Intelligent sorting system based on photoelectric identification to realize full bag detection function

Abstract

This application relates to the field of express delivery full-bag detection technology, specifically to an intelligent sorting system based on photoelectric recognition to achieve full-bag detection. This system includes: a data acquisition module for real-time acquisition of distance values ​​from each sensor to its respective detection position using sensors installed at the bag opening; a data analysis module for pre-setting the full-bag detection cycle, acquiring the neighboring discrimination of each moment within each cycle for a single sensor, to filter the moments where the distance values ​​are suspected to be from the falling of packages; acquiring the classification discrimination feature values ​​of each suspected package moment within each cycle, to filter the actual package moments from the suspected package moments; and a full-bag judgment module for evaluating whether the bag is full. This application aims to improve the intelligent sorting effect of packages by enhancing the accuracy of full-bag detection.

Description

Technical Field

[0001] This application relates to the field of full-bag detection technology for express delivery, specifically to an intelligent sorting system that uses photoelectric recognition to achieve full-bag detection. Background Technology

[0002] With the development of automation technology, automated logistics sorting is becoming increasingly popular. Automated logistics sorting uses sorting conveyor lines to transfer packages and utilizes identification technologies such as barcode recognition and RFID to sequentially place packages destined for the same place into packing bags. Once the packing bag is full, it is sealed for express delivery. Therefore, accurately detecting whether the packing bag is full is crucial for improving sealing efficiency and reducing the risk of missing or incorrect packages.

[0003] However, traditional methods for detecting full-bag filling typically rely on photoelectric distance sensors to determine whether the material inside the bag interferes with light propagation. While this method works for objects with fixed shapes, it presents problems with flexible bags. When a package is placed into the bag, it causes vibration, affecting the photoelectric distance sensor's measurement results. This leads to reduced accuracy in full-bag detection using traditional methods, ultimately impacting the efficiency of intelligent package sorting. Summary of the Invention

[0004] In view of the above, it is necessary to provide an intelligent sorting system based on photoelectric recognition to realize full bag detection function. Compared with the traditional intelligent sorting system based on full bag detection function, the intelligent sorting effect of packages can be improved by improving the accuracy of full bag detection.

[0005] The intelligent sorting system based on photoelectric recognition for full-bag detection in this application adopts the following technical solution:

[0006] One embodiment of this application provides an intelligent sorting system for full-bag detection based on photoelectric recognition, wherein the system contains:

[0007] The data acquisition module uses sensors installed at the opening of the packaging bag to collect the distance values ​​from the sensors to their respective detection positions in real time.

[0008] The data analysis module is used to preset the full-bag detection cycle. For a single sensor, it analyzes the changes in distance values ​​in the nearest time periods within each cycle to obtain the neighboring region distinction within each cycle. This allows it to filter the time periods where the distance values ​​are suspected to be the landing points of packages within each cycle, and to use these as the suspected package times. By combining the neighboring region distinction and distance values ​​of the suspected package times within each cycle, it obtains the classification distinction feature values ​​of the suspected package times within each cycle, and to filter the actual package times from the suspected package times.

[0009] The full bag determination module is used to obtain full bag detection values ​​by analyzing the distribution of actual package times within each cycle. This is used to assess whether the packaging bags are full within each cycle. If the bags are not full, a package determination value is obtained based on the distance values ​​at other times within each cycle, excluding actual package times. This value is then combined with the full bag detection value to assess the full bag status in the next cycle in real time. If the bags are still not full, the full bag detection value and package determination value are updated, and full bag detection continues in subsequent cycles until all sensors detect that the bags are full.

[0010] In one embodiment, each sensor and its detection position are symmetrical about the central axis of the bag opening.

[0011] In one embodiment, obtaining the dissimilarity of the nearest neighbor includes:

[0012] The difference between the distance values ​​at each time point and its adjacent time points is recorded as the distance change value at each time point;

[0013] Calculate the dispersion of the distance change values ​​for the first half and the second half of the nearest neighbor time periods at each time point; calculate the mean of all distance change values ​​corresponding to the minimum value of the dispersion.

[0014] The degree of difference between the distance change value at each time point within each period and the mean value is measured;

[0015] The nearest neighbor distinguishability is the product of the measure of the degree of difference and the minimum value.

[0016] In one embodiment, obtaining the suspected package time includes:

[0017] Within each period, the distribution of the discriminative power of the nearest neighbor is calculated using the distance values ​​collected by all sensors, and the segmentation threshold of the discriminative power of the nearest neighbor is obtained.

[0018] Each time point in each cycle where the difference between neighboring areas is less than or equal to the segmentation threshold is considered a suspected package time point.

[0019] In one embodiment, obtaining the classification distinguishing feature value includes:

[0020] The distance values ​​of each suspected package time within each period are mapped to positive numbers, and the classification distinguishing feature value is the ratio of the neighboring distinguishability of each suspected package time within each period to the positive number.

[0021] In one embodiment, obtaining the actual package time includes:

[0022] The segmentation threshold is obtained for the classification distinguishing feature values ​​of all suspected package moments in each period. Each suspected package moment in each period whose classification distinguishing feature value is greater than the segmentation threshold is taken as the actual package moment.

[0023] In one embodiment, the process of assessing whether the packaging bag is full in each cycle is as follows:

[0024] If the last real package time in each period is the last data collection time, consecutive real package times are grouped into a package subset. The total number of real package times in each package subset is counted. The maximum value among the total number of real package times in all package subsets except the last package subset in time sequence is counted. The product of the maximum value and a preset constant greater than 1 is used as the full bag detection value. If the total number of real package times in the last package subset in time sequence is greater than the full bag detection value, it is determined that the package is full at the last time of each period. If the total number of real package times in the last package subset in time sequence is less than or equal to the full bag detection value, it is determined that the package is not full at the last time of each period.

[0025] If the last real package time in each cycle is not the last data collection time, it is determined that the package is not full at the last time in each cycle.

[0026] In one embodiment, the package determination value is the minimum distance value among all times in each period except for the actual package time.

[0027] In one embodiment, the process of real-time evaluation of the bag fullness status in the next cycle is as follows:

[0028] In the next cycle, each time when the distance value is less than the package judgment value is taken as the real package time. The number of consecutive real package times is counted in real time. If the number of real package times counted in real time is greater than the full bag detection value, the packaging bag is determined to be full.

[0029] In one embodiment, the bag is determined to be full when all sensors detect that the bag is full; otherwise, the bag is determined to be falsely full.

[0030] This application has at least the following beneficial effects:

[0031] This application analyzes changes in distance values ​​to obtain the distinguishability of nearby neighborhoods, which helps to screen suspected package times. It can effectively distinguish changes caused by the package falling from changes caused by factors other than package falling, such as the shaking of the packing bag itself, reducing misjudgments caused by factors such as packing bag shaking and improving the accuracy of full bag detection. By combining the distinguishability of nearby neighborhoods and distance values ​​to obtain classification distinguishing feature values, it can accurately extract the time when the distance value formed by the package falling can be used, which can more accurately identify the presence of the package and provide a more reliable basis for subsequent full bag detection.

[0032] Furthermore, by obtaining full-bag detection values ​​based on the actual distribution of packages at any given time, the full-bag status within each cycle is evaluated. Combined with the package judgment value, the full-bag status in the next cycle is evaluated in real time. This allows the intelligent sorting system to dynamically track the filling status of the packing bags, improving the flexibility and adaptability of full-bag detection. When no full bag is detected, the full-bag detection value and package judgment value are continuously updated, and detection continues in subsequent cycles. This continuous optimization process improves the accuracy of full-bag detection, thereby enhancing the intelligent sorting effect of packages. Attached Figure Description

[0033] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0034] Figure 1 A block diagram of an intelligent sorting system based on photoelectric recognition to achieve full-bag detection is provided in this application;

[0035] Figure 2 This is a schematic diagram of the sensor distribution;

[0036] Figure 3 This is a flowchart of an intelligent sorting system that uses photoelectric recognition to detect full bags. Detailed Implementation

[0037] In the description of the embodiments in this application, the words "exemplary," "or," and "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of the words "exemplary," "or," and "for example" is intended to present the relevant concepts in a specific manner.

[0038] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. It should be understood that, unless otherwise stated, " / " in this application means "or".

[0039] It should also be noted that the terms "first" and "second" in this application are used to distinguish similar objects, rather than to describe a specific order or sequence.

[0040] The following description, in conjunction with the accompanying drawings, details the specific scheme of the intelligent sorting system based on photoelectric recognition to achieve full-bag detection provided in this application.

[0041] Please see Figure 1 The diagram shows a block diagram of an intelligent sorting system based on photoelectric recognition to realize full bag detection according to an embodiment of this application. The system includes: intelligent recognition module 101, data acquisition module 102, data analysis module 103, and full bag judgment module 104.

[0042] The intelligent identification module 101 is used to identify the destination of the package and deliver the package to the sorting port.

[0043] During intelligent parcel sorting, images are captured to obtain the parcel's waybill. A scanner quickly scans the barcode on the waybill, accurately extracting the barcode information to determine the parcel's destination. This destination is then matched with a pre-set sorting point to pinpoint the specific sorting location. Afterward, the parcel is transported to the designated sorting point via a conveyor belt.

[0044] The data acquisition module 102 is used to collect the distance values ​​from each sensor to its respective detection position in real time using the sensors installed at the opening of the packaging bag.

[0045] The sorting port has a bag opener that opens the packaging bags. Multiple photoelectric distance sensors are installed at the bag opening to collect real-time distance measurements from each sensor to its respective detection position. The detection positions are located at the bag opening, and all photoelectric distance sensors and their respective detection positions are symmetrical about the same central axis of the bag opening. However, when using photoelectric distance sensors to measure distances, obstructions such as packaging bags or packages may exist. Therefore, the actual measured distance is the distance from the photoelectric distance sensor to the obstruction. A schematic diagram of the sensor distribution is shown below. Figure 2 As shown. To avoid the influence of the dimensions and numerical scale of the distance values ​​on subsequent analysis, the distance values ​​collected by each photoelectric ranging sensor were normalized.

[0046] In this embodiment, since the opening of the packaging bag is close to a circle, N photoelectric ranging sensors are evenly installed in half of the circumference of the packaging bag opening to ensure that all areas of the packaging bag can be detected. The value of N is 3, and the acquisition frequency of the photoelectric ranging sensor is 100Hz. The values ​​of N and acquisition frequency are preset by human. The implementer can set them according to the actual situation. This application does not impose any special restrictions.

[0047] In this embodiment, the distance value is normalized using the Min-Max normalization method. The Min-Max normalization method is a well-known technique and will not be described in detail here.

[0048] The data analysis module 103 is used to preset the full-bag detection cycle. For a single sensor, it analyzes the changes in the distance values ​​in the nearest time periods within each cycle to obtain the neighboring region distinction within each cycle. This allows it to filter the time periods where the distance values ​​suspected of being formed by the falling of a package are located, and to use these as the suspected package times. By combining the neighboring region distinction and distance values ​​of the suspected package times within each cycle, it obtains the classification distinction feature values ​​of the suspected package times within each cycle, and to filter the actual package times from the suspected package times.

[0049] (1) For a single sensor, by analyzing the changes in the distance values ​​in the nearest time periods of each period, the difference between the nearest neighbors of each period is obtained, so as to filter the time when the distance value suspected to be formed by the falling of the package is located in each period, and use it as the time of each suspected package.

[0050] The preset full-bag detection cycle is used to perform the following analysis on the distance values ​​collected by a single photoelectric ranging sensor.

[0051] In this embodiment, the cycle length is 1 minute. The cycle length is preset by the user and can be set by the implementer according to the actual situation. This application does not impose any special restrictions.

[0052] When using photoelectric distance sensors to collect distance values, the trajectory of the packages falling through the sensor's detection area is crucial because the packages need to be placed into packing bags. Therefore, the distance value collected by the sensor includes not only the distance from the packing bag to the sensor but also the distance from the package to the sensor as it falls into the detection area. Due to the flexible nature of the packing bags and their susceptibility to wind, they are prone to deformation, causing the bag opening to shift and resulting in fluctuations in the collected distance values. However, these fluctuations exhibit a degree of continuity. Furthermore, the initial distance value from the package to the sensor as it passes through the detection area differs significantly from the initial distance from the packing bag to the sensor.

[0053] Since packages are typically packed in cardboard boxes, which usually have a flat structure, the distance values ​​collected between the package and the photoelectric distance sensor usually change almost linearly. Therefore, the difference between the distance values ​​at each moment and the next adjacent moment is calculated and recorded as the distance change value at each moment. This value is used to characterize the change in the distance values ​​collected by the photoelectric distance sensor. Interpolation is then performed on the distance change value at the last moment. Because the change in the distance value from the packing bag to the photoelectric distance sensor is non-linear, while the change in the distance value from the package to the photoelectric distance sensor is linear or nearly linear, the change in the distance value within the nearest neighboring time period at each moment can measure the characteristics of the distance value change, thus helping to filter the distance values ​​formed by the package falling.

[0054] In this embodiment, the difference between distance values ​​is the absolute value of the difference.

[0055] In this embodiment, cubic spline interpolation is used to interpolate the distance variation value. The cubic spline interpolation method is a well-known technique and will not be described in detail here.

[0056] Based on the above analysis, the dispersion of distance change values ​​in the first half and second half of the nearest time period of each time is calculated respectively; the mean of all distance change values ​​corresponding to the minimum value of the dispersion is calculated, which is used to characterize the average level of distance change values ​​of the part closest to the distance change values ​​of each time.

[0057] The degree of difference between the distance change value at each time point within each period and the mean value is measured. The product of the measured degree of difference and the minimum value is used as the nearest neighbor discriminant degree at each time point within each period. Specifically, if the dispersion of the distance change value in the first half of the nearest neighbor time period is less than the dispersion of the distance change value in the second half of the nearest neighbor time period, the mean of the dispersion of the distance change value in the first half of the nearest neighbor time period is calculated; otherwise, the mean of the dispersion of the distance change value in the second half of the nearest neighbor time period is calculated.

[0058] In this embodiment, a window centered at each time point and with a length of 1×M is used as the nearest neighbor time period for each time period. The value of M is 11. While ensuring that M is a positive odd number, the implementer can set the specific value of M according to the actual situation. If there is insufficient data, since the distance from the packaging bag to the photoelectric ranging sensor may change non-linearly, cubic spline interpolation is used to interpolate the missing values. The values ​​of the first half of the nearest neighbor time period are intervals. Integers within a range, where the values ​​of the latter half of the nearest neighbor time interval are in the range [missing information]. Integers within.

[0059] It should be noted that the dispersion of data refers to the degree of unevenness in the distribution of data, which can be achieved by calculating the standard deviation, interquartile range, variance, etc. This application does not impose any special restrictions on this.

[0060] It should be noted that: difference refers to the distinction between data, which can be achieved by calculating the absolute value of the difference, the square of the difference, the ratio, etc. This application does not impose any special restrictions on this.

[0061] In this embodiment, the dispersion of the distance change values ​​is the standard deviation. The method for measuring the degree of difference between the distance change values ​​and the mean is to calculate the absolute value of the difference between the distance change values ​​and the mean.

[0062] It should be noted that in existing analyses, the difference between data and the mean of neighboring data can be used to reflect the differences between the data. However, in this application, the flexibility of the packing bag may cause non-linear changes in the data collected by the photoelectric ranging sensor, resulting in a small difference between the data and neighboring data, making it difficult to accurately reflect the difference between the two. Furthermore, the distance change from the packing bag to the photoelectric ranging sensor is non-linear, while the distance change from the package to the photoelectric ranging sensor is linear. The data dispersion caused by these two types of changes differs significantly. Therefore, by considering the difference between the data and its neighboring data, as well as the data dispersion, the degree of difference between the data and its neighboring data can be more accurately characterized. The smaller the calculated neighborhood discrepancy at each time point, the more likely that each time point represents the distance value formed by the package's fall; conversely, the larger the discrepancy, the more likely that each time point represents the distance value formed by the flexibility of the packing bag.

[0063] (2) By combining the neighboring differences and distance values ​​of each suspected package time in each period, the classification difference feature values ​​of each suspected package time in each period are obtained, so as to filter each real package time from the suspected package time.

[0064] When the calculated nearest neighbor distinguishability at each time point is small, the distance value at each time point may be due to the package falling, the packaging bag deforming slightly, or the packaging bag not deforming at all. Therefore, within each period, the distribution of nearest neighbor distinguishability calculated using the distance values ​​collected by all sensors is used to obtain the segmentation threshold for nearest neighbor distinguishability. Each time point in each period where the nearest neighbor distinguishability is less than or equal to the segmentation threshold is designated as a suspected package time point, used to characterize the time point in each period where the suspected package falling distance value is located.

[0065] In this embodiment, the Otsu threshold segmentation algorithm is used to obtain the segmentation threshold for the distinguishability of nearby neighborhoods. The Otsu threshold segmentation algorithm is a well-known technology and will not be described in detail in this application. As other implementation methods, based on the ability to obtain the segmentation threshold for the distinguishability of nearby neighborhoods, implementers may use other existing technologies, such as iterative threshold segmentation, global threshold segmentation, etc. This application does not impose any special restrictions.

[0066] Furthermore, since the distances from the package and the packing bag to the photoelectric distance sensor are different, the package is closer to the photoelectric distance sensor than the packing bag. Therefore, the distance value collected by the photoelectric distance sensor can be used to distinguish between the packing bag and the package.

[0067] Based on the above analysis, by combining the neighboring discriminant values ​​and distance values ​​of each suspected package moment within each period, the classification discriminant feature values ​​of each suspected package moment within each period are obtained. These features are used to extract the moments where the distance values ​​formed by the package's fall occur within each period. The expression is as follows:

[0068] In the formula, The classification distinguishing feature value represents the time of the j-th suspected package within the i-th period; This represents the neighborhood distinguishability of the j-th suspected package within the i-th period; the distance value of the j-th suspected package within the i-th period is mapped to a positive number. This represents a positive number obtained by mapping the distance value at the j-th suspected package time within the i-th period. The purpose of mapping the distance value to a positive number is to avoid a denominator of 0.

[0069] It should be noted that there are many ways to map data to positive numbers. Specifically, it can be achieved by calculating the sum of the data and a preset value greater than 0, or by using the data as the exponent of an exponential function with the natural constant as the base. This application does not impose any special restrictions on this.

[0070] In this embodiment, the distance value is calculated and compared with a preset value greater than 0. The sum of these values ​​maps the distance values ​​to positive numbers. The value is preset by a person, and the implementer can set it according to the actual situation. In this embodiment, The value is 1.

[0071] It should be noted that, based on existing data analysis, the distance values ​​collected by the photoelectric ranging sensor can determine whether an object exists within the detection area. In this application, because the package is in a moving state when the packaging bag is not full, and the packaging bag may also change shape, the collected distance values ​​are constantly changing. Therefore, distance values ​​alone cannot accurately characterize whether a package exists in the detection area. However, the classification distinguishing feature values ​​at the suspected package moment can characterize the characteristics of the packaging bag and the package. Therefore, by combining the classification distinguishing feature values ​​and the distance values, the characteristics of the package and the packaging bag can be more accurately characterized. The larger the calculated classification distinguishing feature values ​​at each suspected package moment, the more likely the distance value at each suspected package moment is the distance value formed by the package falling.

[0072] The full bag determination module 104 is used to obtain full bag detection values ​​by analyzing the distribution of actual package times in each cycle, in order to assess whether the packaging bag is full in each cycle. If it is not full, it obtains a package determination value based on the distance values ​​of other times in each cycle besides the actual package times, and then combines it with the full bag detection value to assess the full bag status of the packaging bag in the next cycle in real time. If it is still not full, it updates the full bag detection value and the package determination value, and continues to detect full bags in subsequent cycles until all sensors detect that the bag is full.

[0073] Since the range of values ​​for the classification distinguishing feature values ​​of the falling package and the packaging bag are different, a segmentation threshold is obtained for the classification distinguishing feature values ​​of all suspected package moments in each period. Each suspected package moment in each period whose classification distinguishing feature value is greater than the segmentation threshold is taken as the actual package moment, which is used to characterize the moment where the distance value formed by the falling package is located.

[0074] In this embodiment, the Otsu threshold segmentation algorithm is used to obtain the segmentation threshold of the classification distinguishing feature value. As other implementation methods, based on the ability to obtain the segmentation threshold of the classification distinguishing feature value, the implementer may use other existing technologies, such as iterative threshold segmentation, global threshold segmentation, etc. This application does not impose any special restrictions.

[0075] Taking the i-th cycle as an example, determine whether the last real package time in the i-th cycle is the last data collection time. If it is, group consecutive real package times into a package subset, count the total number of real package times in each subset, and calculate the maximum value among all real package times except for the last one in the time sequence. Multiply this maximum value by a preset constant greater than 1 as the full bag detection value. If the total number of real package times in the last package subset is greater than the full bag detection value, the bag is considered full at the last time of the i-th cycle. If the total number of real package times in the last package subset is less than or equal to the full bag detection value, the bag is considered not full at the last time of the i-th cycle. The minimum distance value among all times except the real package time in the i-th cycle is used as the package judgment value. If the bag is not full at the last time of the i-th cycle, use the full bag detection value and package judgment value obtained in the i-th cycle to perform real-time full bag detection in the (i+1)-th cycle.

[0076] If the last actual package time in the i-th period is not the last data collection time, it is determined that the bag is not full at the last time in the i-th period. The maximum value among the total number of actual package times in all package subsets in the i-th period is calculated, and the product of the maximum value and the preset constant greater than 1 is used as the full bag detection value. The minimum distance value among all times other than the actual package time in the i-th period is used as the package judgment value. If the bag is not full at the last time in the i-th period, the full bag detection value and package judgment value obtained in the i-th period are used to perform real-time full bag detection in the (i+1)-th period.

[0077] Within the (i+1)th cycle, each moment where the distance value is less than the package judgment value is considered a real package moment. The number of consecutive real package moments is counted in real time. If the count of real package moments exceeds the full bag detection value, the bag is considered full; otherwise, the bag is considered incomplete. If the bag is still considered incomplete at the end of the (i+1)th cycle, the full bag detection value and package judgment value are recalculated based on all distance values ​​collected within the (i+1)th cycle. This process of determining whether the bag is full continues in subsequent cycles until the bag is full.

[0078] In this embodiment, the value of the preset constant greater than 1 is 1.2, and the value of the preset constant greater than 1 is calculated from experimental data.

[0079] Furthermore, when all N photoelectric ranging sensors detect a full bag, it is determined that the bag is full and a new bag needs to be replaced. When not all N photoelectric ranging sensors detect a full bag, it is determined that the bag is falsely full, meaning the bag is not truly filled. The flowchart of the intelligent sorting system based on photoelectric recognition for full bag detection is as follows: Figure 3 As shown.

[0080] In summary, this application analyzes the changes in distance values ​​to obtain the distinguishability of nearby neighborhoods, which helps to screen suspected package times. It can effectively distinguish changes caused by the package falling from changes caused by factors other than package falling, such as the shaking of the packing bag itself, reducing misjudgments caused by factors such as packing bag shaking and improving the accuracy of full bag detection. By combining the distinguishability of nearby neighborhoods and distance values ​​to obtain classification distinguishing feature values, it can accurately extract the time when the distance value formed by the package falling can be obtained, which can more accurately identify the existence of the package and provide a more reliable basis for subsequent full bag detection.

[0081] Furthermore, by obtaining full-bag detection values ​​based on the actual distribution of packages at any given time, the full-bag status within each cycle is evaluated. Combined with the package judgment value, the full-bag status in the next cycle is evaluated in real time. This allows the intelligent sorting system to dynamically track the filling status of the packing bags, improving the flexibility and adaptability of full-bag detection. When no full bag is detected, the full-bag detection value and package judgment value are continuously updated, and detection continues in subsequent cycles. This continuous optimization process improves the accuracy of full-bag detection, thereby enhancing the intelligent sorting effect of packages.

[0082] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than that shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. In the descriptions corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in a different order than disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. Each block in a block diagram and / or flowchart, and combinations of blocks in a block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0083] It will be apparent to those skilled in the art that this application is not limited to the details of the exemplary embodiments described above, and that this application can be implemented in other specific forms without departing from its essential characteristics. Therefore, the embodiments described above should be considered exemplary and non-limiting in all respects.

Claims

1. An intelligent sorting system based on photoelectric identification to realize full bag detection function, characterized in that, The system contains: The data acquisition module uses sensors installed at the opening of the packaging bag to collect the distance values ​​from the sensors to their respective detection positions in real time. The data analysis module is used to preset the full-bag detection cycle. For a single sensor, it analyzes the changes in distance values ​​in the nearest time periods within each cycle to obtain the neighboring region distinction within each cycle. This allows it to filter the time periods where the distance values ​​are suspected to be the landing points of packages within each cycle, and to use these as the suspected package times. By combining the neighboring region distinction and distance values ​​of the suspected package times within each cycle, it obtains the classification distinction feature values ​​of the suspected package times within each cycle, and to filter the actual package times from the suspected package times. The full bag determination module is used to obtain full bag detection values ​​by analyzing the distribution of actual package times within each cycle. This is used to assess whether the packaging bags are full within each cycle. If the bags are not full, a package determination value is obtained based on the distance values ​​at other times within each cycle, excluding actual package times. This value is then combined with the full bag detection value to assess the full bag status in the next cycle in real time. If the bags are still not full, the full bag detection value and package determination value are updated, and full bag detection continues in subsequent cycles until all sensors detect that the bags are full. 2.The intelligent sorting system based on photoelectric identification to realize full bag detection function according to claim 1, wherein Each sensor and its detection position are symmetrical about the central axis of the packaging bag opening. 3.The intelligent sorting system based on photoelectric identification to realize full bag detection function according to claim 1, characterized in that, The acquisition of the dissimilarity of the nearest neighbor includes: The difference between the distance values ​​at each time point and its adjacent time points is recorded as the distance change value at each time point; Calculate the dispersion of the distance change values ​​for the first half and the second half of the nearest neighbor time periods at each time point; calculate the mean of all distance change values ​​corresponding to the minimum value of the dispersion. The degree of difference between the distance change value at each time point within each period and the mean value is measured; The nearest neighbor distinguishability is the product of the measure of the degree of difference and the minimum value. 4.The intelligent sorting system based on photoelectric identification to realize full bag detection function according to claim 1, characterized in that, The acquisition of the suspected package time includes: Within each period, the distribution of the discriminative power of the nearest neighbor is calculated using the distance values ​​collected by all sensors, and the segmentation threshold of the discriminative power of the nearest neighbor is obtained. Each time point in each cycle where the difference between neighboring areas is less than or equal to the segmentation threshold is considered a suspected package time point. 5.The intelligent sorting system based on photoelectric identification to realize full bag detection function according to claim 1, characterized in that, The acquisition of the classification distinguishing feature values ​​includes: The distance values ​​of each suspected package time within each period are mapped to positive numbers, and the classification distinguishing feature value is the ratio of the neighboring distinguishability of each suspected package time within each period to the positive number. 6.The intelligent sorting system based on photoelectric identification to realize full bag detection function according to claim 1, wherein, The acquisition of the actual package moment includes: The segmentation threshold is obtained for the classification distinguishing feature values ​​of all suspected package moments in each period. Each suspected package moment in each period whose classification distinguishing feature value is greater than the segmentation threshold is taken as the actual package moment.

7. The intelligent sorting system for full-bag detection based on photoelectric recognition as described in claim 1, characterized in that, The process for assessing whether the packaging bags are full in each period is as follows: If the last real package time in each period is the last data collection time, consecutive real package times are grouped into a package subset. The total number of real package times in each package subset is counted. The maximum value among the total number of real package times in all package subsets except the last package subset in time sequence is counted. The product of the maximum value and a preset constant greater than 1 is used as the full bag detection value. If the total number of real package times in the last package subset in time sequence is greater than the full bag detection value, it is determined that the package is full at the last time of each period. If the total number of real package times in the last package subset in time sequence is less than or equal to the full bag detection value, it is determined that the package is not full at the last time of each period. If the last real package time in each cycle is not the last data collection time, it is determined that the packaging bag is not full at the last time in each cycle.

8. The intelligent sorting system for full-bag detection based on photoelectric recognition as described in claim 1, characterized in that, The package determination value is the minimum distance value among all time points in each period except for the actual package time.

9. The intelligent sorting system for full-bag detection based on photoelectric recognition as described in claim 1, characterized in that, The process of real-time assessment of the bag fullness status in the next cycle is as follows: In the next cycle, each time when the distance value is less than the package judgment value is taken as the real package time. The number of consecutive real package times is counted in real time. If the number of real package times counted in real time is greater than the full bag detection value, the packaging bag is determined to be full.

10. The intelligent sorting system for full-bag detection based on photoelectric recognition as described in claim 1, characterized in that, If all sensors detect that the bag is full, the bag is determined to be full; otherwise, the bag is determined to be falsely full.

Images