A cargo label recognition method based on smart logistics warehouse management

Abstract

This invention relates to the field of label recognition technology, and more particularly to a method for identifying cargo labels based on smart logistics warehouse management. The method includes the following steps: deploying a recognition device in the shelving area, comprising a radio frequency antenna, infrared sensing, and a light intensity acquisition module; real-time monitoring of cargo displacement by infrared sensing triggering the sequential activation of the radio frequency antenna to form a time difference; calculating spatial displacement based on the difference and controlling the light intensity module to enhance local illumination to obtain a high-contrast label image; extracting edges and characters from the image and matching them with the radio frequency code to generate a cargo identification confirmation signal, achieving high-precision dynamic recognition. This invention integrates radio frequency identification, infrared sensing, and optical acquisition technologies to achieve cargo displacement-triggered recognition, time-series differential analysis, and reflection signal enhancement, thereby significantly improving the accuracy, stability, and intelligence level of cargo label recognition in smart warehouses.

Description

Technical Field

[0001] This invention relates to the field of label recognition technology, and in particular to a method for identifying cargo labels based on intelligent logistics warehouse management. Background Technology

[0002] In modern smart logistics warehouse management, cargo tag identification technology is a key component for achieving automated cargo management, precise positioning, and efficient scheduling. Rapid identification of cargo tags can significantly improve the intelligence level of warehousing operations, reduce manual intervention, and enhance logistics efficiency.

[0003] However, existing technologies mostly rely on a single radio frequency identification (RFID) method, which is easily affected by signal blockage, reflection interference, and blind spots in environments with dense multi-shelf structures, leading to problems such as decreased identification accuracy, positioning lag, or repeated identification. At the same time, the lack of a sensing mechanism that links with the dynamic status of goods makes it impossible to achieve synchronous response and optical-assisted identification when goods are moving, thus limiting the real-time performance and stability of the system. Summary of the Invention

[0004] Therefore, it is necessary to provide a cargo label identification method based on intelligent logistics warehouse management to solve at least one of the above-mentioned technical problems.

[0005] To achieve the above objectives, a cargo label identification method based on intelligent logistics warehouse management is provided, the method comprising the following steps:

[0006] Step S1: Deploy multiple sets of label recognition devices for near-field sensing in the warehouse shelf area. Each set of label recognition devices includes a radio frequency identification antenna, an infrared sensing unit, and a light intensity acquisition module. The radio frequency identification antenna is used to acquire the radio frequency response signal of the goods label, and the infrared sensing unit is used to detect the presence status of the goods on the shelf and trigger the corresponding recognition antenna to turn on.

[0007] Step S2: When the infrared sensing unit detects the displacement of the goods, it sequentially activates the radio frequency identification antennas in adjacent areas, so that the tag response signal forms a time series response difference between adjacent areas;

[0008] Step S3: Calculate the spatial displacement vector of the cargo label based on the time series response difference, and control the light intensity acquisition module to start local illumination according to the change of the spatial displacement vector to enhance the reflection signal on the label surface, thereby generating a high-contrast label image;

[0009] Step S4: Sharpen the edges and locate the characters in the high-contrast label image to extract the character encoding information of the goods label;

[0010] Step S5: Physically match the character encoding information with the electronic tag number corresponding to the RFID signal. If the matching results are consistent, generate a cargo identification confirmation signal.

[0011] The present invention has the following beneficial effects:

[0012] I. By combining Radio Frequency Identification (RFID) technology with infrared sensing and optical acquisition technology, high-precision dynamic identification and positioning of stored goods are achieved. Compared with traditional methods that rely solely on RFID, this invention uses an infrared sensing unit to achieve real-time detection of the presence and displacement of goods. It can automatically trigger the orderly activation of RFID antennas in adjacent areas the instant the goods move, thereby forming time-series response difference data. This enables accurate displacement tracking and identification of goods within the storage space, improving the response speed and spatial resolution of the identification system.

[0013] II. By analyzing the time-series difference and rate variation trends of RF response signals within adjacent identification areas, a dynamic signal trend modeling method based on the alternating characteristics of acceleration and deceleration phases is proposed. This method can effectively suppress multipath interference and tag signal overlap in complex warehousing environments. Through time-series response difference correction processing, highly stable time-domain feature data is generated, significantly improving the accuracy of main response antenna selection and the robustness of signal recognition. In particular, by setting the sampling time interval and rate threshold range, refined detection of cargo movement details can be achieved, providing a reliable temporal basis for tag recognition in dynamic scenarios.

[0014] Third, by controlling local illumination through spatial displacement vector changes, adaptive enhancement and directional correction of the reflected light signal from the cargo label are achieved. Through multi-level processing steps such as photoelectric signal normalization, time integration, and sensitivity conversion, directional correction reflection intensity data related to the incident angle can be obtained, and high-contrast label images can be generated based on this. This process effectively improves the clarity and accuracy of character recognition on the label surface, solving the recognition difficulties caused by low light, occlusion, and uneven reflection in warehouses. In summary, this invention not only achieves intelligent, dynamic, and high-precision cargo label recognition, but also significantly enhances the stability and practicality of the system in complex warehousing environments. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating the steps of a cargo label identification method based on intelligent logistics warehouse management.

[0016] Figure 2 for Figure 1 A detailed flowchart illustrating the implementation steps of step S2.

[0017] Figure 3This is a schematic diagram of a label recognition device for a cargo label recognition method based on intelligent logistics warehouse management according to this application;

[0018] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0019] The technical method of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0020] Furthermore, the accompanying drawings are merely illustrative of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0021] It should be understood that although the terms "first," "second," etc., may be used herein to describe various units, these units should not be limited by these terms. These terms are used merely to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0022] To achieve the above objectives, please refer to Figures 1 to 3 A method for identifying cargo tags based on intelligent logistics warehouse management, the method comprising the following steps:

[0023] Step S1: Deploy multiple sets of label recognition devices for near-field sensing in the warehouse shelf area. Each set of label recognition devices includes a radio frequency identification antenna, an infrared sensing unit, and a light intensity acquisition module. The radio frequency identification antenna is used to acquire the radio frequency response signal of the goods label, and the infrared sensing unit is used to detect the presence status of the goods on the shelf and trigger the corresponding recognition antenna to turn on.

[0024] In one embodiment, reference may be made to Figure 3To achieve high-precision label recognition of goods within the warehouse shelving area, multiple sets of label recognition devices are evenly deployed on each layer of the warehouse shelving and along both sides of the aisles. Each set of devices is fixedly installed on the shelf support columns or beams, and its sensing distance covers the front edge of the shelf to the front of the goods placement area, ensuring comprehensive coverage of goods of different heights and volumes.

[0025] Each tag identification device includes a radio frequency identification (RFID) antenna, an infrared sensing unit, and a light intensity acquisition module. The RFID antenna is connected to the identification main controller via a coaxial feed line and is used to transmit identification interrogation signals within a preset frequency band (e.g., 920MHz to 925MHz) and receive the reflected response signals from the product tags. By employing linearly polarized or circularly polarized antenna structures, the signal coverage angle can be adjusted according to the shelf placement direction, thereby reducing inter-antenna interference.

[0026] The infrared sensing unit is located at the front end of the RFID antenna and is used to detect the presence of goods on the shelf in real time. When the infrared sensor detects a reflected signal from an object in front of it and the reflection intensity exceeds a set threshold, the system determines that there are goods in that location and sends a trigger signal to the identification controller. The controller then activates the corresponding RFID antenna, enabling on-demand activation and reducing antenna power consumption during idle operation.

[0027] In addition, the light intensity acquisition module is used to sense the ambient light intensity in the shelving area. The main controller automatically adjusts the transmission power and receiving sensitivity of the infrared sensing unit based on the light intensity data to adapt to changes in day and night or fluctuations in lighting conditions within the warehouse, ensuring the stability and accuracy of infrared detection under different lighting conditions.

[0028] In a preferred embodiment, the radio frequency identification antenna and infrared sensing unit of each group of label identification devices are connected to the centralized control terminal through a bus structure. The control terminal dynamically controls the start and stop of the identification devices in different shelf areas by real-time monitoring of the infrared sensing signals, so as to realize zoned identification and energy-saving operation.

[0029] With the above configuration, this embodiment can realize automatic detection of the presence status of goods and adaptive triggering of the label recognition device in the warehouse shelf area, reduce redundant scanning operations, improve label recognition efficiency and system operation stability, and provide an accurate perception basis for subsequent goods label matching and warehouse information updates.

[0030] Step S2: When the infrared sensing unit detects the displacement of the goods, it sequentially activates the radio frequency identification antennas in adjacent areas, so that the tag response signal forms a time series response difference between adjacent areas;

[0031] In one embodiment, when the infrared sensing unit detects displacement of goods on the shelf—that is, when the intensity or echo time of the infrared reflected signal continuously changes beyond a set threshold—it determines that a goods movement event has occurred at that location. The infrared sensing unit transmits the detection signal to the main control processing unit in real time. Based on the direction and amplitude of the goods displacement, the main control unit determines the range of adjacent recognition areas that need to be activated.

[0032] Subsequently, the main control unit activates the RFID antennas in the target area and its adjacent areas sequentially according to the preset timing logic. When activated, each antenna transmits an RFID interrogation signal at microsecond intervals and receives the response signal from the cargo tag. Due to the slight displacement of the cargo position, the relative spatial distance between the tag and each antenna, as well as the signal propagation path length, changes accordingly, resulting in slight differences in amplitude, phase, and time delay in the response signals received by adjacent antennas.

[0033] The tag response signals acquired by the sequentially activated RFID antennas are time-stamped, and a time-series signal set is generated in the main control unit. By comparing the response signal strength and arrival time of different antennas in adjacent activation periods, the time-series response difference is calculated, thereby obtaining the response characteristic trajectory of the cargo tag under spatial location changes.

[0034] In a preferred embodiment, the main control unit is further provided with a dynamic threshold adjustment mechanism to automatically adjust the signal sampling sensitivity according to the intensity of ambient electromagnetic interference. When an increase in background noise is detected, the activation interval between adjacent antennas is appropriately extended and the average number of signal samples is increased to enhance the stability and discriminative power of the response difference.

[0035] Furthermore, to prevent signal overlap caused by simultaneous displacement of multiple adjacent storage locations, this embodiment introduces area identification coding between each group of RFID antennas. The main control unit performs area identification separation on the acquired signals based on the coding sequence, ensuring that the response differences between different storage locations can be accurately distinguished.

[0036] Step S3: Calculate the spatial displacement vector of the cargo label based on the time series response difference, and control the light intensity acquisition module to start local illumination according to the change of the spatial displacement vector to enhance the reflection signal on the label surface, thereby generating a high-contrast label image;

[0037] In one embodiment, after receiving the time-series response difference acquired by the RFID antennas in adjacent areas, the main control unit performs signal analysis and position inversion. First, by calculating the arrival time difference and amplitude difference of the response signals of adjacent antennas, the relative displacement components of the tag in the three-dimensional coordinate system are determined, and the spatial displacement vector of the cargo tag is generated accordingly.

[0038] Spatial displacement vectors are used to represent the direction and distance of the tag's movement within the detection cycle. The vector direction reflects the movement trend of the goods, and the vector magnitude corresponds to the displacement of the goods. The main control unit uses continuously acquired time-series response difference data to perform multi-frame fitting and dynamic smoothing of the tag's movement, filtering out errors caused by occasional signal fluctuations, thereby obtaining a stable and high-precision spatial displacement vector sequence.

[0039] When the rate of change of the spatial displacement vector exceeds a set threshold, it indicates that the goods may be under manual handling, rotation, or rearrangement. At this point, the light intensity acquisition module is triggered to enter local lighting mode. The light intensity acquisition module includes several adjustable light source units distributed above, on the sides, and behind the shelf, which can automatically select the optimal lighting angle based on the direction of displacement.

[0040] Based on the direction information of the spatial displacement vector, the main control unit calculates the local spatial azimuth angle of the label and controls the corresponding light source unit to start to achieve directional supplemental lighting. When the label moves to an area with insufficient lighting or reflective interference, the illumination intensity and light source angle are automatically adjusted so that the light forms a near-perpendicular reflection path with the label surface to improve the recognizability of the reflected signal.

[0041] During the operation of the light intensity acquisition module, its internal photosensitive detection element synchronously acquires the reflected light signal from the tag surface and associates the reflected light intensity data with the radio frequency response signal. The two types of signals are then fused to form a high-contrast tag image dataset.

[0042] To ensure image sharpness, a short-exposure capture strategy is executed after each local illumination. By limiting the exposure time and light intensity threshold, a significant brightness difference is made between the label character area and the background area, thereby obtaining a label image with sharp reflection boundaries and high image contrast.

[0043] In a preferred embodiment, the light intensity acquisition module is also equipped with an automatic light decay compensation algorithm, which can automatically adjust the output power according to the brightness decay rate of consecutive frame images to prevent image quality degradation due to light source aging or changes in ambient light.

[0044] Step S4: Sharpen the edges and locate the characters in the high-contrast label image to extract the character encoding information of the goods label;

[0045] In one embodiment, after obtaining the high-contrast label image, the image data is first input to the image preprocessing module. The preprocessing module eliminates local brightness unevenness caused by differences in illumination angle or reflection through filtering and normalization operations, ensuring the stability of subsequent edge feature extraction. This module preferably employs an adaptive histogram equalization algorithm based on local contrast enhancement to dynamically reconstruct the overall grayscale distribution of the image, thereby enhancing the brightness contrast between the label character area and the background area.

[0046] The preprocessed image is fed into the edge sharpening unit. This unit calculates the edge response of the image using a multi-scale edge detection operator and performs local gradient enhancement on the character contour region. Specifically, it extracts grayscale gradient data in the horizontal and vertical directions at the edges of the label characters and adjusts the sharpening weights based on the gradient magnitude difference, making the character boundaries clearer.

[0047] During edge enhancement, morphological dilation and erosion are simultaneously employed to eliminate isolated noise points caused by reflections or stains on the label surface. In the final sharpened image, the grayscale difference between character strokes and the background is further amplified, providing a reliable feature base for subsequent character localization.

[0048] Subsequently, the image analysis module proceeds to the character localization stage. This module first identifies continuous high-contrast regions in the image using a connected component detection algorithm, and then filters candidate regions that match the character's morphological characteristics based on region area, aspect ratio, and geometric constraints of the boundary rectangle. For cases where characters are joined or broken, local repair is performed using projection segmentation and boundary reconnection techniques to restore the complete shape of the characters.

[0049] In a preferred embodiment, spatial arrangement analysis of candidate character regions is also performed in conjunction with the label's structural template information. By comparing character spacing and row / column alignment features, non-character pattern regions (such as trademarks or borders) can be automatically eliminated. Once the candidate regions pass template matching verification, the final character positioning result can be determined.

[0050] For the located character region, the character extraction unit uses the binarized image within the region to extract the character contour. Through dynamic thresholding and edge closure correction, the outer boundary of the character region is accurately described as a vector contour and encoded into a standardized character image patch.

[0051] Finally, the encoding and recognition module serializes and sorts the extracted character image blocks, and completes the recognition of character encoding information based on a pre-trained character recognition model. The recognition results include the unique label number, batch number, or barcode information, and are output in digital form to the central database for cargo tracking and warehouse management.

[0052] Step S5: Physically match the character encoding information with the electronic tag number corresponding to the RFID signal. If the matching results are consistent, generate a cargo identification confirmation signal.

[0053] In one embodiment, after obtaining the tag character encoding information output by the character recognition module, the character encoding information is first input to the central identification control unit. The identification control unit communicates with the RFID module via an internal bus and reads the electronic tag number information corresponding to the current shelf area from the RFID module. This electronic tag number was obtained from a previous RFID antenna scan and has been stored in a local buffer.

[0054] To ensure the timeliness and accuracy of the matching process, the character encoding information and the radio frequency identification information are standardized before matching. This process includes character set unification, length correction, and redundancy bit verification to ensure that the two identification data are structurally consistent and to eliminate matching errors caused by encoding differences.

[0055] Next, the physical matching phase begins. The matching module uses a dual-channel comparison algorithm to extract the serial number data features from the character encoding channel and the radio frequency signal channel, respectively, and establishes a one-to-one mapping table in memory. By comparing the byte sequence, data check bits, and logical address information of the serial number features, it determines whether the two belong to the same cargo entity.

[0056] If a perfect match is detected during the comparison process, a goods identification confirmation signal is generated. This confirmation signal includes the goods' unique identification number, identification timestamp, identification device number, and the location information on the shelf. The confirmation signal is transmitted to the warehouse management main control system via a wired or wireless communication interface for updating the goods' inbound and outbound status and synchronizing inventory.

[0057] In another embodiment, when there is a partial inconsistency between the character encoding information and the RFID number, the system automatically triggers a secondary matching verification mechanism. This mechanism includes comparison and correction based on signal strength and time synchronization, reactivating the local RFID antenna to reacquire the electronic tag number signal, and performing a second matching based on historical cached data to prevent misjudgments caused by signal interference or reflection.

[0058] If discrepancies persist after secondary verification, an abnormal match record will be generated and marked as awaiting manual review; no confirmation signal will be generated at this time. This abnormal match information will be uploaded to the monitoring platform for subsequent tag anomaly investigation and device self-check operations.

[0059] In a preferred embodiment, the identification confirmation signal can also be linked with the infrared sensing status of the shelf area. If the identification confirmation signal remains valid after the goods are moved, the goods are considered to have been successfully identified. If the sensing status does not match the confirmation signal, the label re-identification process will be automatically initiated.

[0060] As an example of the present invention, reference is made to Figure 2 As shown, step S2 in this example includes:

[0061] Step S21: When the infrared sensing unit detects that the goods in the shelf area have moved, the displacement detection signal output by the sensing unit is collected to generate goods displacement trigger data;

[0062] Step S22: Determine the adjacent identification area range of the cargo based on the cargo displacement trigger data, and select the corresponding radio frequency identification antenna group as the target identification group;

[0063] Step S23: Activate each radio frequency identification antenna in the target identification group in sequence to form a continuous radio frequency transmission cycle in the time series;

[0064] Step S24: Receive the radio frequency response signal returned by the cargo tag in each radio frequency transmission cycle and generate the corresponding tag response dataset;

[0065] Step S25: Arrange the label response datasets between adjacent regions in chronological order and calculate the difference between adjacent data to obtain the time series response difference.

[0066] In one embodiment, when the infrared sensing unit detects displacement of goods within the shelf area, the sensing unit outputs a continuous displacement detection signal. This signal includes a trigger timestamp, trigger point coordinates, and trigger amplitude. The control unit collects this signal in real time and filters and denoises it according to a preset trigger threshold to generate goods displacement trigger data, which is used to identify the occurrence of goods movement events.

[0067] Based on the cargo displacement trigger data and the spatial layout information of the warehouse shelves, the adjacent identification area range where the cargo is located is determined. The control unit selects the corresponding RFID antenna group in this area as the target identification group through a mapping table, and loads the antenna group number and spatial coordinate information into the identification scheduling module to ensure that subsequent RFID operations are performed on the target area.

[0068] The RFID antennas within the target identification group are controlled by sequential activation. Specifically, the control unit sends activation commands sequentially according to a preset RFID transmission cycle, causing each RFID antenna to form a continuous transmission sequence in time. Each antenna emits an RFID signal upon activation, and simultaneously records the antenna's transmission status, transmission power, and activation time in a local cache for subsequent data comparison and time series analysis.

[0069] During each radio frequency transmission cycle, the corresponding cargo tag responds with a returned radio frequency signal. The receiving module acquires the returned signal in real time and extracts information such as tag number, signal strength, phase, and return time to form a tag response dataset. The dataset contains a complete scan record of each antenna on the cargo tag to support subsequent spatial positioning and motion analysis.

[0070] The tag response datasets of adjacent identification areas are arranged in chronological order. The control unit calculates the response difference between consecutive frames to obtain a time-series response difference. This time-series response difference reflects the displacement of goods between adjacent areas and provides a data basis for subsequent spatial displacement vector calculations. The difference calculation includes analysis of signal strength difference, time of arrival difference, and phase change, and the direction and speed of goods movement can be determined based on the magnitude of the difference.

[0071] In a preferred embodiment, the time series response difference is also compared with historical scanning data to filter out false response differences caused by environmental interference or radio frequency reflection, thereby ensuring that the generated time series response difference is accurate and reliable, and can provide stable data support for the generation of high-contrast label images and subsequent recognition steps.

[0072] Preferably, step S25 includes:

[0073] Extract the initial static timestamps and dynamic timestamps of the label response dataset between adjacent regions to obtain the initial time series and dynamic time series;

[0074] Based on the sorting and comparison results of the initial time series and the dynamic time series, the RFID antenna with the highest signal strength in the adjacent area is selected as the main response antenna;

[0075] Extract three consecutive time points from the time series of the main response antenna and calculate the average interval between each time point;

[0076] The average interval is synchronously compared with the dynamic time series of adjacent antenna groups to obtain time series differential data;

[0077] Time series response difference data is generated based on the rate of change of time series difference data.

[0078] In one embodiment, time information is extracted from the tag response dataset between adjacent areas. The control unit acquires the initial static timestamp of each antenna and the dynamic timestamp collected during the scanning process, and organizes them into an initial time series and a dynamic time series, respectively, to identify the occurrence time and time sequence of each tag response signal.

[0079] The initial and dynamic time series were sorted and compared, and the RFID antenna with the strongest response in the adjacent recognition area was selected as the main response antenna based on the signal strength ranking results. The main response antenna was used in subsequent analysis to determine the spatial displacement direction and amount of the cargo tag, thereby improving recognition accuracy.

[0080] For the main response antenna, the control unit extracts three consecutive time points from its response time series and calculates the average interval between each time point to obtain the time sampling characteristics of the main response antenna. The average interval is used to measure the sampling stability and response continuity of the radio frequency signal, ensuring the reliability of the time series analysis.

[0081] The average time interval of the main response antenna is synchronously compared with the dynamic time series of adjacent antenna groups, and time series differential data is obtained through time matching calculation. The differential data reflects the actual displacement delay and response changes of the cargo between adjacent areas, and is used to further analyze the direction, speed and distance of cargo movement.

[0082] Time-series response difference data is generated based on the rate of change of time-series difference data. Specifically, the difference values ​​at consecutive time points are statistically analyzed and filtered in chronological order to obtain the response change amplitude and trend of each tag in adjacent regions. This time-series response difference data is used in subsequent steps to calculate the spatial displacement vector of the goods and provides a data foundation for the generation of high-contrast tag images.

[0083] Preferably, generating time series response difference data based on the rate of change of time series difference data includes:

[0084] Trend analysis is performed on the rate of change of time series difference data to identify acceleration and deceleration phases.

[0085] The average rate of change is extracted within the alternating intervals of acceleration and deceleration, and the rate difference between each interval is calculated.

[0086] Rate difference is mapped onto time series coordinates to form a time series change curve;

[0087] The dynamic offset value of the response signal is determined by the slope change of the time series curve, and the dynamic offset value is weighted and fused with the original time series difference data to generate the corrected time series response difference data.

[0088] In one embodiment, trend analysis is performed on the time-series difference data. The control unit identifies acceleration and deceleration segments in the data based on the rate of difference change at consecutive time points, i.e., intervals where the difference value increases or decreases significantly over time, in order to capture the dynamic characteristics of the cargo tag during its movement between adjacent areas.

[0089] For the identified acceleration and deceleration segments, the average rate of change is extracted within each alternating interval, and the rate difference between adjacent intervals is calculated. By statistically analyzing the rate differences between each interval, the control unit can quantify the acceleration and deceleration changes of the tag's response signal during spatial movement, providing a data basis for subsequent dynamic correction.

[0090] The control unit maps the rate difference to a time series coordinate system, forming a continuous time series variation curve. This curve reflects the dynamic change trend of the tag's response signal over time, including acceleration, deceleration, and steady-state phases, allowing the system to intuitively assess the dynamic offset of the response signal at different time points.

[0091] The dynamic offset value of the response signal is determined based on the change in the slope of the time series curve. A large change in slope indicates a significant acceleration or deceleration of the response signal, and a dynamic offset compensation value corresponding to the time point is generated accordingly.

[0092] Finally, the control unit performs a weighted fusion of the dynamic offset value and the original time series difference data to obtain the corrected time series response difference data. During the weighted fusion process, the weights can be dynamically adjusted according to the slope change, ensuring that higher correction weights are given in high dynamic change sections, while the original signal characteristics are preserved in stable sections, thereby generating accurate, continuous, and corrected data that reflects the actual displacement state of the cargo.

[0093] Preferably, trend analysis is performed based on the rate of change of the time series difference data to identify acceleration and deceleration phases, including:

[0094] Based on the rate of change of the time series difference data, the instantaneous rate of change between consecutive points is calculated with a sampling time interval of 0.1s to 1.0s, and the rate of change greater than 0.05 to 0.2 is considered. The interval is defined as the acceleration segment, with the rate change below -0.05 to -0.2. The interval is determined to be a deceleration segment.

[0095] In one embodiment, the control unit processes the time-series differential data at preset sampling time intervals, the sampling time intervals being set within the range of 0.1 seconds to 1.0 seconds. The instantaneous rate of change between consecutive time points is calculated sequentially to capture the acceleration and deceleration characteristics of the cargo tag during its movement between adjacent areas.

[0096] For the calculated instantaneous rate of change, the control unit sets the acceleration threshold to 0.05 to 0.2. The deceleration threshold is -0.05 to -0.2. By comparing the instantaneous rate of change with the aforementioned threshold point by point, the interval where the rate exceeds the acceleration threshold is determined as the acceleration segment, and the interval where the rate is below the deceleration threshold is determined as the deceleration segment.

[0097] During the determination process, time points that continuously exceed the threshold can be aggregated, merging adjacent acceleration or deceleration time points into a single continuous interval to reduce noise impact and generate stable acceleration and deceleration segment identifiers. Each interval records its start and end time points, interval length, and average rate of change for subsequent rate difference calculation and dynamic offset correction.

[0098] Preferably, the method for confirming the main response antenna further includes:

[0099] When there are at least two or more radio frequency identification antennas with the highest signal strength in an adjacent area, the time-series response signals of each high-intensity antenna in the continuous detection period are acquired sequentially.

[0100] Compare the stability of the response signals of each high-intensity antenna;

[0101] The RFID antenna with the highest stability was selected as the main response antenna.

[0102] In one embodiment, when the signal strength of at least two or more RFID antennas in an adjacent area reaches its maximum value, the control unit sequentially extracts the time-series response signals of these high-intensity antennas within a continuous detection period. The detection period can be set in the range of 50 milliseconds to 500 milliseconds to ensure the real-time performance of signal acquisition.

[0103] Stability analysis was performed on the extracted time-series response signals of each high-intensity antenna, including calculating the variance of the signal amplitude, peak-to-peak fluctuation, and response repeatability, to evaluate the fluctuation of each antenna signal within a continuous detection period. Antennas with higher stability exhibit smaller amplitude fluctuations and higher response repeatability.

[0104] The control unit sorts the high-intensity antennas according to their stability indicators and selects the RFID antenna with the highest stability as the main response antenna. The main response antenna is used for subsequent time-series response difference calculation and spatial displacement vector extraction to ensure the accuracy and reliability of the response data.

[0105] Preferably, to prevent the influence of transient interference or occasional abnormal signals, a weighted average of the stability of multiple consecutive detection cycles can be performed before confirming the main response antenna, so as to further improve the robustness of the main response antenna selection.

[0106] Preferably, comparing the stability of the response signals of each high-intensity antenna includes:

[0107] Obtain the RF response signal amplitude sequence for each high-intensity antenna;

[0108] The amplitude sequence is analyzed point by point to calculate the amplitude change rate and identify the time period in which the amplitude is continuous without abrupt changes as the amplitude continuity interval;

[0109] The duration of the continuous interval of the statistical amplitude is used to form the response duration data of each high-intensity antenna;

[0110] Based on the response duration data and the average amplitude and response duration over the continuous amplitude interval, calculate the stability index value for each antenna;

[0111] The stability index values ​​are sorted to determine the most stable main response antenna in the adjacent region.

[0112] In one embodiment, the amplitude sequence of the radio frequency response signal of the radio frequency identification antenna with the highest signal strength in each adjacent region is obtained sequentially. The amplitude sequence includes amplitude data points within a continuous sampling period. The sampling period can be set in the range of 50 milliseconds to 500 milliseconds to ensure the real-time performance and continuity of the data.

[0113] The amplitude sequence of each antenna is analyzed point by point to calculate the amplitude change rate, and the time period in which the amplitude remains continuous without abrupt changes is identified as the amplitude continuity interval. The criterion for determining the amplitude continuity interval can be set as the amplitude change between adjacent data points not exceeding 0.05 to 0.2 units, in order to exclude the influence of occasional interference or noise.

[0114] The duration of each continuous amplitude interval is recorded as response duration data. For the same high-intensity antenna, the average duration of multiple continuous amplitude intervals can be calculated to obtain a more accurate assessment of response stability.

[0115] The stability index is calculated based on the response duration data and the average amplitude over a continuous amplitude interval for each antenna. The stability index comprehensively considers amplitude uniformity, duration, and amplitude fluctuation; a higher value indicates a more stable response signal from the antenna.

[0116] The stability index values ​​of all high-intensity antennas were sorted, and the RFID antenna with the highest stability index value was selected as the main response antenna. This main response antenna was used for subsequent time series difference calculation and spatial displacement vector extraction to ensure the accuracy and reliability of tag identification data.

[0117] Preferably, step S3, which involves controlling the light intensity acquisition module to activate local illumination based on the change in the spatial displacement vector to enhance the reflective signal from the label surface, includes:

[0118] The target visual area of ​​the cargo label surface under illumination is determined based on the direction and amplitude of the change in the spatial displacement vector.

[0119] The light intensity acquisition module activates local illumination in the target visualization area, and adjusts the light source brightness and illumination angle to obtain the adjusted reflected signal intensity.

[0120] When the intensity of the reflected signal is greater than or equal to the preset threshold for the intensity of the reflected signal, the corresponding optical image of the cargo label is acquired;

[0121] Contrast adjustment and exposure correction are performed on the optical image of the cargo label to obtain a high-contrast label image.

[0122] In one embodiment, the direction and magnitude of displacement are first analyzed based on the spatial displacement vector change of the cargo label to determine the target visual area of ​​the cargo label surface under illumination conditions. The target visual area can be the area of ​​the strongest reflection on the cargo label surface, and its size is proportional to the displacement amplitude. A more accurate prediction area can be obtained by accumulating multiple displacement data.

[0123] The light intensity acquisition module activates local illumination in the aforementioned target visualization area. The light intensity acquisition module includes an adjustable brightness light source and a rotatable illumination angle mechanism. Based on the material characteristics of the cargo label surface and the direction of spatial displacement, the brightness and illumination angle of the light source are automatically adjusted to optimize the reflected signal from the label surface, while avoiding overexposure or shadowed areas.

[0124] Under localized lighting conditions, the intensity of the reflected signal from the surface of the cargo label is monitored in real time. When the intensity of the reflected signal reaches or exceeds a preset threshold (e.g., an adjustable threshold in the range of 50 to 200 lux), the optical image acquisition device is triggered to capture the corresponding optical image of the cargo label. The acquired image can be grayscale or color, and the resolution can be set to 1024×1024 pixels or higher to ensure the accuracy of subsequent character recognition.

[0125] The acquired optical images of the cargo labels are further processed, including contrast adjustment, exposure correction, and noise filtering, to improve the clarity and legibility of the label characters.

[0126] Preferably, the method for obtaining the intensity of the reflected signal includes:

[0127] The directional light source unit in the light intensity acquisition module emits a local illumination beam toward the target visualization area, causing directional reflection on the label surface;

[0128] The light intensity acquisition module uses a photoelectric sensing unit to receive the light signal reflected from the label surface, and then uses a focusing lens group to converge the reflected light.

[0129] The converged reflected light is input into a photoelectric conversion element to convert the optical signal into an electrical signal, thereby obtaining a light intensity response voltage signal;

[0130] The intensity of the reflected signal is obtained by calculating the intensity value of the reflected signal based on the light intensity response voltage signal.

[0131] In one embodiment, the light intensity acquisition module emits a local illumination beam toward the target visualization area of ​​the cargo label through its directional light source unit. The beam is a directional light with adjustable brightness to ensure that the label surface produces stable directional reflection at a specific angle, avoiding signal weakening caused by light source scattering or ambient light interference.

[0132] The photoelectric sensing unit within the light intensity acquisition module receives the light signal reflected from the label surface. To ensure the concentration of the light signal and measurement accuracy, the reflected light is first converged through a focusing lens group. The focal length and optical path position of the lens group can be adjusted according to the label size and the distance to the light source, so that the reflected light is effectively converged onto the effective photosensitive area of ​​the photoelectric conversion element.

[0133] The converged optical signal is input to a photoelectric conversion element (such as a photodiode or photodiode array) to convert the optical signal into an electrical signal, resulting in a light intensity response voltage signal. The sampling frequency and sensitivity of the photoelectric conversion element can be preset according to the label surface material, reflectivity, and light source power to ensure that the dynamic range of the output signal is sufficient to cover changes in reflection intensity.

[0134] Based on the light intensity response voltage signal, combined with the preset sensitivity coefficient and sampling parameters of the photoelectric conversion element, the reflected signal intensity value of the cargo label is calculated. The obtained reflected signal intensity can be used to dynamically determine whether the label is in a high-contrast identifiable state, and drive the light intensity acquisition module to adjust the local illumination brightness or illumination angle, thereby realizing the real-time optimization of the label's reflected signal.

[0135] Preferably, the calculation of the reflected signal intensity value based on the light intensity response voltage signal includes:

[0136] The incident light angle and incident light intensity are obtained based on the optical signal;

[0137] The amplitude of the light intensity response voltage signal is normalized to obtain the normalized light intensity response voltage signal.

[0138] The average voltage value within a unit sampling period is obtained by performing time integration on the normalized light intensity response voltage signal.

[0139] The average voltage value is converted into an optical power response value using a preset sensitivity coefficient in the photoelectric conversion element;

[0140] Based on the cosine relationship between the optical power response value and the incident light angle, the direction-corrected reflection intensity data is confirmed;

[0141] The reflected signal intensity is obtained by calculating the ratio of the direction-corrected reflection intensity data to the incident light intensity.

[0142] In one embodiment, the angle and intensity of the incident light are measured based on the light signal data acquired by the light intensity acquisition module. The angle of the incident light can be calculated from the fixed position of the light source and the normal direction of the label surface, while the intensity of the incident light is obtained from the output power of the light source and the area of ​​the irradiated area.

[0143] The light intensity response voltage signal output by the photoelectric conversion element is normalized to obtain the normalized light intensity response voltage signal. The purpose of normalization is to eliminate amplitude differences caused by different sampling periods and fluctuations in light source power, ensuring the stability and comparability of subsequent calculations.

[0144] The normalized light intensity response voltage signal is integrated over time to obtain the average voltage value within a unit sampling period. This integral averaging effectively suppresses transient noise or illumination flicker in the optical signal, improving the signal-to-noise ratio of the measurement.

[0145] The optical power response value is obtained by multiplying the average voltage value per unit sampling period by the preset sensitivity coefficient of the photoelectric conversion element. The sensitivity coefficient of the photoelectric conversion element is set based on the element's manufacturing parameters and calibration data, and is used to accurately map the voltage signal to the actual optical power.

[0146] To account for the directional effect of label surface reflection, a direction correction is performed based on the cosine relationship between the optical power response value and the incident light angle, generating direction-corrected reflection intensity data. Finally, the ratio of the direction-corrected reflection intensity data to the incident light intensity is calculated to obtain the final reflected signal intensity value.

[0147] Of particular importance is the physical matching of character encoding information with the electronic tag number corresponding to the RFID signal, which includes:

[0148] Generate a character sequence based on the character position order of character encoding information;

[0149] The character sequence is mapped one-to-one with the list of electronic tag numbers acquired by the RFID antenna. Spatial correlation confirmation is performed based on the spatial position of the character sequence in the tag image and the active area position of the antenna signal. When the confirmation result is true, a cargo identification confirmation signal is generated.

[0150] In one embodiment, after obtaining the character encoding information of the goods label, the character encoding information is first processed by position serialization. Specifically, based on the arrangement direction and spacing information of the characters in the high-contrast label image, the pixel coordinate position of each character is extracted, and a character sequence is generated according to the arrangement order of the characters from left to right or from top to bottom, so as to reflect the logical arrangement order of the label content in the image space.

[0151] Subsequently, a list of electronic tag numbers output by the RFID antenna module during the identification period is obtained. This list corresponds to the tag response signals detected by each RFID antenna in different active areas. Each electronic tag number contains spatial location information and signal strength information of the identification antenna, indicating the physical location range of the tag.

[0152] The character sequence is mapped one by one to the list of electronic tag numbers, and a spatial association model is established based on the spatial coordinates of the characters in the image and the active area of ​​the RFID antenna. By comparing whether the area where each character in the character sequence is located overlaps or is adjacent to the coverage area of ​​the RFID antenna, the source of the tag signal corresponding to that character is determined. If the spatial overlap is greater than a preset threshold (e.g., above 80%), it is determined that the character and the corresponding electronic tag number are physically associated.

[0153] After spatial association confirmation, a consistency check is performed on the mapping results. If the electronic tag number corresponding to the character sequence matches the number returned by the RFID module, and the timestamps are within the same recognition period, a cargo identification confirmation signal is generated to indicate that the tag image recognition result and the electronic RFID tag recognition result of the cargo are completely consistent.

[0154] Of particular importance is the spatial correlation confirmation based on the spatial location of the character sequence in the tag image and the location of the active area of ​​the antenna signal. This also includes addressing situations where multiple tags overlap or characters are partially missing.

[0155] Extract label region location data and character boundary contour information from multi-frame high-contrast label images;

[0156] Time-series fitting is performed on the displacement trajectory of the same label area in different sampling periods to generate label motion trajectory curves;

[0157] By comparing the pixel distribution of the same character in the previous and next frames based on the label motion trajectory curve, the possible regions of the missing character can be inferred.

[0158] Local pixel interpolation and brightness compensation calculations are performed within the possible area, and the character contours generated during the completion process are compared with the character sequences of the preceding and following frames for sequence consistency. If the character recognition results are continuous and the matching degree exceeds the set standard, the sequence completion process is confirmed to be effective.

[0159] The corrected character sequence is then matched again with the corresponding electronic tag number.

[0160] In one embodiment, the spatial location of the identified character sequence in the tag image is first correlated with the location of the activated area of ​​the RFID antenna. The control unit then performs spatial matching between the coordinates of the activated area of ​​the antenna and the pixel coordinates of the tag characters in the image frame to confirm the spatial correspondence between the character sequence and the tag to which the RFID signal belongs.

[0161] When multiple overlapping labels or missing characters are detected within the same field of view, the regional location data and character boundary contour information of each label are extracted from continuously acquired high-contrast label images. This extraction process identifies the rectangular boundaries of the labels using edge detection operators and extracts the closed pixel regions of the character outer contours for subsequent trajectory fitting analysis.

[0162] The processing module performs time-series fitting on the displacement trajectory of the same tag area under different sampling periods to generate tag motion trajectory curves. The time-series fitting uses polynomial smoothing or spline interpolation methods to ensure the continuity and stability of trajectory changes, thereby accurately describing the tag's motion trend within the field of view.

[0163] Based on the label's motion trajectory curve, the pixel distribution positions of the same character in adjacent frames are compared to determine the displacement deviation of the character during motion. When a character is detected to be partially missing or occluded between consecutive frames, the processing module infers the possible range of the missing character based on the trajectory prediction results.

[0164] Within the inferred possible region, local pixel interpolation and brightness compensation operations are performed. Pixel interpolation is weighted by the gray-level gradient changes of adjacent regions, while brightness compensation is corrected by referencing the average brightness value of the same character region in previous and subsequent frames, thereby restoring the contour features of the missing character.

[0165] After the completion process is completed, the recovered character outlines are compared with the character sequences of the preceding and following frames for sequence consistency. If the matching degree of the character recognition results in consecutive frames exceeds the set standard (e.g., similarity higher than 95%), the character sequence completion process is confirmed to be effective and used as the final corrected character sequence.

[0166] Finally, the corrected character sequence is re-matched and verified against the electronic tag number corresponding to the RFID signal. If the matching result matches, a final cargo identification confirmation signal is generated, achieving robust identification even in cases of multiple tag overlap and character loss.

[0167] Therefore, the embodiments should be considered as exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0168] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A method for identifying cargo labels based on intelligent logistics warehouse management, characterized in that, Includes the following steps: Step S1: Deploy multiple sets of label recognition devices for near-field sensing in the warehouse shelf area. Each set of label recognition devices includes a radio frequency identification antenna, an infrared sensing unit, and a light intensity acquisition module. The radio frequency identification antenna is used to acquire the radio frequency response signal of the goods label, and the infrared sensing unit is used to detect the presence status of the goods on the shelf and trigger the corresponding recognition antenna to turn on. Step S2: When the infrared sensing unit detects the displacement of the goods, it sequentially activates the radio frequency identification antennas in adjacent areas, so that the tag response signal forms a time series response difference between adjacent areas; wherein, step S2 includes the following steps: Step S21: When the infrared sensing unit detects that the goods in the shelf area have moved, the displacement detection signal output by the sensing unit is collected to generate goods displacement trigger data. Step S22: Determine the adjacent identification area range of the cargo based on the cargo displacement trigger data, and select the corresponding radio frequency identification antenna group as the target identification group; Step S23: Activate each radio frequency identification antenna in the target identification group in sequence to form a continuous radio frequency transmission cycle in the time series; Step S24: Receive the radio frequency response signal returned by the cargo tag in each radio frequency transmission cycle and generate the corresponding tag response dataset; Step S25: Arrange the label response datasets between adjacent regions in chronological order and calculate the difference between adjacent data to obtain the time series response difference; wherein, step S25 includes: Extract the initial static timestamps and dynamic timestamps of the label response dataset between adjacent regions to obtain the initial time series and dynamic time series; Based on the sorting and comparison results of the initial time series and the dynamic time series, the RFID antenna with the highest signal strength in the adjacent area is selected as the main response antenna; Extract three consecutive time points from the time series of the main response antenna and calculate the average interval between each time point; The average interval is synchronously compared with the dynamic time series of adjacent antenna groups to obtain time series differential data; Time series response difference data is generated based on the rate of change of the time series difference data; Step S3: Calculate the spatial displacement vector of the cargo label based on the time series response difference, and control the light intensity acquisition module to start local illumination according to the change of the spatial displacement vector to enhance the reflection signal on the label surface, thereby generating a high-contrast label image; Step S4: Sharpen the edges and locate the characters in the high-contrast label image to extract the character encoding information of the goods label; Step S5: Physically match the character encoding information with the electronic tag number corresponding to the RFID signal. If the matching results are consistent, generate a cargo identification confirmation signal.

2. The cargo label identification method based on intelligent logistics warehouse management according to claim 1, characterized in that, Generating time series response difference data based on the rate of change of time series difference data includes: Trend analysis is performed on the rate of change of time series difference data to identify acceleration and deceleration phases. The average rate of change is extracted within the alternating intervals of acceleration and deceleration, and the rate difference between each interval is calculated. Rate difference is mapped onto time series coordinates to form a time series change curve; The dynamic offset value of the response signal is determined by the slope change of the time series curve, and the dynamic offset value is weighted and fused with the original time series difference data to generate the corrected time series response difference data.

3. The cargo label identification method based on intelligent logistics warehouse management according to claim 2, characterized in that, Trend analysis is performed on the rate of change of time series difference data to identify acceleration and deceleration phases, including: Based on the rate of change of the time series difference data, the instantaneous rate of change between consecutive points is calculated with a sampling time interval of 0.1s to 1.0s, and the rate of change greater than 0.05 to 0.2 is considered. The interval is defined as the acceleration segment, with the rate change below -0.05 to -0.

2. The interval is determined to be a deceleration segment.

4. The cargo label identification method based on intelligent logistics warehouse management according to claim 1, characterized in that, The verification methods for the main response antenna also include: When there are at least two or more radio frequency identification antennas with the highest signal strength in an adjacent area, the time-series response signals of each high-intensity antenna in the continuous detection period are acquired sequentially. Compare the stability of the response signals of each high-intensity antenna; The RFID antenna with the highest stability was selected as the main response antenna.

5. The cargo label identification method based on intelligent logistics warehouse management according to claim 4, characterized in that, The stability of the response signals of various high-intensity antennas was compared, including: Obtain the RF response signal amplitude sequence for each high-intensity antenna; The amplitude sequence is analyzed point by point to calculate the amplitude change rate and identify the time period in which the amplitude is continuous without abrupt changes as the amplitude continuity interval; The duration of the continuous interval of the statistical amplitude is used to form the response duration data of each high-intensity antenna; Based on the response duration data and the average amplitude and response duration over the continuous amplitude interval, calculate the stability index value for each antenna; The stability index values ​​are sorted to determine the most stable main response antenna in the adjacent region.

6. The cargo label identification method based on intelligent logistics warehouse management according to claim 1, characterized in that, Step S3, which involves controlling the light intensity acquisition module to activate local illumination to enhance the reflective signal from the tag surface based on the change in the spatial displacement vector, includes: The target visual area of ​​the cargo label surface under illumination is determined based on the direction and amplitude of the change in the spatial displacement vector. The light intensity acquisition module activates local illumination in the target visualization area, and adjusts the light source brightness and illumination angle to obtain the adjusted reflected signal intensity. When the intensity of the reflected signal is greater than or equal to the preset threshold for the intensity of the reflected signal, the corresponding optical image of the cargo label is acquired; Contrast adjustment and exposure correction are performed on the optical image of the cargo label to obtain a high-contrast label image.

7. The cargo label identification method based on intelligent logistics warehouse management according to claim 6, characterized in that, Methods for obtaining the intensity of reflected signals include: The directional light source unit in the light intensity acquisition module emits a local illumination beam toward the target visualization area, causing directional reflection on the label surface; The light intensity acquisition module uses a photoelectric sensing unit to receive the light signal reflected from the label surface and then uses a focusing lens group to converge the reflected light. The converged reflected light is input into a photoelectric conversion element to convert the optical signal into an electrical signal, thereby obtaining a light intensity response voltage signal; The intensity of the reflected signal is obtained by calculating the intensity value of the reflected signal based on the light intensity response voltage signal.

8. The cargo label identification method based on intelligent logistics warehouse management according to claim 7, characterized in that, The calculation of the reflected signal intensity value based on the light intensity response voltage signal includes: The incident light angle and incident light intensity are obtained based on the optical signal; The amplitude of the light intensity response voltage signal is normalized to obtain the normalized light intensity response voltage signal. The average voltage value within a unit sampling period is obtained by performing time integration on the normalized light intensity response voltage signal. The average voltage value is converted into an optical power response value using a preset sensitivity coefficient in the photoelectric conversion element; Based on the cosine relationship between the optical power response value and the incident light angle, the direction-corrected reflection intensity data is confirmed; The reflected signal intensity is obtained by calculating the ratio of the direction-corrected reflection intensity data to the incident light intensity.

Images