A Method and System for Quality Inspection of Packaging Paper Box Printing Based on Machine Vision

Abstract

This invention discloses a method and system for inspecting the printing quality of packaging cartons based on machine vision, belonging to the field of packaging printing inspection technology. The method includes the following steps: setting an electrostatic sensing strip along the conveying path for the packaging cartons printing quality inspection, collecting discharge flash signals and calculating the discharge trajectory, determining the starting position of each discharge event, and generating a list of electrostatic starting point anchor points; based on the list of electrostatic starting point anchor points, inserting a micro-delay window into the exposure sequence of the machine vision inspection device to limit the time interval of the discharge flash signals to the non-exposure interval, forming a timing avoidance trajectory. This invention, through the coordinated control of electrostatic sensing and timing avoidance, makes the light field distribution of the reflected image on the printed surface more uniform and stable, thereby significantly improving the image clarity and judgment reliability of machine vision inspection, ensuring the continuity and stability of the inspection data, and achieving high-precision identification and long-term stable imaging control of printing quality inspection through this dynamic feedback method.

Description

Technical Field

[0001] This invention relates to the field of packaging and printing inspection technology, specifically to a method and system for inspecting the printing quality of packaging paper boxes based on machine vision inspection. Background Technology

[0002] Machine vision-based inspection of packaging cardboard box printing quality refers to the process of non-contact automatic identification and judgment of patterns, text, colors, registration positions, and surface defects on the printed surface of packaging cardboard boxes using high-precision industrial cameras, light sources, image acquisition devices, and visual analysis technology. This inspection method acquires printing images in real time on the production line, establishes standard templates, and compares them frame by frame. It can detect minute abnormalities such as color differences, stains, scratches, omissions, and misalignments within milliseconds. Its core lies in replacing manual visual inspection with visual perception. Through optical imaging, feature extraction, edge recognition, and color distribution analysis, it achieves continuous monitoring and dynamic judgment of printing quality, thereby ensuring the accuracy, consistency, stability, and batch traceability of printing results. It is an important technological approach for achieving intelligent inspection and quality control in modern packaging production.

[0003] The existing technology has the following shortcomings: In the current process of packaging carton printing quality inspection, since the inspection equipment mostly operates in a constant-speed conveyor and high-frequency lighting environment, static electricity easily accumulates between the carton, conveyor belt, and inspection components in the conveying path when the ambient humidity is low. When the static charge increases in a local potential difference, transient discharge occurs, releasing a high-energy light spot signal for a very short time. This light spot usually has randomness, transient nature, and high brightness characteristics, and its spectral distribution is very similar to the reflectivity of printing ink. After the machine vision inspection device captures this discharge light spot within a fixed exposure cycle, it often misidentifies it as a bright spot, dirt spot, or ink abnormality on the printed surface, thus triggering an incorrect printing defect alarm. Such false alarms not only lead to distorted inspection data but may also cause automatic rejection or shutdown of the entire line, resulting in production cycle disorder and failure of quality judgment, making it difficult to guarantee the stability and reliability of the printing inspection process.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for inspecting the printing quality of packaging paper boxes based on machine vision inspection, so as to solve the problems in the background art mentioned above.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for inspecting the printing quality of packaging paper boxes based on machine vision inspection, comprising the following steps:

[0007] An electrostatic sensing strip is set up on the conveying path for quality inspection of printed paper boxes to collect discharge flash signals and calculate the discharge trajectory, determine the starting position of each discharge event, and generate a list of electrostatic starting point anchor points.

[0008] Based on the list of electrostatic starting point anchors, a micro-delay window is inserted into the exposure sequence of the machine vision inspection device to limit the time interval of the discharge flash signal to the non-exposure interval, thus forming a timing avoidance trajectory.

[0009] Based on the timing avoidance trajectory, a polarization suppression curtain is arranged on the imaging plane to perform polarization separation on the reflected light signal entering the exposure area, generating a reflection purification mask to reduce the interference reflection of the discharge spot.

[0010] A reflection cleansing mask is used to perform a weighted spreading operation on the image edge channel to extract the contour features of the real printed image from optical noise and establish a stable anchor point sequence across frames.

[0011] Based on the cross-frame stable anchor sequence, a breathing-type ion dissipation traction mechanism is driven to control the ion wind and ground coupling torque in a frequency-shifting manner, dynamically adjust the exposure duty cycle, and construct a closed-loop control process for electrostatic suppression and imaging stability.

[0012] Preferably, the steps for generating the electrostatic initiation anchor point list are as follows:

[0013] Electrostatic sensing strips are laid at key locations along the conveying path for quality inspection of printed packaging boxes. The electrostatic sensing strips are distributed along the entire conveying direction of the packaging boxes to collect electrostatic potential change signals generated during the conveying process in real time.

[0014] After the electrostatic signal acquisition is completed, the acquired potential change signal is segmented and extracted. The discharge flash signal is extracted by time correlation comparison and the time point and region of the discharge event are determined.

[0015] After extracting the discharge flash signal, the discharge trajectory is calculated based on the distribution location of the electrostatic sensing strip, combined with the spatial layout of the transmission path, to determine the propagation direction and coverage of the released charge.

[0016] After the discharge trajectory calculation is completed, the starting position of each discharge event is determined based on the concentration of the trajectory distribution and the duration characteristics. The discharge direction, duration and energy change amplitude are recorded, and a list of electrostatic starting point anchor points is generated.

[0017] Preferably, the steps for generating the temporal avoidance trajectory are as follows:

[0018] Based on the time attributes and spatial distribution information in the list of electrostatic initiation anchor points, an exposure timing reference for the machine vision inspection device is established, and the time window of each discharge event is time-aligned with the exposure trigger signal.

[0019] After establishing the exposure timing reference, the start time of the exposure trigger signal is slightly delayed according to the time distribution pattern of the discharge event, so that the exposure start time avoids the high-energy release period of the discharge signal.

[0020] After the exposure trigger signal completes the delay adjustment, the time difference of consecutive exposure cycles is smoothed to form a continuous time delay curve to construct a complete timing avoidance trajectory.

[0021] After the timing avoidance trajectory is formed, the exposure cycle length, exposure duration, image transmission interval and exposure start phase are updated synchronously according to the trajectory change pattern to establish a timing avoidance network to isolate the discharge light signal.

[0022] Preferably, the micro-delay adjustment is achieved through a time delay adjustment unit. The time delay adjustment unit receives control parameters from the timing mapping table to delay the rising edge of the exposure trigger signal, so that the exposure start time lags behind the energy release period of the discharge flash signal, thereby keeping the exposure action and the electrostatic discharge process separated on the time axis, ensuring the illumination stability and imaging uniformity of the image acquisition process.

[0023] Preferably, the steps for generating the reflection cleansing mask are as follows:

[0024] According to the timing avoidance trajectory, a polarization suppression curtain is arranged at the front end of the imaging plane. The polarization suppression curtain consists of a transparent substrate layer, a polarization filter layer and a reflection reduction layer, which is used to perform preliminary polarization screening on the reflected light entering the exposure area.

[0025] After the polarization anti-reflection curtain is installed, the polarization orientation is dynamically controlled by a micro-driving component set at the edge of the polarization filter layer, so that the polarization direction alternates between the exposure range and the non-exposure range to filter out the reflection of the discharge spot.

[0026] After polarization orientation adjustment is completed, the reflected light signal passing through the polarization suppression curtain is spatially separated by the optical guidance component, so that the high-energy reflected light is guided to the light absorption groove and the effective reflected light is retained to enter the imaging plane.

[0027] After completing the spatial separation of the optical signal, a reflective purification mask is generated on the imaging plane. The reflective purification mask is formed by a combination of a transparent transmission layer and a light-absorbing filter layer, which is used to weaken residual interference light and balance the reflected brightness.

[0028] Preferably, the polarization suppression curtain maintains a fixed tilt angle with the imaging plane, the orientation direction of the polarization filter layer is perpendicular to the polarization direction of the discharge spot when switching between the exposure range and the non-exposure range, and the light-transmitting aperture array of the reflective purification mask is consistent with the polarization direction, so as to ensure that the light path is continuous and stable during exposure and maintain a uniform distribution of the imaging light field.

[0029] Preferably, the steps for generating a cross-frame stable anchor sequence are as follows:

[0030] After completing the optical filtering of the reflective purification mask, the purified image signal is divided into edge channels. During the signal output stage of the photosensitive element, the imaging plane is divided into parallel image acquisition channels and a light intensity compensation band is set to enhance the edge response.

[0031] After the edge channels are formed, the light field intensity of each channel is reweighted and spread using a reflective purification mask. The edge features of the printed pattern are enhanced by the diffusion and redistribution of light energy through the light-transmitting aperture array.

[0032] After completing the light intensity weighting, the outline shape of the printed pattern is extracted by continuously scanning the light intensity change trend of each edge channel, and a transition zone is set between the channels to ensure the continuity of the outline.

[0033] After extracting the complete outline of the printed image, a cross-frame stable anchor point sequence is established based on the time sequence of consecutive exposure frames. The anchor point motion trajectory is formed by continuously matching the anchor point positions, thus maintaining the stability of image recognition.

[0034] Preferably, during the weighted spreading process, the aperture and spacing of the light-transmitting hole array of the reflective purification mask change step by step according to the light field brightness gradient, and the angle of the light-transmitting hole wall is consistent with the direction of the channel boundary, so that the light energy diffuses along the edge direction and forms an energy superposition area in the longitudinal direction, thereby further improving the brightness contrast and imaging clarity of the printed pattern boundary.

[0035] Preferably, the following steps are taken to implement a breathing-type ion dissipation traction mechanism driven by a cross-frame stable anchor point sequence, controlling the ion wind and ground coupling torque in a frequency-shifting manner, and dynamically adjusting the exposure duty cycle to form a closed-loop control process:

[0036] After the cross-frame stable anchor point sequence is established, the spatial position coordinates, time series distribution and brightness fluctuation trend in the anchor point sequence are converted into electrostatic environment change feedback signals, which are used to drive the breathing ion dissipation traction mechanism.

[0037] After the breathing ion dissipation traction mechanism receives the anchor sequence input signal, it adjusts the ion wind output cycle and exposure sequence through the frequency misalignment control device to keep the ion wind and the exposure process misaligned on the time axis and achieve electrostatic neutralization.

[0038] While the ion wind is output at an off-frequency, the spatial balance between ion release and electrostatic neutralization is controlled by adjusting the grounding coupling torque, so that the charge flow direction remains stable and a unidirectional potential release channel is formed.

[0039] After the ion wind frequency offset output and ground coupling torque balance adjustment form a stable operation, the exposure duty cycle is dynamically adjusted to form a closed-loop coordinated control between imaging illumination intensity and electrostatic discharge state.

[0040] A machine vision-based inspection system for packaging paper box printing quality includes an electrostatic sensing module, a timing control module, an optical purification module, an image extraction module, and a closed-loop control module.

[0041] The electrostatic sensing module sets up an electrostatic sensing strip on the conveying path for the printing quality inspection of packaging paper boxes, collects discharge flash signals and calculates the discharge trajectory, determines the starting position of each discharge event, and generates a list of electrostatic starting point anchor points.

[0042] The timing control module, based on the list of electrostatic starting point anchor points, inserts a micro-delay window into the exposure timing of the machine vision inspection device, limiting the time interval of the discharge flash signal to the non-exposure interval, thus forming a timing avoidance trajectory.

[0043] The optical purification module arranges polarization suppression curtains on the imaging plane according to the timing avoidance trajectory, performs polarization separation on the reflected light signal entering the exposure area, and generates a reflection purification mask to reduce the interference reflection of the discharge spot.

[0044] The image extraction module uses a reflection purification mask to perform a weighted spreading operation on the image edge channel to extract the contour features of the real printed image from optical noise and establish a stable anchor point sequence across frames.

[0045] The closed-loop control module, based on the cross-frame stable anchor point sequence, drives the breathing ion dissipation traction mechanism, controls the ion wind and ground coupling torque in a frequency-shifting manner, dynamically adjusts the exposure duty cycle, and constructs a closed-loop control process for electrostatic suppression and imaging stability.

[0046] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0047] This invention introduces a coordinated control mechanism of electrostatic sensing and timing avoidance on the conveying path for packaging carton printing quality inspection. This effectively isolates electrostatic discharge behavior both temporally and spatially, preventing instantaneous flashes from entering the exposure area and interfering with imaging. By setting a micro-delay window within the exposure cycle and combining it with a polarization suppression structure, high-brightness transient signals in the optical path are suppressed, resulting in a more uniform and stable light field distribution in the reflected image from the printed surface. This significantly improves the image clarity and reliability of machine vision inspection, ensuring the continuity and stability of the inspection data.

[0048] This invention achieves dynamic coupling between imaging stability and electrostatic dissipation by constructing a closed-loop control process centered on a cross-frame stable anchor point sequence. The breathing-type ion wind traction mechanism maintains the potential balance of the detection environment under frequency misalignment drive, and the exposure duty cycle is adjusted in real time according to the imaging state, thus creating a complementary relationship between electrostatic dissipation and optical acquisition. Through this dynamic feedback method, the detection process maintains stable illumination and interference-free images even under high-speed production conditions, achieving high-precision identification of printing quality and stable imaging control for long-term operation. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0050] Figure 1 This is a flowchart of the packaging paper box printing quality inspection method based on machine vision inspection according to the present invention.

[0051] Figure 2 This is a schematic diagram of the modules of the packaging paper box printing quality inspection system based on machine vision inspection according to the present invention. Detailed Implementation

[0052] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0053] This invention provides, for example Figure 1 The machine vision-based inspection method for packaging paper box printing quality inspection shown includes the following steps:

[0054] An electrostatic sensing strip is set up on the conveying path for quality inspection of printed paper boxes to collect discharge flash signals and calculate the discharge trajectory, determine the starting position of each discharge event, and generate a list of electrostatic starting point anchor points.

[0055] An electrostatic sensing strip is installed along the conveyor path for quality inspection of printed packaging boxes. Through continuous acquisition and spatial positioning of electrostatic signals, the system captures discharge flash signals, calculates discharge trajectories, determines the discharge initiation position, and generates a list of electrostatic initiation point anchors. The specific steps of this process are as follows:

[0056] Electrostatic sensing strips are laid at key locations along the conveyor path for packaging carton printing quality inspection. These strips are distributed along the entire conveying direction of the carton, covering the entire path from the carton's entry into the inspection area to its exit. The sensing strips consist of a conductive polymer layer, a shielding layer, and a flexible support layer, with a highly sensitive conductive coating on the surface and an embedded distributed wire mesh. The wire mesh uses a staggered layout, enabling the sensing strips to capture electrostatic potential changes in multiple directions. This ensures immediate response to charge accumulation when friction or separation occurs between the carton, conveyor belt, guide rollers, tension rollers, and the inspection device housing. Each sensing strip has a signal output terminal connected to a reference ground terminal at both ends. The signal output terminal is connected to a signal acquisition circuit via shielded wires to collect potential changes at various locations along the conveying direction. As the carton moves along the conveyor path, the sensing strips sense the potential difference caused by friction in real time, outputting an electrostatic response signal that varies over time. This continuous sensing signal not only reflects the trend of electrostatic charge accumulation but also reveals the spatial differences in charge distribution along the conveying path, providing a continuous source of raw data for subsequent discharge signal identification.

[0057] After acquiring electrostatic signals, the acquired potential change signals are segmented and extracted to obtain discharge flash signals. Electrostatic discharge typically occurs the instant the potential difference reaches the dielectric breakdown threshold, manifesting as a pulse signal with an extremely short duration. Since packaging boxes, conveyor belts, and guide components may discharge simultaneously at different locations, it is necessary to determine the temporal correlation of signals acquired from multiple sensing bands. By comparing the sequence of signal changes across sensing bands, the time point and approximate area of ​​each discharge event can be determined. For example, when the first and third sensing bands generate high-amplitude pulses within a very short time interval, while the signal from the second sensing band is in a transitional state, it can be inferred that the main release area of ​​this discharge event is located between the conveyor path positions corresponding to the first and third sensing bands. Through this time-series-based signal comparison method, multiple individual discharge events can be extracted from continuous potential change curves, and the start signal, peak signal, and attenuation signal of each discharge event are labeled, thereby constructing a flash signal dataset that reflects the characteristics of discharge activity. This dataset will provide a temporal basis for the next stage of spatial trajectory calculation.

[0058] After extracting the discharge flash signal, the spatial trajectory of the discharge event is calculated based on the spatial layout of the conveyor path. The packaging carton conveyor path typically consists of multiple conveyor belts, guide rollers, and detection stations. Electrostatic sensing belts are distributed at fixed intervals on the surfaces and sidewalls of these structures. Using the actual installation position of each sensing belt as a coordinate reference, the propagation direction and coverage of the charge release in space can be deduced by comparing the occurrence times of the discharge peak collected by different sensing belts. Specifically, when the discharge signal first appears on the sensing belt upstream of the carton, and then a similar waveform appears on the adjacent downstream sensing belt, it can be determined that the discharge event propagates from upstream to downstream; conversely, when the signal first appears on the downstream sensing belt, it indicates that the discharge is released from downstream to upstream. By recording the order of appearance and intensity differences of the discharge signal on each sensing belt, the spatial distribution of the discharge trajectory on the conveyor path can be obtained. For areas where discharges occur frequently, the concentrated area of ​​discharge activity can also be obtained through cumulative statistics. This trajectory information reflects the propagation characteristics of electrostatic energy release in the conveyor path, providing a quantifiable basis for determining the starting position of charge release.

[0059] After the discharge trajectory calculation is completed, the starting position of each discharge event is determined based on the concentration and duration characteristics of the trajectory distribution, and a list of electrostatic starting point anchors is generated. Specifically, the starting point of each discharge event trajectory is taken as an electrostatic starting point anchor, and its position coordinates in the transport path coordinate system, discharge direction, discharge duration, and discharge energy variation amplitude are recorded. When multiple discharge events occur at the same or adjacent locations, the system marks this area as a high-frequency discharge zone. All electrostatic starting point anchors are arranged in spatial and temporal order to form an electrostatic starting point anchor list. This list is stored in tabular form and contains complete information such as anchor number, anchor coordinates, discharge direction angle, discharge duration, energy intensity level, and discharge time window. The electrostatic starting point anchor list not only reflects the spatial distribution of electrostatic discharge sources in the transport path but also provides a precise reference for subsequent exposure timing adjustments. When the packaging carton enters the detection area, the detection device can identify areas prone to discharge in advance based on the anchor point list and avoid the time windows of these areas in subsequent exposure control stages, thereby reducing the probability of electrostatic discharge flashes entering the exposed image.

[0060] Based on the list of electrostatic starting point anchors, a micro-delay window is inserted into the exposure sequence of the machine vision inspection device to limit the time interval of the discharge flash signal to the non-exposure interval, thus forming a timing avoidance trajectory.

[0061] After generating the list of electrostatic initiation anchor points, the exposure sequence of the machine vision inspection device is precisely adjusted based on the discharge initiation position, discharge direction, discharge time window, and discharge energy intensity information recorded in the list. A micro-delay window is inserted into the exposure cycle to limit the time interval of the discharge flash signal to the non-exposure interval, forming a complete timing avoidance trajectory, thereby ensuring the stability and reliability of image acquisition. The specific steps of this process are as follows:

[0062] Based on the temporal attributes and spatial distribution information in the electrostatic initiation anchor point list, an exposure timing reference is established for the machine vision inspection device. The electrostatic initiation anchor point list records the start time, duration, discharge energy intensity, and position coordinates on the transport path for each discharge event. To ensure the machine vision inspection device maintains time synchronization with the discharge events, an adjustable timing mapping table is first established in the device's control section. This timing mapping table uses the exposure trigger signal as the reference axis, dividing each exposure cycle into four time periods: exposure preparation, exposure acquisition, image reading, and signal transmission. Subsequently, the time window of each discharge event in the electrostatic initiation anchor point list is mapped to the corresponding exposure cycle, and the relative time difference between each discharge event and the exposure trigger signal is determined through time alignment. Through this process, the exposure trigger cycle and the discharge occurrence cycle can be precisely superimposed on the time axis, thus laying an accurate time reference foundation for inserting micro-delay windows.

[0063] After establishing the timing reference, the start time of the exposure trigger signal is slightly delayed based on the temporal distribution pattern of discharge events. The exposure trigger signal is output by the control device and is used to activate the camera's photosensitive element to enter the exposure acquisition stage. Under normal conditions, the exposure trigger signal period is fixed, and the exposure interval is constant. Because electrostatic discharge has the characteristics of short duration and high energy, its duration is usually much shorter than the exposure cycle. Therefore, by applying a slight delay to the exposure trigger signal, the high-brightness flash time window of the discharge signal can be effectively avoided. The delay amount is set according to the discharge duration and discharge time distribution characteristics in the electrostatic starting point anchor point list, and is usually between tens of microseconds and hundreds of microseconds. In specific implementation, an independent time delay adjustment unit is set in the detection device. This unit receives control parameters from the timing map table and delays the rising edge of the trigger pulse to make the exposure start time lag behind the high-energy release period of the discharge event. When the packaging box enters a known high-incidence area of ​​discharge, the delay unit automatically predicts the time and position of the impending discharge based on the time stamps in the anchor point list and synchronously delays the next frame exposure cycle. In this way, the exposure time of each frame can be ensured to avoid the time interval of the electrostatic discharge flash signal, so that the light spot generated by electrostatic discharge is released before the camera is exposed, thereby avoiding the image sensor from capturing interfering light.

[0064] After the exposure trigger signal completes a slight delay adjustment, to maintain the stability of continuous detection, a smooth and continuous delay curve needs to be formed during multi-frame exposure, thus constructing a complete temporal avoidance trajectory. Since the packaging box passes continuously along the conveyor path, electrostatic discharge events in different areas exhibit randomness and overlap on the time axis. Therefore, the exposure delay cannot be a single abrupt operation but should gradually transition between multiple exposure cycles. To this end, after each exposure delay, the detection device compares the current exposure time with the previous exposure time and calculates the time difference. This time difference is used to adjust the starting point of the next exposure delay, making the change in exposure time continuous and gradual. As the detection process continues, these tiny time offsets form a continuous curve on the time axis. The fluctuations of this curve reflect the continuous avoidance of discharge events by the exposure action. Through continuous temporal adjustment, a stable avoidance pattern is formed along the entire detection conveyor path in the time dimension. Even if the location or time of a discharge event changes slightly, it will not affect the overall exposure rhythm, thus maintaining the uniformity of the detected image and the consistency of illumination.

[0065] After forming a complete timing avoidance trajectory, the exposure control parameters need to be updated synchronously according to the trajectory's changing patterns to ensure the anti-interference stability of the detection process during continuous operation. Exposure control parameters include exposure cycle length, exposure duration, image transmission interval, and exposure start phase. By continuously comparing the timing avoidance trajectory with the electrostatic initiation anchor point list, the exposure control parameters can be dynamically updated, allowing the detection device to adapt to different discharge environments in the detection of different batches of packaging cartons. When the detection device reaches a new batch of printed packaging cartons, the control unit first reads the timing avoidance trajectory formed in the previous detection cycle as an initial reference. Subsequently, based on the electrostatic initiation anchor point list generated for the current batch, it calculates the new discharge time distribution characteristics and superimposes the time offset in the new trajectory with the original trajectory to obtain a multi-layer timing avoidance model. This model ensures that the exposure process remains separated from the discharge flash signal, meaning the energy release of the discharge signal always lies within the non-exposure range. When the detection device is in continuous sampling mode, the exposure trigger unit and the delay adjustment unit communicate synchronously at a fixed period to ensure the stability and repeatability of the delay. Ultimately, through this series of timing control operations, the machine vision inspection device forms a time-avoidance network consisting of multiple micro-delay windows during the packaging box printing quality inspection process. Each time node in this network corresponds to a potential discharge event window, and all exposure actions are avoided according to these time nodes, ensuring that the discharge light signal is completely isolated from the exposure stage in the time dimension. Through this complete process, exposure control not only achieves active avoidance of discharge flash spots but also ensures continuous balance of illumination intensity during imaging. This enables the machine vision inspection device to stably output clear, electrostatic interference-free printed images in high-speed production environments, providing a reliable foundation for subsequent optical cleaning and image recognition processes.

[0066] Based on the timing avoidance trajectory, a polarization suppression curtain is arranged on the imaging plane to perform polarization separation on the reflected light signal entering the exposure area, generating a reflection purification mask to reduce the interference reflection of the discharge spot.

[0067] After establishing the timing avoidance trajectory, to prevent the bright flashes generated by electrostatic discharge from interfering with the imaging results by entering the exposure area, a polarization suppression curtain is placed on the imaging plane. Polarization separation is performed on the reflected light signals entering the exposure area, and a reflection purification mask is generated based on this to weaken the interference reflections of the electrostatic discharge spots, thus keeping the imaging optical path stable and clean. The specific steps are as follows:

[0068] Based on the established temporal avoidance trajectory, a polarization suppression curtain is placed at the front end of the imaging plane. The polarization suppression curtain is positioned along the optical axis of the machine vision inspection device and covers the entire effective area of ​​the imaging optical path, ensuring that reflected light from the surface of the printed cardboard box must pass through this structure before entering the photosensitive element. The polarization suppression curtain consists of three optical layers: a transparent substrate layer, a polarization filter layer, and a reflection reduction layer. The transparent substrate layer is made of high-molecular-weight polycarbonate material, and its thickness is precisely controlled to ensure that light does not deform during transmission. The polarization filter layer is composed of a polymer-oriented thin film, with regularly arranged metal micro / nano particles covering its surface. The orientation of the metal particles is consistent with the orientation direction of the thin film, used to limit the vibration direction of the transmitted light waves. The reflection reduction layer is a functional layer formed by a carbon-based light-absorbing coating and a porous microcrystalline film, capable of absorbing some stray light with incident angles deviating from the optical axis and reducing specular reflection of high-energy reflected light. The polarization suppression curtain is tilted at a certain angle to the imaging plane. This tilt angle is determined based on the optical axis height of the exposure area and the lens focal length, ensuring that the polarization direction of the reflected light forms a fixed angle with the normal to the photosensitive surface after passing through the polarization suppression curtain. This structural arrangement allows for preliminary polarization screening of reflected light from the surface of the printed paper box in space, creating the necessary conditions for subsequent optical signal separation.

[0069] After the polarization suppression curtain is deployed, its polarization orientation is dynamically adjusted based on the temporal distribution characteristics of the exposed and unexposed areas in the timing avoidance trajectory, ensuring synchronization with the exposure control process on the time axis. Polarization orientation adjustment is achieved through a micro-drive component located at the edge of the polarization filter layer. This component consists of a flexible piezoelectric ceramic sheet, a torsion bracket, and a micro-elastic connecting rod. Under the influence of an electrical signal, the piezoelectric ceramic sheet undergoes a slight deformation, driving the torsion bracket to apply an angle adjustment force to the polarization filter layer. The micro-elastic connecting rod ensures that the polarization filter layer can spring back to its original position after deformation, thus achieving periodic deflection. During device operation, the control unit outputs electrical signals to the micro-drive component based on real-time data from the timing avoidance trajectory, causing the polarization filter layer to alternately switch between the exposed and unexposed areas. When in the non-exposure zone, the polarization filter layer is oriented in the same direction as the main polarization direction of the reflected discharge spot. At this time, the high-brightness instantaneous component of the reflected light is filtered out or guided out of the imaging light path. When in the exposure zone, the polarization filter layer is adjusted to be perpendicular to the polarization direction of the imaging light, allowing the normal reflected light from the printed surface to fully enter the imaging plane. Through this continuous orientation switching, the polarization anti-reflection curtain dynamically coordinates with the exposure process in the time dimension, ensuring that only effective light from the printed surface is received by the photosensitive element during exposure, while the discharge flashes are isolated during the polarization screening stage.

[0070] After polarization orientation adjustment, the reflected light signal passing through the polarization suppression curtain undergoes spatial separation and energy purification to ensure that only light meeting imaging conditions enters the exposure area. The polarized reflected light first enters the optical guiding assembly, which consists of multiple sets of transmission lenses and refractive reflectors to rearrange the light according to the incident angle and polarization direction. The transmission lenses are made of low-dispersion glass with an anti-reflection coating to reduce secondary reflections. The refractive reflectors are composed of an aluminum-coated reflective layer and a high-hardness silicon-based support layer, and their installation angle is consistent with the tilt angle of the polarization suppression curtain. This ensures that the filtered light is guided along the optical axis to the photosensitive plane, while light not allowed to pass is refracted into the absorption groove for energy dissipation. Through this spatial separation structure, high-energy reflected light from electrostatic discharge is guided into the absorption groove, preventing bright spots from interfering with the imaging plane; while conventional reflected light from the printed pattern smoothly enters the imaging area, forming a uniform and directional light field. During operation, the optical guiding component maintains a stable optical axis alignment, enabling light rays at different times to form a stable polarization coupling relationship between the polarization suppression curtain and the optical guiding component, thereby further reducing the spatial influence of the discharge spot.

[0071] After spatial separation of the optical signal, a reflective purification mask is generated on the imaging plane based on the effective reflected light distribution through the polarization suppression curtain and optical guiding components. This mask is used to weaken residual interference light and balance reflected brightness. The reflective purification mask is composed of a transparent transmission layer and a light-absorbing filter layer, and is installed between the imaging plane and the photosensitive element. The transparent transmission layer is made of high-transmittance silicon oxide material, and its surface is processed with uniformly arranged micron-sized honeycomb light-transmitting holes. The aperture of each light-transmitting hole is precisely designed according to the exposure light intensity to ensure that the light transmission ratio in different areas remains consistent. The light-absorbing filter layer is located below the transmission layer and is composed of a mixture of nano-carbon black composite film and titanium oxide particles. The particle distribution density gradually changes, and it is used to absorb residual high-energy transient components in the light. When light passes through the transmission layer, the brighter reflected light is partially attenuated, while the lower-energy effective reflected light is completely transmitted, thus forming a uniform light field on the photosensitive surface. Through the action of the reflective purification mask, the true pattern on the printed surface is clearly presented on the photosensitive element, while the high-brightness spots generated by electrostatic discharge cannot form visible interference because the energy is dissipated in the filter layer. The reflective purification mask is updated synchronously with the orientation change of the polarization suppression curtain during the detection operation, and its light-transmitting aperture array is consistent with the polarization direction to ensure the continuity and stability of the light path throughout the exposure period.

[0072] A reflection cleansing mask is used to perform a weighted spreading operation on the image edge channel to extract the contour features of the real printed image from optical noise and establish a stable anchor point sequence across frames.

[0073] After the initial purification of the optical signal using the reflective cleansing mask, in order to further remove optical noise from the printed image, extract the true outline features of the printed pattern from the purified light field, and establish a stable anchor point sequence in continuously acquired multiple frames of images to ensure that machine vision inspection maintains imaging consistency and recognition stability in a high-speed production environment, a reweighting operation needs to be performed using a reflective cleansing mask in the image edge channel. The specific steps are as follows:

[0074] After optical filtering with the reflective cleansing mask, the cleaned image signal is divided into edge channels. After exposure, the imaging plane forms a uniformly distributed optical image, composed of reflected signals from the printed pattern and a small amount of background scattered light. To spatially separate the detailed areas of the printed pattern, during the signal output stage of the photosensitive element, the entire imaging plane is divided into several parallel image acquisition channels, each corresponding to a detection band on the printed surface. The division method is determined based on the exposure resolution and the camera's field of view, ensuring that the width of each channel completely covers the printed lines, text edges, and color block boundaries. After channel division, the light signal is transmitted sequentially according to the channel direction, and the output signal of each channel is acquired independently to avoid light intensity interference between different areas. Since the reflective cleansing mask has weakened the influence of high-brightness discharge light, the brightness difference of the light signals in each channel is small, resulting in a stable light intensity gradient. To further enhance the optical response of the edge region, an intensity compensation band is set at the edge of each channel during the channel segmentation process. This compensation band controls the incident angle so that the edge reflected light is concentrated at the edge pixels of the photosensitive element with a higher energy density, thereby making the grayscale changes at the image edge more obvious and providing a basis for subsequent contour enhancement.

[0075] After the edge channels are formed, a reflective cleansing mask is used to perform a weighted spread of the light field intensity in each channel to enhance the edge features of the printed pattern. The core of weighted spread is to redistribute the light energy in the printed outline area by controlling the lateral diffusion and vertical balance of light intensity within the channel. Specifically, the reflective cleansing mask has a uniformly arranged array of micron-sized light-transmitting holes. When light passes through these holes, it undergoes minute diffusion based on the hole diameter, hole spacing, and hole wall angle. When the bright areas in the imaging light field pass through these hole arrays, the light energy diffuses along the channel boundary, thus creating an energy superposition effect at the edge; while in the dark areas of the light field, the light-transmitting holes allow some outer light to penetrate, supplementing areas with insufficient light energy. In this way, the light intensity distribution within the channel forms a gradient that increases from the center to the edge, relatively enhancing the edge brightness of the printed pattern and making the boundary lines appear with higher contrast in the optical image. Meanwhile, in the longitudinal direction of the light field, the slight difference in thickness of the reflective purification mask causes a minute phase delay in the transmitted light. This delay accumulates during consecutive frame exposures, forming a slight light field shift and creating traceable overlapping areas in the light energy distribution between different frames. Through weighted spreading processing, the edge information of the printed pattern is continuously enhanced, while the energy of optical noise is uniformly dispersed, avoiding false boundaries caused by local reflections.

[0076] After completing the intensity reweighting process, contour features are extracted from the enhanced edge channel light field to separate the true structure of the printed surface from the cleaned image. Since the reflection cleansing mask and reweighting have created clear brightness differences in the boundary lines of the printed pattern, the contour shape of the printed pattern is gradually determined at this stage by continuously scanning the intensity change trends in each edge channel. The scanning operation is performed along the channel direction, with the intensity value read point by point from the start to the end of each channel, and the locations of abrupt intensity changes are used as potential boundary points. When adjacent channels show intensity changes at the same location, it indicates the existence of a true printing boundary at that location; however, if a channel exhibits an isolated abrupt change without synchronous changes in adjacent channels, it is identified as optical noise. This scanning method based on light field consistency allows for the gradual spatial depiction of the outer contour of printed text, the shape of pattern lines, and the position of color block boundaries. To prevent breaks during edge extraction, a transition zone is set at the intensity junctions between channels. This zone consists of partially overlapping pixels between channels, used to connect adjacent contour segments and ensure edge continuity. Through this process, the main structure of the printed image is completely depicted on the imaging plane, with clear and continuous outlines that accurately reflect the distribution boundaries of the printing ink.

[0077] After extracting the complete outline of the printed image, a cross-frame stable anchor point sequence is established based on the temporal order of consecutive exposure frames to maintain the consistency of image features across multiple frames. The anchor point sequence is built upon the outline features extracted in the previous step, with key edge points in each frame selected as anchor points. The selection criteria for anchor points are stable position, constant brightness, and significant changes in boundary curvature. By analyzing the light intensity distribution at the same location in consecutive frames, the spatial coordinates of the anchor points can be determined. As the packaging box moves along the conveyor path, the anchor points in each frame will undergo slight spatial displacement. By continuously matching the anchor point positions in adjacent frames, the anchor point movement trajectory can be formed. If an anchor point maintains a stable relative position across multiple frames, it is classified as a cross-frame stable anchor point. All stable anchor points are arranged in spatial order to form a cross-frame anchor point sequence. This sequence records the morphological continuity of the printed pattern on the time axis and can be used to track minute offsets, stretching, or deformations during the printing process. To further enhance the stability of anchor points, during the anchor point update process, the new frame image is compared with the anchor point sequence of the previous frame. If the distance and brightness difference between the newly added contour point and the existing anchor points are within a set range, it is included in the sequence, realizing the dynamic extension of the anchor points. Through this continuous anchor point tracking mechanism, machine vision inspection can maintain the consistency of image recognition under high-speed operation and ensure the stability and reliability of print quality inspection results.

[0078] Based on the cross-frame stable anchor sequence, a breathing-type ion dissipation traction mechanism is driven to control the ion wind and ground coupling torque in a frequency-shifting manner, dynamically adjust the exposure duty cycle, and construct a closed-loop control process for electrostatic suppression and imaging stability.

[0079] After obtaining the cross-frame stable anchor point sequence, in order to further eliminate the influence of electrostatic accumulation on the printing inspection imaging process and to establish a dynamic coordination relationship between exposure timing and charge release, a breathing-type ion dissipation traction mechanism is driven by the cross-frame stable anchor point sequence. This mechanism controls the ion wind and grounding coupling torque in a staggered frequency manner and dynamically adjusts the exposure duty cycle during imaging. This creates a continuous closed-loop control between charge suppression and optical exposure, achieving electrostatic dissipation and image stabilization in the printing inspection process. The specific steps are as follows:

[0080] After the cross-frame stable anchor point sequence is established, the spatial coordinates, temporal distribution, and brightness fluctuation trends in the sequence are converted into feedback signals of electrostatic environment changes, serving as the input for driving the breathing-type ion dissipation traction mechanism. The cross-frame stable anchor point sequence records the edge coordinate changes and light intensity changes of the printed image in continuous frame imaging. The brightness fluctuation of each anchor point reflects the dynamic changes of electrostatic influence in the image acquisition area. When the brightness of the anchor point shifts periodically, it indicates that electrostatic charge accumulates on the surface of that area and interferes with the local optical path. Therefore, the control unit performs paired analysis of the light intensity change amplitude and time offset of each anchor point to obtain the temporal characteristic signal of charge distribution. This signal is converted into voltage amplitude change and transmitted to the control terminal of the breathing-type ion dissipation traction mechanism. The breathing-type ion dissipation traction mechanism is installed above the packaging carton conveying path, and its output channel extends along the length of the detection area, covering the entire imaging area. The device contains multiple ion release electrode arrays, each connected to an independent airflow channel, with a breathing-type adjustment cavity at the front end of the airflow channel. The regulating chamber controls the output speed and direction of the airflow through the periodic expansion and contraction of an internal diaphragm, causing the ion wind to exhibit a periodic, breathing-like flow pattern. Through this structure, when the control unit outputs an anchor point feedback signal, the ion dissipation device can automatically adjust the flow rate and output sequence of the ion wind according to the charge distribution intensity, providing an adjustable ion flow source for the subsequent electrostatic dissipation process.

[0081] After receiving the input signal from the anchor point sequence, the breathing-type ion dissipation traction mechanism uses a frequency misalignment control device to adjust the ion wind output cycle and exposure sequence, ensuring that the ion wind and exposure process do not interfere with each other on the time axis. The frequency misalignment control device, consisting of a control electrode, a drive diaphragm, and a variable frequency oscillator, is installed upstream of the ion emission cavity. When the detection device is in the exposure phase, the frequency misalignment control device reduces the output frequency of the oscillator, slowing down the potential change of the ion emission electrode, causing the drive diaphragm to contract, reducing the airflow pressure, and weakening the ion wind output. After exposure, the oscillator frequency increases, the electrode potential change accelerates, the diaphragm expands, and the ion wind intensity increases, thus achieving an output mode that is misaligned with the exposure cycle. Under this frequency misalignment, the ion wind flow tends to be stable during the exposure period, avoiding airflow disturbances that could cause fluctuations in the optical path; while during non-exposure periods, the ion wind output is enhanced, allowing charged particles to diffuse into the detection area in a shorter time, rapidly neutralizing the static charge accumulated in the air. To further improve static dissipation efficiency, the frequency misalignment period of the ion wind is automatically adjusted according to the rate of change of the anchor point sequence. When the brightness of the anchor point changes frequently, the control device shortens the ion wind output cycle to make ion release more concentrated; when the anchor point is stable and the brightness does not change significantly, the ion wind output cycle is appropriately extended to maintain the dynamic balance of the electrostatic field. Through this time-displacement-based ion wind control method, the electrostatic dissipation process and the imaging exposure process avoid each other, ensuring that the imaging plane is not disturbed by airflow during optical acquisition.

[0082] While controlling the frequency of the ion wind output, the grounding coupling torque is adjusted to achieve a spatial equilibrium between the ion release and electrostatic neutralization processes, ensuring the directionality and continuity of electrostatic dissipation. The grounding coupling torque depends on the potential difference between the ion wind and the conductive grounding layer. The conductive grounding layer is located below the conveying path, maintaining a fixed distance from the surface of the packaging box. Its interior is composed of layers of highly conductive copper foil and carbon fiber conductive mesh, with an anti-oxidation coating on the surface to stabilize conductivity. Charged particles released by the ion dissipation traction mechanism pass through the detection area under the guidance of the airflow, reacting with suspended static charges in the air. Some charges are directly neutralized, while the remaining uncombined charges are guided to the grounding layer by the airflow. After the detection device has been running for a period of time, a stable potential gradient forms near the grounding layer. To prevent reverse potential accumulation, the control unit adjusts the voltage amplitude of the ion emission electrode and the airflow velocity to change the magnitude of the coupling torque. When the cross-frame stable anchor point sequence reflects increased brightness fluctuations at the image edges, it indicates rising electrostatic interference. The control device increases the electrode voltage, accelerating the movement of charged particles and increasing the coupling torque, thereby speeding up charge migration to the ground layer. When the anchor point sequence shows stable image light intensity, the control device decreases the voltage output, weakening the coupling torque to prevent excessive ionization in the air. Through this synchronous adjustment, the charge flow direction throughout the detection area remains stable, forming a unidirectional potential release channel from above the imaging area to the ground layer, achieving continuous dissipation of electrostatic energy.

[0083] After the ion wind frequency shift output and ground coupling torque balance adjustment achieve stable operation, the exposure duty cycle of the machine vision inspection device is dynamically adjusted to establish a real-time closed-loop relationship between imaging illumination intensity and electrostatic discharge state. Exposure duty cycle refers to the proportion of time the photosensitive element is in a photosensitive state during the exposure cycle. When electrostatic interference is enhanced, the exposure time needs to be shortened to avoid overexposure because discharge flashes may cause sudden changes in illumination; conversely, when the electrostatic field is stable, the exposure time can be appropriately extended to improve image brightness and clarity. Before each frame exposure, the control unit reads the latest state of the cross-frame stable anchor point sequence and determines the exposure time parameters under the current environment by combining the ion wind output intensity and ground torque value. When the ion wind is in a high output state, the exposure duty cycle is correspondingly reduced, allowing the camera to complete illumination in a shorter time; when the ion wind is in the breathing decay stage, the exposure duty cycle is increased, allowing for more complete image acquisition. The magnitude of the exposure time adjustment is automatically corrected based on the illumination balance table, which is composed of brightness statistics from consecutive frames. After each frame image is acquired, the control unit updates the exposure parameters in real time. By dynamically coordinating the ion wind intensity, coupling torque, and exposure time, a mutually constraining cyclical system is formed between electrostatic discharge, airflow stabilization, and optical acquisition. The entire closed loop continuously repeats over time: the anchor point sequence provides optical state feedback, the breathing ion wind dissipates charge, the grounding layer conducts potential current, and the exposure control unit corrects the illumination response. Through this continuous process, packaging carton printing quality inspection can maintain stable imaging over long periods in high-speed production environments, preventing brightness drift, shadow errors, and image jitter caused by electrostatic accumulation, achieving high-precision and high-stability printing quality inspection and optical recognition.

[0084] This invention introduces a coordinated control mechanism of electrostatic sensing and timing avoidance on the conveying path for packaging carton printing quality inspection. This effectively isolates electrostatic discharge behavior both temporally and spatially, preventing instantaneous flashes from entering the exposure area and interfering with imaging. By setting a micro-delay window within the exposure cycle and combining it with a polarization suppression structure, high-brightness transient signals in the optical path are suppressed, resulting in a more uniform and stable light field distribution in the reflected image from the printed surface. This significantly improves the image clarity and reliability of machine vision inspection, ensuring the continuity and stability of the inspection data.

[0085] This invention achieves dynamic coupling between imaging stability and electrostatic dissipation by constructing a closed-loop control process centered on a cross-frame stable anchor point sequence. The breathing-type ion wind traction mechanism maintains the potential balance of the detection environment under frequency misalignment drive, and the exposure duty cycle is adjusted in real time according to the imaging state, thus creating a complementary relationship between electrostatic dissipation and optical acquisition. Through this dynamic feedback method, the detection process maintains stable illumination and interference-free images even under high-speed production conditions, achieving high-precision identification of printing quality and stable imaging control for long-term operation.

[0086] This invention provides, for example Figure 2 The packaging cardboard box printing quality inspection system shown includes an electrostatic sensing module, a timing control module, an optical purification module, an image extraction module, and a closed-loop control module.

[0087] The electrostatic sensing module sets up an electrostatic sensing strip on the conveying path for the printing quality inspection of packaging paper boxes, collects discharge flash signals and calculates the discharge trajectory, determines the starting position of each discharge event, and generates a list of electrostatic starting point anchor points.

[0088] The timing control module, based on the list of electrostatic starting point anchor points, inserts a micro-delay window into the exposure timing of the machine vision inspection device, limiting the time interval of the discharge flash signal to the non-exposure interval, thus forming a timing avoidance trajectory.

[0089] The optical purification module arranges polarization suppression curtains on the imaging plane according to the timing avoidance trajectory, performs polarization separation on the reflected light signal entering the exposure area, and generates a reflection purification mask to reduce the interference reflection of the discharge spot.

[0090] The image extraction module uses a reflection purification mask to perform a weighted spreading operation on the image edge channel to extract the contour features of the real printed image from optical noise and establish a stable anchor point sequence across frames.

[0091] The closed-loop control module, based on the cross-frame stable anchor point sequence, drives the breathing ion dissipation traction mechanism, controls the ion wind and ground coupling torque in a frequency-shifting manner, dynamically adjusts the exposure duty cycle, and constructs a closed-loop control process for electrostatic suppression and imaging stability.

[0092] The packaging paper box printing quality inspection method based on machine vision inspection provided in this embodiment of the invention is implemented by the above-mentioned packaging paper box printing quality inspection system based on machine vision inspection. For details of the specific methods and processes of the packaging paper box printing quality inspection system based on machine vision inspection, please refer to the above-mentioned embodiment of the packaging paper box printing quality inspection method based on machine vision inspection, which will not be repeated here.

[0093] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for inspecting the printing quality of packaging paper boxes based on machine vision inspection, characterized in that, Includes the following steps: An electrostatic sensing strip is set up on the conveying path for quality inspection of printed paper boxes to collect discharge flash signals and calculate the discharge trajectory, determine the starting position of each discharge event, and generate a list of electrostatic starting point anchor points. Based on the list of electrostatic starting point anchors, a micro-delay window is inserted into the exposure sequence of the machine vision inspection device to limit the time interval of the discharge flash signal to the non-exposure interval, thus forming a timing avoidance trajectory. Based on the timing avoidance trajectory, a polarization suppression curtain is arranged on the imaging plane to perform polarization separation on the reflected light signal entering the exposure area, generating a reflection purification mask to reduce the interference reflection of the discharge spot. A reflection cleansing mask is used to perform a weighted spreading operation on the image edge channel to extract the contour features of the real printed image from optical noise and establish a stable anchor point sequence across frames. Based on the cross-frame stable anchor sequence, a breathing-type ion dissipation traction mechanism is driven to control the ion wind and ground coupling torque in a frequency-shifting manner, dynamically adjust the exposure duty cycle, and construct a closed-loop control process for electrostatic suppression and imaging stability.

2. The method for inspecting the printing quality of packaging paper boxes based on machine vision inspection according to claim 1, characterized in that, The steps to generate the list of electrostatic initiation anchor points are as follows: Electrostatic sensing strips are laid at key locations along the conveying path for quality inspection of printed packaging boxes. The electrostatic sensing strips are distributed along the entire conveying direction of the packaging boxes to collect electrostatic potential change signals generated during the conveying process in real time. After the electrostatic signal acquisition is completed, the acquired potential change signal is segmented and extracted. The discharge flash signal is extracted by time correlation comparison and the time point and region of the discharge event are determined. After extracting the discharge flash signal, the discharge trajectory is calculated based on the distribution location of the electrostatic sensing strip, combined with the spatial layout of the transmission path, to determine the propagation direction and coverage of the released charge. After the discharge trajectory calculation is completed, the starting position of each discharge event is determined based on the concentration of the trajectory distribution and the duration characteristics. The discharge direction, duration and energy change amplitude are recorded, and a list of electrostatic starting point anchor points is generated.

3. The method for inspecting the printing quality of packaging paper boxes based on machine vision inspection according to claim 2, characterized in that, The steps for generating a time-series obstacle avoidance trajectory are as follows: Based on the time attributes and spatial distribution information in the list of electrostatic initiation anchor points, an exposure timing reference for the machine vision inspection device is established, and the time window of each discharge event is time-aligned with the exposure trigger signal. After establishing the exposure timing reference, the start time of the exposure trigger signal is slightly delayed according to the time distribution pattern of the discharge event, so that the exposure start time avoids the high-energy release period of the discharge signal. After the exposure trigger signal completes the delay adjustment, the time difference of consecutive exposure cycles is smoothed to form a continuous time delay curve to construct a complete timing avoidance trajectory. After the timing avoidance trajectory is formed, the exposure cycle length, exposure duration, image transmission interval and exposure start phase are updated synchronously according to the trajectory change pattern to establish a timing avoidance network to isolate the discharge light signal.

4. The method for inspecting the printing quality of packaging paper boxes based on machine vision inspection according to claim 3, characterized in that, The micro-delay adjustment is achieved through the time delay adjustment unit. The time delay adjustment unit receives the control parameters in the timing map table to delay the rising edge of the exposure trigger signal, so that the exposure start time lags behind the energy release period of the discharge flash signal. This keeps the exposure action and the electrostatic discharge process separated on the time axis, ensuring the illumination stability and imaging uniformity of the image acquisition process.

5. The method for inspecting the printing quality of packaging paper boxes based on machine vision inspection according to claim 3, characterized in that, The steps for generating a reflection cleansing mask are as follows: A polarization suppression curtain is placed at the front end of the imaging plane according to the timing avoidance trajectory. The polarization suppression curtain consists of a transparent substrate layer, a polarization filter layer and a reflection reduction layer. After the polarization anti-reflection curtain is installed, the polarization orientation is dynamically controlled by a micro-driving component set at the edge of the polarization filter layer, so that the polarization direction alternates between the exposure range and the non-exposure range to filter out the reflection of the discharge spot. After polarization orientation adjustment is completed, the reflected light signal passing through the polarization suppression curtain is spatially separated by the optical guidance component, so that the high-energy reflected light is guided to the light absorption groove and the effective reflected light is retained to enter the imaging plane. After completing the spatial separation of the optical signal, a reflective purification mask is generated on the imaging plane. The reflective purification mask is formed by combining a transparent transmission layer and a light-absorbing filter layer.

6. The method for inspecting the printing quality of packaging paper boxes based on machine vision inspection according to claim 5, characterized in that, The polarization suppression curtain maintains a fixed tilt angle with the imaging plane. The orientation of the polarization filter layer is perpendicular to the polarization direction of the discharge spot when switching between the exposure and non-exposure zones. The light-transmitting aperture array of the reflection purification mask is consistent with the polarization direction.

7. The method for inspecting the printing quality of packaging paper boxes based on machine vision inspection according to claim 5, characterized in that, The steps for generating a cross-frame stable anchor sequence are as follows: After completing the optical filtering of the reflective purification mask, the purified image signal is divided into edge channels. During the signal output stage of the photosensitive element, the imaging plane is divided into parallel image acquisition channels and a light intensity compensation band is set to enhance the edge response. After forming the edge channels, the light field intensity of each channel is reweighted and spread using a reflective purification mask. The edge features of the printed pattern are enhanced by the diffusion and redistribution of light energy through the light-transmitting aperture array. After completing the light intensity weighting, the outline shape of the printed pattern is extracted by continuously scanning the light intensity change trend of each edge channel, and a transition area is set between the channels to ensure the continuity of the outline. After extracting the complete outline of the printed image, a cross-frame stable anchor point sequence is established based on the time sequence of consecutive exposure frames. The anchor point motion trajectory is formed by continuously matching the anchor point positions, thus maintaining the stability of image recognition.

8. The method for inspecting the printing quality of packaging paper boxes based on machine vision inspection according to claim 7, characterized in that, During the weighted spreading process, the aperture and spacing of the light-transmitting hole array of the reflective purification mask change step by step according to the light field brightness gradient. The angle of the light-transmitting hole wall is consistent with the direction of the channel boundary, so that the light energy diffuses along the edge direction and forms an energy superposition area in the longitudinal direction.

9. The method for inspecting the printing quality of packaging paper boxes based on machine vision inspection according to claim 7, characterized in that, The following are the steps of the closed-loop control process based on a cross-frame stable anchor sequence-driven breathing-type ion dissipation traction mechanism, which controls the coupling torque between the ion wind and grounding in a frequency-shifting manner and dynamically adjusts the exposure duty cycle: After the cross-frame stable anchor point sequence is established, the spatial position coordinates, time series distribution and brightness fluctuation trend in the anchor point sequence are converted into electrostatic environment change feedback signals, which are used to drive the breathing ion dissipation traction mechanism. After the breathing ion dissipation traction mechanism receives the anchor sequence input signal, it adjusts the ion wind output cycle and exposure sequence through the frequency misalignment control device so that the ion wind and the exposure process are misaligned on the time axis and electrostatic neutralization is achieved. While the ion wind is output at a different frequency, the spatial balance between ion release and electrostatic neutralization is controlled by adjusting the grounding coupling torque, so that the charge flow direction remains stable and a unidirectional potential release channel is formed. After the ion wind frequency offset output and ground coupling torque balance adjustment form a stable operation, the exposure duty cycle is dynamically adjusted to form a closed-loop coordinated control between imaging illumination intensity and electrostatic discharge state.

10. A packaging paper box printing quality inspection system based on machine vision inspection, used to implement the packaging paper box printing quality inspection method based on machine vision inspection as described in any one of claims 1-9, characterized in that, It includes an electrostatic sensing module, a timing control module, an optical purification module, an image extraction module, and a closed-loop control module: The electrostatic sensing module sets up an electrostatic sensing strip on the conveying path for the printing quality inspection of packaging paper boxes, collects discharge flash signals and calculates the discharge trajectory, determines the starting position of each discharge event, and generates a list of electrostatic starting point anchor points. The timing control module, based on the list of electrostatic starting point anchor points, inserts a micro-delay window into the exposure timing of the machine vision inspection device, limiting the time interval of the discharge flash signal to the non-exposure interval, thus forming a timing avoidance trajectory. The optical purification module arranges polarization suppression curtains on the imaging plane according to the timing avoidance trajectory, performs polarization separation on the reflected light signal entering the exposure area, and generates a reflection purification mask to reduce the interference reflection of the discharge spot. The image extraction module uses a reflection purification mask to perform a weighted spreading operation on the image edge channel to extract the contour features of the real printed image from optical noise and establish a stable anchor point sequence across frames. The closed-loop control module, based on the cross-frame stable anchor point sequence, drives the breathing-type ion dissipation traction mechanism, controls the ion wind and ground coupling torque in a frequency-shifting manner, dynamically adjusts the exposure duty cycle, and constructs a closed-loop control process for electrostatic suppression and imaging stability.

Images