angle and the offset of the reagent strip carried on the material lower cover do not meet the preset pose condition, the pose detection result can be “reagent strip position unqualified”. The preset pose condition can be that the reagent strip skew angle is less than or equal to a preset skew angle threshold and / or the offset is less than or equal to a preset offset threshold.

[0033] Optionally, the reagent strip position detection module can comprise a reagent strip camera. The reagent strip camera can be configured to capture an image of the material lower cover carrying the reagent strip. The control module can be further configured to: first, capture the material lower cover carrying the reagent strip to obtain a material lower cover image; second, determine a reagent strip region in the material lower cover image. The reagent strip region can be a region representing the reagent strip. For example, the reagent strip region can be a coordinate pair representing the reagent strip. In practice, the control module can determine the reagent strip region in the material lower cover image in various ways. Third, detect the skew angle of the reagent strip region to obtain reagent strip skew angle information. The reagent strip skew angle information can comprise a reagent strip skew angle. The reagent strip skew angle information can be information representing the reagent strip skew angle. For example, the reagent strip skew angle information can be "skew 5°". In practice, the control module can determine the angle between the center line of the reagent strip region and the longitudinal or transverse coordinate axis of the pre-set planar coordinate system to obtain the reagent strip skew angle information. Fourth, determine the offset information between the reagent strip region and a target region. The offset information can comprise an offset. The offset information can be information representing the distance between the center of the reagent strip region and the center of the target region. The target region can be a region on the material upper cover for carrying the reagent strip. Finally, generate the pose detection result according to the reagent strip skew angle information and the offset information. In practice, in response to the reagent strip skew angle information and the offset information satisfying a pre-set pose condition, the control module can generate a pose detection result representing that the reagent strip skew angle information and the offset information corresponding to the reagent strip satisfy the pre-set pose condition. In response to the reagent strip skew angle information and the offset information not satisfying the pre-set pose condition, the control module can generate a pose detection result representing that the reagent strip skew angle and the offset corresponding to the reagent strip do not satisfy the pre-set pose condition.

[0034] Optionally, the control module can be further configured to, in response to the pose detection result not satisfying the preset pose condition, control the reagent strip grabbing mechanism 202 to adjust the position of the reagent strip in the material lower cover based on the reagent strip skew angle information and the offset information. In practice, first, the control module can generate an adjustment path of the reagent strip grabbing mechanism 202 based on the reagent strip skew angle information and the offset information. The adjustment path can include a moving direction and a moving distance. As an example, the control module can input the reagent strip skew angle information and the offset information into a pre-trained adjustment path generation model to obtain the adjustment path of the reagent strip grabbing mechanism 202. The adjustment path generation model can be a machine learning model that takes reagent strip skew angle information and offset information as input and outputs an adjustment path. The control module can control the reagent strip grabbing mechanism 202 to move based on the adjustment path to adjust the position of the reagent strip in the material lower cover. In this way, when the reagent strip in the material lower cover is skewed or offset, the reagent strip grabbing mechanism 202 can be controlled to adjust the position of the reagent strip in the material lower cover.

[0035] Optionally, the control module can be further configured to: first, capture the material lower cover carried by the tray arranged on the conveying mechanism 3 by the reagent strip camera to obtain a material lower cover image. Then, perform position recognition on the material lower cover image to obtain material lower cover position information. The material lower cover position information can be information representing the position of the material lower cover. For example, the material lower cover position information can be a coordinate pair representing the position of the material lower cover. In practice, the control module can recognize the position of the material lower cover image in various ways to obtain the material lower cover position information. Finally, based on the material lower cover position information, the control module controls the reagent strip grabbing mechanism 202 to grab the reagent strip transported on the reagent strip feeding module 102 into the interior of the material lower cover conveyed by the conveying mechanism 3. In practice, the control module can input the material lower cover position information into a pre-trained grabbing path generation model to obtain a grabbing path of the reagent strip grabbing mechanism 202. The grabbing path generation model can be a machine learning model that takes a material lower cover image as input and outputs a grabbing path of the reagent strip grabbing mechanism 202. In this way, the reagent strip can be accurately placed in the material lower cover.

[0036] Optionally, the visual detection module 4 can further comprise a lower cover position detection module. The lower cover position detection module can be a visual detection module 4 for detecting the position of the tray. The lower cover position detection module can comprise a lower cover camera. The lower cover camera can be a camera for capturing images of the tray. The control module can be further configured to: first, capture an image of the tray located below the lower cover grabbing mechanism 203 on the conveying mechanism 3 by the lower cover camera, to obtain a tray image. Second, determine tray position information corresponding to the tray image according to the tray image. The tray position information can be information representing the position of the tray. For example, the tray position information can be a pair of coordinates representing the position of the tray. In practice, the control module can identify the position of the tray image in various ways to obtain the tray position information. According to the tray position information, the control module controls the lower cover grabbing mechanism 203 to grab the material lower cover transported on the lower cover feeding module 103 into the tray on the conveying mechanism 3. In practice, the control module can input the tray position information into a pre-trained material lower cover grabbing path generation model to obtain a material lower cover grabbing path. The material lower cover grabbing path generation model can be a machine learning model that takes a tray image as input and outputs a material lower cover grabbing path. Thus, the material lower cover can be accurately placed in the tray.

[0037] Optionally, the assembly detection module can further comprise an assembly camera. The assembly camera can be a camera for capturing images of the assembled material upper cover and material lower cover. The control module can be further configured to: first, capture an image of the assembled material upper cover and material lower cover by the assembly camera, to obtain an assembly box image. Then, identify stains in the assembly box image to obtain at least one stain area. The stain area can be an area representing a stain on the assembled material upper cover and material lower cover. For example, the stain area can be a pair of coordinates. In practice, the control module can input the assembly box image into a pre-trained stain identification model to obtain the stain area. The stain identification model can be a machine learning model that takes an assembly box image as input and outputs a stain area. Then, determine the area of each stain area in the at least one stain area. Next, determine the sum of the areas of each stain area as the stain area. Finally, generate a defect detection result according to the stain area. The defect detection result can be information representing whether the assembled material upper cover and material lower cover have a stain with an area greater than a preset stain area. In practice, in response to the area of the stain area being greater than the preset stain area, a defect detection result is generated indicating that the assembled material upper cover and material lower cover have a stain with an area greater than the preset stain area.

[0038] Optionally, the assembly detection module can include an assembly camera and a contrast assembly camera. The contrast assembly camera can be a camera configured to capture images of the assembled material upper cover and the material lower cover. The contrast assembly camera can be adjacent to the assembly camera. The contrast assembly camera can have a different signal-to-noise ratio than the assembly camera. The control module can perform the following steps:

[0039] In a first step, the assembly camera captures images of the assembled material upper cover and the material lower cover to obtain an assembly box image.

[0040] In a second step, the assembly box image is identified to obtain at least one stain area. The stain area can be a region in the assembly box image representing a stain on the assembled material upper cover and the material lower cover. For example, the stain area can be a coordinate pair. In practice, the control module can identify the stain area in the assembly box image in various ways.

[0041] In a third step, the assembly box image is cropped according to the at least one stain area to obtain at least one stain image. The stain image can be an image corresponding to the stain area in the assembly box image.

[0042] In a fourth step, the image features corresponding to each stain image in the at least one stain image are determined to obtain stain image features. The stain image features can be image features representing stains. For example, the stain image features can be an image feature vector of the stain. In practice, the control module can determine the image features corresponding to the stain image by various image feature extraction algorithms to obtain the stain image features. As an example, the control module can determine the image features corresponding to the stain image by the SIFT (Scale-Invariant Feature Transform) algorithm to obtain the stain image features.

[0043] In a fifth step, the contrast assembly camera captures images of the assembled material upper cover and the material lower cover to obtain a contrast assembly box image.

[0044] In a sixth step, the contrast assembly box image is identified to obtain at least one contrast stain area. The contrast stain area can be a region in the contrast assembly box image representing a stain on the assembled material upper cover and the material lower cover. For example, the contrast stain area can be a coordinate pair. In practice, the control module can identify the contrast stain area in the assembly box image in various ways.

[0045] In the seventh step, the comparison assembly box image is cropped according to the at least one comparison stain area to obtain at least one comparison stain image. The comparison stain image can be an image corresponding to the comparison stain area in the comparison assembly box image.

[0046] In the eighth step, an image feature corresponding to each comparison stain image in the at least one comparison stain image is determined to obtain a comparison stain image feature. The comparison stain image feature can be an image feature representing a stain. For example, the comparison stain image feature can be an image feature vector of the stain. In practice, the control module can determine the image feature corresponding to the comparison stain image by various image feature extraction algorithms to obtain the comparison stain image feature. As an example, the control module can determine the image feature corresponding to the comparison stain image by the SIFT algorithm to obtain the comparison stain image feature.

[0047] In the ninth step, a stain image corresponding to a comparison stain image feature with a similarity greater than a preset similarity threshold in the at least one stain image is determined as a target stain image. The preset similarity threshold can be a preset similarity threshold representing that a stain image corresponding to a similarity greater than the threshold is a target stain image. In practice, the control module can first determine the similarity between each stain image feature and each comparison stain image feature to obtain a stain image feature similarity group. As an example, the control module can determine the similarity between each stain image feature and each comparison stain image feature by various image similarity algorithms. Finally, a stain image corresponding to an image feature similarity greater than the preset similarity threshold in each obtained stain image feature similarity group is determined as a target stain image.

[0048] In the tenth step, a sum of the areas of the determined target stain images is determined as a target stain area. In practice, first, the control module can determine the area of each target stain image to obtain a stain area, and then the control module can determine the sum of each obtained stain area as the target stain area.

[0049] In the eleventh step, a defect detection result is generated according to the target stain area. In practice, in response to the area of the target stain area being greater than a preset stain area, the control module can generate a defect detection result representing that the assembled material upper cover and the material lower cover have stains with a corresponding area greater than the preset stain area.

[0050] The optional content above is one of the invention points of the embodiments of the present disclosure, and solves the second technical problem mentioned in the background that "smaller stains cannot be accurately distinguished, further resulting in a lower qualified rate of the assembled reagent strip." The reason for further resulting in a lower qualified rate of the assembled reagent strip is that smaller stains cannot be accurately distinguished. If the above factors are solved, the qualified rate of the assembled reagent strip can be further improved. In order to achieve this effect, the above assembly detection module of the present disclosure includes an assembly camera and a contrast assembly camera, wherein the contrast assembly camera is adjacent to the assembly camera, and the signal-to-noise ratio corresponding to the contrast assembly camera is different from the signal-to-noise ratio corresponding to the assembly camera; the control module is further configured to: capture the assembled material upper cover and the material lower cover by the assembly camera to obtain an assembly box image; perform stain identification on the assembly box image to obtain at least one stain area; cut the assembly box image according to the at least one stain area to obtain at least one stain image; determine the image features corresponding to each stain image in the at least one stain image to obtain stain image features; capture the assembled material upper cover and the material lower cover by the contrast assembly camera to obtain a contrast assembly box image; perform stain identification on the contrast assembly box image to obtain at least one contrast stain area; cut the contrast assembly box image according to the at least one contrast stain area to obtain at least one contrast stain image; determine the image features corresponding to each contrast stain image in the at least one contrast stain image to obtain contrast stain image features; determine the stain image in which the similarity of the corresponding stain image features and the contrast stain image features in the at least one stain image is greater than a preset similarity threshold as a target stain image; determine the sum of the areas of the determined target stain images as a target stain area; and generate a defect detection result according to the target stain area.The assembly detection module comprises an assembly camera and a contrast assembly camera. The contrast assembly camera is adjacent to the assembly camera, and the signal-to-noise ratio corresponding to the contrast assembly camera is different from the signal-to-noise ratio corresponding to the assembly camera. The control module is further configured to: capture the assembled upper cover and lower cover of the material by the assembly camera to obtain an assembly box image; identify the stains in the assembly box image to obtain at least one stain area; cut the assembly box image according to the at least one stain area to obtain at least one stain image; determine the image features corresponding to each stain image in the at least one stain image to obtain stain image features; capture the assembled upper cover and lower cover of the material by the contrast assembly camera to obtain a contrast assembly box image; identify the stains in the contrast assembly box image to obtain at least one contrast stain area; cut the contrast assembly box image according to the at least one contrast stain area to obtain at least one contrast stain image; determine the image features corresponding to each contrast stain image in the at least one contrast stain image to obtain contrast stain image features; determine the stain image in which the similarity between the stain image features and the contrast stain image features is greater than a preset similarity threshold as a target stain image; determine the sum of the areas of the target stain images as a target stain area; and generate a defect detection result according to the target stain area. In this way, the assembled upper cover and lower cover of the material are captured by two cameras with different signal-to-noise ratios at the same time, and the stain features corresponding to the two groups of images are compared, so that the stain image with high similarity can be determined as the target stain, the accuracy of identifying the stain is improved, and the pass rate of the assembled reagent strip is further improved.

[0051] Optionally, the assembly detection module can comprise a light sensor and a light source. The light sensor and the light source can be arranged on the assembly detection module.

[0052] Optionally, the control module can be further configured to:

[0053] In the first step, the ambient light intensity is determined by the light sensor. The ambient light intensity can be the light intensity in the environment corresponding to the assembly device.

[0054] In the second step, the distance between the light source and the tray located below the light source is determined. In practice, the control module can determine the distance between the light source and the tray located below the light source by the associated distance measuring sensor.

[0055] A third step is to determine a target light intensity corresponding to the light source according to the ambient light intensity and the distance. The target light intensity can be the light intensity set by the light source to make the obtained assembly box image clearest. In practice, the target light intensity can be obtained by analyzing the distance between the light source and the tray located below the light source and the ambient light intensity by an artificial intelligence chip included in the control module. The machine learning model carried by the artificial intelligence chip is trained by a training sample set. The training sample set includes the distance between the sample light source and the sample reagent strip camera device, the sample ambient light intensity, and the sample light intensity. The machine learning model is trained by taking the distance between the sample light source and the tray located below the sample light source and the sample ambient light intensity as input and taking the sample light intensity as expected output.

[0056] A fourth step is to adjust the light intensity corresponding to the light source according to the target light intensity. In practice, the control module can adjust the light intensity of the light source to the target light intensity.

[0057] The optional content as an invention point of an embodiment of the present disclosure solves the third technical problem mentioned in the background that the assembled reagent strip cannot be accurately determined to be skewed or offset, which further leads to a lower qualified rate of the assembled reagent strip. The reason for further leading to a lower qualified rate of the assembled reagent strip is as follows: the assembled reagent strip cannot be accurately determined to be skewed or offset. If the above factors are solved, the qualified rate of the assembled reagent strip can be further improved. In order to achieve this effect, the assembly detection module of the present disclosure includes a light sensor and a light source. The light sensor and the light source are both arranged on the assembly detection module. The control module is further configured to determine the ambient light intensity by the light sensor; determine the distance between the light source and the tray located below the light source. According to the ambient light intensity and the distance, the target light intensity corresponding to the light source is determined; and the light intensity corresponding to the light source is adjusted according to the target light intensity. Thus, by determining the light intensity of the light source through the distance between the light source and the tray located below the light source and the ambient light intensity, the clarity of the obtained image can be improved, so that the assembled reagent strip can be accurately determined to be skewed or offset, and the qualified rate of the assembled reagent strip can be further improved.

[0058] The various embodiments of the present disclosure have the following beneficial effects: the assembly device based on visual detection of some embodiments of the present disclosure can reduce the damage rate of reagent strips and improve the assembly efficiency and the qualified rate of the assembled reagent strips. Specifically, the reason why the related assembly method causes the reagent strip damage and the low assembly efficiency and qualified rate of the assembled reagent strips is that: for some special reagent strips, manual picking is easy to cause the reagent strip pollution, and for the small-sized reagent strips, manual picking is inconvenient, and the staff is easy to ignore some details of the pollution. Based on this, the assembly device based on visual detection of some embodiments of the present disclosure comprises a control module, a feeding module, a conveying mechanism, a grabbing mechanism, a visual detection module and a pressing mechanism, wherein the control module is in communication connection with the feeding module, and the feeding module comprises an upper cover feeding module, a lower cover feeding module and a reagent strip feeding module. The conveying mechanism is in communication connection with the control module, and at least one tray is arranged on the conveying mechanism, and each tray in the at least one tray is used for carrying a material lower cover. The grabbing mechanism is in communication connection with the control module, and the grabbing mechanism is located above the conveying mechanism, and the grabbing mechanism comprises an upper cover grabbing mechanism, a lower cover grabbing mechanism and a reagent strip grabbing mechanism, wherein the reagent strip grabbing mechanism is located between the upper cover grabbing mechanism and the lower cover grabbing mechanism. The control module is configured to control the lower cover grabbing mechanism to grab the material lower cover transported on the lower cover feeding module into the tray arranged on the conveying mechanism, control the reagent strip grabbing mechanism to grab the reagent strip transported on the reagent strip feeding module into the material lower cover transported by the conveying mechanism, and control the upper cover grabbing mechanism to grab the material upper cover transported on the upper cover feeding module above the material lower cover carrying the reagent strip. The pressing mechanism is in communication connection with the control module, and the pressing mechanism is located on one side of the upper cover grabbing mechanism, and the pressing mechanism is used for pressing the material upper cover and the material lower cover transported on the conveying mechanism to realize the assembly of the material upper cover and the material lower cover. The visual detection module is in communication connection with the control module, and the visual detection module comprises an assembly detection module, wherein the control module is further configured to perform defect detection on the assembled material upper cover and the material lower cover by the assembly detection module to obtain a defect detection result. The control module is further configured to control the conveying mechanism to convey the assembled material upper cover and the material lower cover to a storage box in response to the defect detection result representing that the assembled material upper cover and the material lower cover do not have defects. Because the assembly device grabs the reagent strip through the reagent strip grabbing mechanism, manual contact with the reagent strip can be avoided, and the damage rate of the reagent strip is reduced. Also, the assembled reagent strip can be subjected to defect detection through the assembly detection module, and only the assembled reagent strip without defects is transmitted, thereby the qualified rate of the assembled reagent strip can be improved.Thus, the vision detection based assembling device of some embodiments of the present disclosure can reduce the damage rate of reagent strips and improve the assembling efficiency and the qualified rate of assembled reagent strips.

[0059] With reference to the accompanying drawings, the vision detection based assembling method of some embodiments of the present disclosure is illustrated. Figure 2 The vision detection based assembling method includes the following steps:

[0060] In step 201, the lower cover transported on the lower cover feeding module is grabbed by the lower cover grabbing mechanism to the inside of the tray arranged on the conveying mechanism.

[0061] In some embodiments, the execution subject of the vision detection based assembling method (for example Figure 1 The vision detection based assembling device shown can grab the lower cover transported on the lower cover feeding module by the lower cover grabbing mechanism to the inside of the tray arranged on the conveying mechanism. The assembling device includes a control module, a feeding module, a conveying mechanism, a grabbing mechanism, a vision detection module and a pressing mechanism. The feeding module includes an upper cover feeding module, a lower cover feeding module and a reagent strip feeding module. The grabbing mechanism includes an upper cover grabbing mechanism, a lower cover grabbing mechanism and a reagent strip grabbing mechanism.

[0062] In step 202, the reagent strip transported on the reagent strip feeding module is grabbed by the reagent strip grabbing mechanism to the inside of the lower cover transported by the conveying mechanism.

[0063] In some embodiments, the execution subject can grab the reagent strip transported on the reagent strip feeding module by the reagent strip grabbing mechanism to the inside of the lower cover transported by the conveying mechanism.

[0064] In step 203, the upper cover transported on the upper cover feeding module is grabbed by the upper cover grabbing mechanism above the lower cover carrying the reagent strip.

[0065] In some embodiments, the execution subject can grab the upper cover transported on the upper cover feeding module by the upper cover grabbing mechanism above the lower cover carrying the reagent strip.

[0066] In step 204, the upper cover and the lower cover transported on the conveying mechanism are pressed by the pressing mechanism to realize the assembly of the upper cover and the lower cover.

[0067] In some embodiments, the execution subject can press the upper cover and the lower cover transported on the conveying mechanism by the pressing mechanism to realize the assembly of the upper cover and the lower cover.

[0068] At step 205, the assembled material upper cover and the material lower cover are subjected to defect detection by the assembly detection module, and a defect detection result is obtained.

[0069] In some embodiments, the above-mentioned execution subject can perform defect detection on the assembled material upper cover and the material lower cover by the above-mentioned assembly detection module, and obtain a defect detection result.

[0070] At step 206, in response to the defect detection result indicating that the assembled material upper cover and the material lower cover have no defects, the conveying mechanism is controlled to convey the assembled material upper cover and the material lower cover to the storage bin.

[0071] In some embodiments, the above-mentioned execution subject can control the above-mentioned conveying mechanism to convey the assembled material upper cover and the material lower cover to the storage bin in response to the defect detection result indicating that the assembled material upper cover and the material lower cover have no defects.

[0072] The above various embodiments of the present disclosure have the following beneficial effects: the visual detection-based assembly method of some embodiments of the present disclosure can reduce the damage rate of reagent strips and improve the assembly efficiency and the qualified rate of the assembled reagent strips. Specifically, the reason why the related assembly methods cause the reagent strip damage and the low assembly efficiency and qualified rate of the assembled reagent strips is that: for some special reagent strips, manual picking is easy to cause the reagent strip pollution, and for the reagent strips with small size, manual picking is inconvenient, and the staff is easy to ignore some details of the pollution. Based on this, the visual detection-based assembly method of some embodiments of the present disclosure comprises: the lower cover grabbing mechanism described above is used to grab the material lower cover transported on the lower cover feeding module to the inside of the tray arranged on the conveying mechanism; the reagent strip grabbing mechanism described above is used to grab the reagent strip transported on the reagent strip feeding module to the inside of the material lower cover conveyed by the conveying mechanism; the upper cover grabbing mechanism described above is used to grab the material upper cover transported on the upper cover feeding module to the upper side of the material lower cover carrying the reagent strip; the pressing mechanism described above is used to press the material upper cover and the material lower cover transported on the conveying mechanism, so as to realize the assembly of the material upper cover and the material lower cover; the assembly detection module described above is used to detect the defects of the assembled material upper cover and the material lower cover, and the defect detection result is obtained; in response to the fact that the defect detection result represents that the assembled material upper cover and the material lower cover do not have defects, the conveying mechanism described above is controlled to convey the assembled material upper cover and the material lower cover to the storage box. Because the reagent strip grabbing mechanism is used to grab the reagent strip by the assembly equipment, manual contact with the reagent strip can be avoided, and the damage rate of the reagent strip is reduced. Also, the defect detection of the assembled reagent strip can be performed by the assembly detection module, and only the assembled reagent strip without defects is conveyed, so that the qualified rate of the assembled reagent strip can be improved. Therefore, the visual detection-based assembly method of some embodiments of the present disclosure can reduce the damage rate of the reagent strip and improve the assembly efficiency and the qualified rate of the assembled reagent strip.

[0073] Some embodiments of the present disclosure also provide a computer program product comprising a computer program which, when executed by a processor, implements any of the above visual detection-based assembly methods.

[0074] The above description is only some preferred embodiments of the present disclosure and the explanation of the applied technical principles. Those skilled in the art should understand that the scope of the application involved in the embodiments of the present disclosure is not limited to the technical solutions formed by the specific combinations of the above technical features, and should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above inventive concept. For example, the technical solutions formed by the mutual replacement of the above features and the technical features with similar functions disclosed in the embodiments of the present disclosure (but not limited to) can be formed.

Claims

1. A vision-based assembly device, comprising: The system comprises a control module, a feeding module, a conveying mechanism, a gripping mechanism, a vision inspection module, and a pressing mechanism. The control module is communicatively connected to the feeding module, and the feeding module includes an upper cover feeding module, a lower cover feeding module, and a reagent strip feeding module; The conveying mechanism is communicatively connected to the control module, and at least one tray is provided on the conveying mechanism, each of the at least one tray being used to carry the material cover; The gripping mechanism is communicatively connected to the control module. The gripping mechanism is located above the conveying mechanism. The gripping mechanism includes an upper cover gripping mechanism, a lower cover gripping mechanism, and a reagent strip gripping mechanism, wherein the reagent strip gripping mechanism is located between the upper cover gripping mechanism and the lower cover gripping mechanism. The control module is configured to control the lower cover gripping mechanism to grip the lower cover of the material being transported on the lower cover feeding module and place it inside the tray set on the conveying mechanism; to control the reagent strip gripping mechanism to grip the reagent strip being transported on the reagent strip feeding module and place it inside the lower cover of the material being transported by the conveying mechanism; and to control the upper cover gripping mechanism to grip the upper cover of the material being transported on the upper cover feeding module and place it above the lower cover of the material carrying the reagent strip. The control module controls the lower cover gripping mechanism to grip the lower cover of the material being transported on the lower cover feeding module and place it inside the tray set on the conveying mechanism according to a preset movement time and movement distance of the lower cover gripping mechanism. The pressing mechanism is communicatively connected to the control module. The pressing mechanism is located on one side of the upper cover gripping mechanism. The pressing mechanism is used to press the upper cover and lower cover of the material being transported on the conveying mechanism together to achieve the assembly of the upper cover and lower cover. The visual inspection module is communicatively connected to the control module. The visual inspection module includes an assembly inspection module. The control module is further configured to perform defect detection on the assembled upper and lower material covers using the assembly inspection module, obtaining defect detection results. The assembly inspection module also includes an assembly camera device. The control module is further configured to: The assembly camera device captures images of the assembled material top cover and material bottom cover to obtain an image of the assembly box. The assembly box image is subjected to blemish recognition to obtain at least one blemish region. The control module inputs the assembly box image into a pre-trained blemish recognition model to obtain the blemish region. The blemish recognition model is a machine learning model that takes the assembly box image as input and the blemish region as output. The assembly box image is cropped according to the at least one blemish region to obtain at least one blemish image. Determine the image features corresponding to each blemish image in at least one blemish image to obtain the blemish image features; By comparing the images of the assembled material top cover and material bottom cover with the assembly camera device, a comparison assembly box image is obtained. The image of the comparison assembly box is subjected to blemish identification to obtain at least one blemish area for comparison. Based on the at least one contrasting blemish area, the contrasting assembly box image is cropped to obtain at least one contrasting blemish image; Determine the image features corresponding to each contrast taint image in at least one contrast taint image to obtain contrast taint image features; The tainted image whose similarity to the corresponding tainted image features and the contrasting tainted image features in the at least one tainted image is greater than a preset similarity threshold is determined as the target tainted image; The sum of the areas of the identified target taint images is used to determine the area of ​​the target taint. Based on the target stain area, generate defect detection results; The control module is also configured to control the conveying mechanism to convey the assembled material top cover and material bottom cover to the storage bin in response to the defect detection result indicating that there are no defects in the assembled material top cover and material bottom cover.

2. The assembly equipment based on vision inspection according to claim 1, wherein, The upper cover gripping mechanism, the lower cover gripping mechanism, and the reagent strip gripping mechanism all include a robotic arm, a cylinder, and a pneumatic nozzle assembly. The pneumatic nozzle assembly includes at least one pneumatic nozzle. The robotic arm is detachably connected to the corresponding pneumatic nozzle assembly. The cylinder is used to provide negative pressure to the corresponding pneumatic nozzle assembly.

3. The assembly equipment based on vision inspection according to claim 1, wherein, The visual inspection module also includes a reagent strip position detection module, and the control module is further configured to perform pose detection on the reagent strip carried on the material cover through the reagent strip position detection module to obtain the pose detection result.

4. The assembly equipment based on vision inspection according to claim 3, wherein, The reagent strip position detection module includes a reagent strip camera device; and The control module is further configured to: The reagent strip camera device is used to capture an image of the material cover containing the reagent strip. Identify the reagent strip area in the image of the material's lower cover; The skew angle of the reagent strip area is detected to obtain reagent strip skew angle information, wherein the reagent strip skew angle information includes the reagent strip skew angle; Determine the offset information between the reagent strip area and the target area, wherein the offset information includes the offset amount; Based on the skew angle information of the reagent strip and the offset information, a pose detection result is generated.

5. The assembly equipment based on vision inspection according to claim 4, wherein, The control module is further configured to: In response to the pose detection result indicating that the skew angle information and the offset information of the reagent strip do not meet the preset pose conditions, the reagent strip gripping mechanism is controlled to adjust the position of the reagent strip carried in the material cover containing the reagent strip according to the skew angle information and the offset information.

6. The assembly equipment based on vision inspection according to claim 4, wherein, The control module is further configured to: The reagent strip camera device captures an image of the material cover on the tray mounted on the conveying mechanism. The position of the material under cover image is identified to obtain the position information of the material under cover; Based on the material cover position information, the reagent strip gripping mechanism is controlled to grip the reagent strips transported on the reagent strip feeding module and place them inside the material cover conveyed by the conveying mechanism.

7. The assembly equipment based on vision inspection according to claim 1, wherein, The visual inspection module further includes a lower cover position detection module, which includes a lower cover camera device; and The control module is further configured to: The lower cover camera device captures an image of the pallet located below the lower cover gripping mechanism on the conveying mechanism. Based on the pallet image, determine the pallet position information corresponding to the pallet image; Based on the pallet position information, the lower cover gripping mechanism is controlled to grip the lower cover of the material being transported on the lower cover feeding module and place it into the inside of the pallet set on the conveying mechanism.

8. A vision-based assembly method, applied to the vision-based assembly equipment according to any one of claims 1-7, wherein, The assembly equipment includes a control module, a feeding module, a conveying mechanism, a gripping mechanism, a vision inspection module, and a pressing mechanism. The feeding module includes an upper cover feeding module, a lower cover feeding module, and a reagent strip feeding module. The gripping mechanism includes an upper cover gripping mechanism, a lower cover gripping mechanism, and a reagent strip gripping mechanism. The method includes: The material lower cover transported on the lower cover feeding module is gripped by the lower cover gripping mechanism and placed into the inside of the tray set on the conveying mechanism; The reagent strips transported on the reagent strip feeding module are grasped by the reagent strip grasping mechanism and placed inside the material cover conveyed by the conveying mechanism. The upper cover gripping mechanism grips the material upper cover transported on the upper cover feeding module and places it above the material lower cover carrying the reagent strip. The control module controls the lower cover gripping mechanism to grip the material lower cover transported on the lower cover feeding module and place it inside the tray set on the conveying mechanism according to the preset movement time and movement distance of the lower cover gripping mechanism. The pressing mechanism presses together the upper and lower covers of the material being transported on the conveying mechanism to achieve the assembly of the upper and lower covers. The assembly inspection module performs defect detection on the assembled upper and lower material covers to obtain defect detection results. The assembly inspection module also includes an assembly camera device. The control module is further configured to: The assembly camera device captures images of the assembled material top cover and material bottom cover to obtain an image of the assembly box. The assembly box image is subjected to blemish recognition to obtain at least one blemish region. The control module inputs the assembly box image into a pre-trained blemish recognition model to obtain the blemish region. The blemish recognition model is a machine learning model that takes the assembly box image as input and the blemish region as output. The assembly box image is cropped according to the at least one blemish region to obtain at least one blemish image. Determine the image features corresponding to each blemish image in at least one blemish image to obtain the blemish image features; By comparing the images of the assembled material top cover and material bottom cover with the assembly camera device, a comparison assembly box image is obtained. The image of the comparison assembly box is subjected to blemish identification to obtain at least one blemish area for comparison. Based on the at least one contrasting blemish area, the contrasting assembly box image is cropped to obtain at least one contrasting blemish image; Determine the image features corresponding to each contrast taint image in at least one contrast taint image to obtain contrast taint image features; The tainted image whose similarity to the corresponding tainted image features and the contrasting tainted image features in the at least one tainted image is greater than a preset similarity threshold is determined as the target tainted image; The sum of the areas of the identified target taint images is used to determine the area of ​​the target taint. Based on the target stain area, generate defect detection results; In response to the defect detection result indicating that there are no defects in the assembled material top cover and material bottom cover, the conveying mechanism is controlled to convey the assembled material top cover and material bottom cover to the storage box.

9. A computer program product comprising a computer program that, when executed by a processor, implements the method according to claim 8.

Images