Special article intelligent identification method and system

Abstract

The invention discloses a special article intelligent identification method and system, and relates to the technical field of article identification, and the method comprises the steps: obtaining the operation power of a sensor array, carrying out the analysis and calculation through combining the classification grade index and the identification influence index of a suspicious article, and obtaining the identification confidence index of the suspicious article; by constructing the classification grade index and setting the upper and lower limit thresholds, the intelligent three-level classification of the articles is realized, the risk grade is effectively distinguished, the misjudgment rate and the omission ratio caused by complete dependence on manual judgment are greatly reduced, workers are liberated from heavy manual preliminary screening, and the work efficiency is improved. Therefore, the overall customs clearance efficiency is obviously improved; furthermore, sensor operation power data are fused, a recognition confidence index is calculated in combination with classification and environment influence information, a scientific and quantitative decision basis is provided for final disposal of suspicious articles, when the confidence degree is low, manual recheck is triggered, and when the confidence degree is high, the suspicious articles are automatically classified as common articles to be released.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of article identification, in particular to a special article intelligent identification method and system. BACKGROUND

[0002] In the customs entry and exit management, the accurate detection and identification of articles are important links for maintaining national security and preventing the circulation of prohibited articles. With the increasing frequency of international trade and the acceleration of personnel flow, the customs department is facing unprecedented challenges, both ensuring efficient customs clearance and strictly controlling the safety of articles to prevent any special articles that may endanger national security and social stability from entering or leaving the country.

[0003] However, the traditional customs article detection method relies too much on manual inspection, and the detection personnel need to check a large number of articles one by one based on experience, which not only has high labor intensity and low efficiency, but also is difficult to meet the increasing demand for customs clearance brought by cross-border trade and personnel flow. On the other hand, although a single sensor data (such as X-ray imaging) is introduced as an auxiliary in some scenarios, due to the single technical means, there is a lack of multi-dimensional information fusion analysis, which leads to a high misjudgment rate. For example, some organic and inorganic substances may show similar characteristics under single imaging technology, resulting in missed detection or false detection. These problems seriously restrict the accuracy and efficiency of customs detection work.

[0004] In view of the above technical defects, the present application provides a solution. SUMMARY

[0005] The purpose of the present application is to solve the problem that the traditional customs article detection method relies too much on manual inspection, and the detection personnel need to check a large number of articles one by one based on experience, which not only has high labor intensity and low efficiency, but also is difficult to meet the increasing demand for customs clearance brought by cross-border trade and personnel flow. On the other hand, although a single sensor data is introduced as an auxiliary in some scenarios, due to the single technical means, there is a lack of multi-dimensional information fusion analysis, which leads to a high misjudgment rate.

[0006] In order to achieve the above purpose, the present application adopts the following technical scheme: a special article intelligent identification method, comprising the following steps: Step one, through the sensor array arranged on the article identification channel, the classification data, influence data and running power of the sensor array in the article identification process are collected, and the collected data is preprocessed to improve the data quality; Step two, the classification data is obtained and analyzed and calculated to obtain the classification level index of the to-be-identified article, and then compared and analyzed with the preset upper limit threshold and lower limit threshold of the classification level, and the article is divided into special article, suspicious article and ordinary article; Step 3: Obtain and analyze the impact data to obtain the identification impact index of suspicious items. The identification impact index is used to reflect the impact of environmental factors on item identification. Step 4: Obtain the operating power of the sensor array and analyze and calculate it in conjunction with the classification level index and identification impact index of the suspicious item, thereby obtaining the identification confidence index of the suspicious item.

[0007] Furthermore, the classification data includes image feature values, odor feature values, and spectral value data of the item, and the influencing data includes temperature, humidity, and light intensity data of the surrounding environment of the item identification.

[0008] Furthermore, the sensor array includes an image sensor, an odor sensor, a spectral sensor, a temperature sensor, a humidity sensor, and a light sensor.

[0009] Furthermore, the calculation process for the classification level index of the item to be identified is as follows: S11. Acquire the image feature values, odor feature values, and spectral value data of the item and perform analysis and calculation; S12. Calculate the classification level index of the item to be identified according to the following formula. : in, The total number of pre-set special items. For the first Image feature values ​​of a specific item The image feature values ​​of the object to be identified. The standard image feature difference is the preset value. For the first The odor characteristic value of a special item, The odor characteristic value of the item to be identified. The difference in odor characteristics is a preset standard. For the first The spectral values ​​of a specific item, For the spectral characteristics of the item to be identified, The standard spectral values ​​of the preset items, These are preset image weighting coefficients. The preset odor weighting coefficient, These are preset spectral weighting coefficients; S13. Obtain the preset upper limit threshold for classification level. and classification level lower limit threshold The classification level index of the item to be identified Comparative analysis, when If the similarity between the item to be identified and the preset special item is low, the item to be identified will be classified as an ordinary item. S14, when If the similarity between the item to be identified and the preset special item is generally low, the item to be identified will be classified as a suspicious item. S15, when If the item to be identified is highly similar to a preset special item, the item to be identified will be classified as a special item, an abnormal alarm will be triggered, and the information of the identified special item will be sent to the abnormal alarm terminal.

[0010] Furthermore, the calculation process for the suspicious item identification impact index is as follows: S21. Acquire and analyze data on the temperature, humidity, and light intensity of the surrounding environment of the object being identified. S22. Calculate the identification impact index of suspicious items according to the following formula. : in, To identify the temperature of the surrounding environment for the object, Identify the upper limit of the suitable temperature for the preset items. Identify the lower limit of the suitable temperature for a preset item. To identify the humidity of the surrounding environment for objects. Identify the upper limit of suitable humidity for the preset items. Identify the lower limit of suitable humidity for preset items. The preset temperature weighting coefficient, The preset humidity weighting coefficient, To identify the ambient light intensity of an object, The standard illumination intensity is the preset standard for item recognition. The identification impact index of suspicious items is used to reflect the degree of influence of environmental factors on item recognition. The higher the value of the identification impact index, the higher the degree of influence of the surrounding environment on item recognition.

[0011] Furthermore, the calculation process for the confidence index of suspicious item identification is as follows: S31. Obtain the operating power of the sensor array and analyze and calculate it in conjunction with the classification level index and identification impact index of the suspicious items; S32. Calculate the identification confidence index of suspicious items according to the following formula. : in, This is a classification level index for suspicious items. The impact index for identifying suspicious items. The weighting coefficients for the preset classification levels. The preset weighting coefficients for environmental impact. For the first The actual operating power of each sensor For the first The rated operating power of each sensor; S33. Obtain the preset recognition confidence threshold. Confidence index for identifying suspicious items Comparative analysis, when If the item is found to be suspicious, it indicates that the item is not trustworthy and staff need to open the box for inspection and identification. S34, when If the suspicious item is deemed to be of high trustworthiness, it does not require staff to open the box for inspection and should be reclassified as a regular item.

[0012] The present invention also provides a special item intelligent identification system, including a data acquisition unit, an item classification unit, an environmental analysis unit, a confidence analysis unit, and an anomaly alarm unit; The data acquisition unit is used to collect classification data, impact data and operating power of the sensor array during the object recognition process through the sensor array arranged on the object recognition channel. The classification data is sent to the object classification unit, the impact data is sent to the environmental analysis unit, and the operating power data of the sensor array is sent to the confidence analysis unit. The item classification unit is used to acquire classification data and perform analysis and calculation to obtain the classification level index of the item to be identified and to classify the item to be identified. The environmental analysis unit is used to acquire impact data and perform analysis and calculation to obtain the identification impact index of suspicious items. The identification impact index is used to reflect the impact of environmental factors on item identification. The confidence analysis unit is used to obtain the operating power of the sensor array and combine it with the classification level index and identification impact index of the suspicious item to perform analysis and calculation, thereby obtaining the identification confidence index of the suspicious item; The abnormal alarm unit is used to trigger an abnormal alarm when a special item is identified, and sends the information of the identified special item to the abnormal alarm terminal so that staff can keep track of the special item information in a timely manner.

[0013] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: This intelligent identification method and system for special items utilizes an array of sensors including image, odor, spectrum, temperature, humidity, and light sensors for multi-source data acquisition. This significantly enriches the dimensions of item characteristic information, overcoming the limitations of traditional single-sensor data and thus greatly improving the comprehensiveness and accuracy of item identification. Simultaneously, by constructing a classification level index and setting upper and lower thresholds, it achieves intelligent three-level classification of items (special, suspicious, and ordinary). This effectively distinguishes risk levels, drastically reducing the false positive and false negative rates caused by relying entirely on manual judgment, and freeing staff from arduous manual initial screening, thereby significantly improving overall customs clearance efficiency. Furthermore, by integrating sensor operating power data and combining classification and environmental impact information, an identification confidence index is calculated, providing a scientific and quantitative basis for the final disposal of suspicious items. When the confidence level is low, manual review is triggered, while when the confidence level is high, the item is automatically classified as ordinary and released. This mechanism, while ensuring a safety baseline, minimizes unnecessary unpacking inspections and optimizes human resource allocation. Attached Figure Description

[0014] Figure 1 A schematic diagram of the method flow of the present invention is shown; Figure 2 A schematic diagram of the system flow of the present invention is shown. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] Example 1: like Figure 1 As shown, a special intelligent object recognition method first uses a sensor array deployed on the object recognition channel. This sensor array includes image sensors, odor sensors, spectral sensors, temperature sensors, humidity sensors, and light sensors to collect classification data, influencing data, and the operating power of the sensor array during the object recognition process. The collected data undergoes preprocessing, specifically data cleaning to remove missing values, outliers, noisy data, and duplicate data. Then, a Min-Max normalization method is used to improve data quality. It should be noted that the classification data includes the object's image feature values, odor feature values, and spectral data, while the influencing data includes the temperature, humidity, and light intensity data of the surrounding environment.

[0017] Then, the classification data is acquired and analyzed to obtain the classification level index of the item to be identified. Then, it is compared and analyzed with the preset upper and lower thresholds of the classification level to classify the item into special items, suspicious items, and ordinary items. The calculation process for the classification level index of the item to be identified is as follows: S11. Acquire the image feature values, odor feature values, and spectral value data of the item and perform analysis and calculation; S12. Calculate the classification level index of the item to be identified according to the following formula. : in, The total number of pre-set special items. For the first Image feature values ​​of a specific item The image feature values ​​of the object to be identified. The standard image feature difference is the preset value. For the first The odor characteristic value of a special item, The odor characteristic value of the item to be identified. The difference in odor characteristics is a preset standard. For the first The spectral values ​​of a specific item, For the spectral characteristics of the item to be identified, The standard spectral values ​​of the preset items, These are preset image weighting coefficients. The preset odor weighting coefficient, These are preset spectral weighting coefficients; S13. Obtain the preset upper limit threshold for classification level. and classification level lower limit threshold The classification level index of the item to be identified Comparative analysis, when If the similarity between the item to be identified and the preset special item is low, the item to be identified will be classified as an ordinary item. S14, when If the similarity between the item to be identified and the preset special item is generally low, the item to be identified will be classified as a suspicious item. S15, when If the item to be identified is highly similar to a preset special item, the item to be identified will be classified as a special item, an abnormal alarm will be triggered, and the information of the identified special item will be sent to the abnormal alarm terminal.

[0018] Next, the impact data is acquired and analyzed to obtain the identification impact index of suspicious items. The identification impact index is used to reflect the impact of environmental factors on item identification. The calculation process for the suspicious item identification impact index is as follows: S21. Acquire and analyze data on the temperature, humidity, and light intensity of the surrounding environment of the object being identified. S22. Calculate the identification impact index of suspicious items according to the following formula. : in, To identify the temperature of the surrounding environment for the object, Identify the upper limit of the suitable temperature for the preset items. Identify the lower limit of the suitable temperature for a preset item. To identify the humidity of the surrounding environment for objects. Identify the upper limit of suitable humidity for the preset items. Identify the lower limit of suitable humidity for preset items. The preset temperature weighting coefficient, The preset humidity weighting coefficient, To identify the ambient light intensity of an object, The standard illumination intensity is the preset standard for item recognition. The identification impact index of suspicious items is used to reflect the degree of influence of environmental factors on item recognition. The higher the value of the identification impact index, the higher the degree of influence of the surrounding environment on item recognition.

[0019] Finally, the operating power of the sensor array is obtained and analyzed in conjunction with the classification level index and identification impact index of the suspicious item to obtain the identification confidence index of the suspicious item. The calculation process for the confidence index of suspicious items is as follows: S31. Obtain the operating power of the sensor array and analyze and calculate it in conjunction with the classification level index and identification impact index of the suspicious items; S32. Calculate the identification confidence index of suspicious items according to the following formula. : in, This is a classification level index for suspicious items. The impact index for identifying suspicious items. The weighting coefficients for the preset classification levels. The preset weighting coefficients for environmental impact. For the first The actual operating power of each sensor For the first The rated operating power of each sensor; S33. Obtain the preset recognition confidence threshold. Confidence index for identifying suspicious items Comparative analysis, when If the item is found to be suspicious, it indicates that the item is not trustworthy and staff need to open the box for inspection and identification. S34, when If the suspicious item is deemed to be of high trustworthiness, it does not require staff to open the box for inspection and should be reclassified as a regular item.

[0020] Example 2: like Figure 2 As shown, a special item intelligent identification system is applied to a special item intelligent identification method, including a data acquisition unit, an item classification unit, an environmental analysis unit, a confidence analysis unit, and an anomaly alarm unit; The data acquisition unit is used to collect classification data, impact data and operating power of the sensor array during the object recognition process through the sensor array arranged on the object recognition channel. The classification data is sent to the object classification unit, the impact data is sent to the environmental analysis unit, and the operating power data of the sensor array is sent to the confidence analysis unit. The item classification unit is used to acquire classification data and perform analysis and calculation to obtain the classification level index of the item to be identified and to classify the item to be identified. The environmental analysis unit is used to acquire impact data and perform analysis and calculation to obtain the identification impact index of suspicious items. The identification impact index is used to reflect the impact of environmental factors on item identification. The confidence analysis unit is used to obtain the operating power of the sensor array and combine it with the classification level index and identification impact index of the suspicious item to perform analysis and calculation, thereby obtaining the identification confidence index of the suspicious item; The abnormal alarm unit is used to trigger an abnormal alarm when a special item is identified, and sends the information of the identified special item to the abnormal alarm terminal so that staff can keep track of the special item information in a timely manner.

[0021] This invention utilizes an array of sensors including image, odor, spectrum, temperature, humidity, and light sensors for multi-source data acquisition. This significantly enriches the dimensions of item characteristic information, overcoming the limitations of traditional single-sensor data and thus greatly improving the comprehensiveness and accuracy of item identification. Simultaneously, by constructing a classification level index and setting upper and lower thresholds, it achieves intelligent three-level classification of items (special, suspicious, and ordinary). This effectively distinguishes risk levels, significantly reducing the false positive and false negative rates caused by relying entirely on manual judgment, and freeing staff from arduous manual initial screening, thereby significantly improving overall customs clearance efficiency. Furthermore, by integrating sensor operating power data and combining classification and environmental impact information to calculate an identification confidence index, it provides a scientific and quantitative basis for the final disposal of suspicious items. When the confidence level is low, manual review is triggered, while when the confidence level is high, the item is automatically classified as ordinary and released. This mechanism, while ensuring a safety baseline, minimizes unnecessary unpacking inspections and optimizes human resource allocation.

[0022] The size of the interval and threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by those skilled in the art for each set of sample data; as long as it does not affect the ratio between the parameter and the quantized value.

[0023] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation. The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for intelligent identification of special items, characterized in that, Includes the following steps: Step 1: Collect classification data, impact data, and operating power of the sensor array during the object recognition process by deploying a sensor array on the object recognition channel, and preprocess the collected data to improve data quality. Step 2: Obtain classification data and perform analysis and calculation to obtain the classification level index of the item to be identified. Then, compare and analyze the index with the preset upper and lower thresholds of the classification level to classify the item into special items, suspicious items, and ordinary items. Step 3: Obtain and analyze the impact data to obtain the identification impact index of suspicious items. The identification impact index is used to reflect the impact of environmental factors on item identification. Step 4: Obtain the operating power of the sensor array and analyze and calculate it in conjunction with the classification level index and identification impact index of the suspicious item, thereby obtaining the identification confidence index of the suspicious item.

2. The intelligent identification method for special items according to claim 1, characterized in that, The classification data includes image feature values, odor feature values, and spectral value data of the item, and the influencing data includes temperature, humidity, and light intensity data of the surrounding environment of the item identification.

3. The intelligent identification method for special items according to claim 1, characterized in that, The sensor array includes an image sensor, an odor sensor, a spectral sensor, a temperature sensor, a humidity sensor, and a light sensor.

4. The intelligent identification method for special items according to claim 1, characterized in that, The calculation process for the classification level index of the item to be identified is as follows: S11. Acquire the image feature values, odor feature values, and spectral value data of the item and perform analysis and calculation; S12. Calculate the classification level index of the item to be identified according to the following formula. : in, The total number of pre-set special items. Let be the image feature value of the th specific item. The image feature values ​​of the object to be identified. The standard image feature difference is the preset value. For the first The odor characteristic value of a special item, The odor characteristic value of the item to be identified. The difference in odor characteristics is a preset standard. For the first The spectral values ​​of a specific item, For the spectral characteristics of the item to be identified, The standard spectral values ​​of the preset items, These are preset image weighting coefficients. The preset odor weighting coefficient, These are preset spectral weighting coefficients; S13. Obtain the preset upper limit threshold for classification level. and classification level lower limit threshold The classification level index of the item to be identified Comparative analysis, when If the similarity between the item to be identified and the preset special item is low, the item to be identified will be classified as an ordinary item. S14, when If the similarity between the item to be identified and the preset special item is generally low, the item to be identified will be classified as a suspicious item. S15, when If the item to be identified is highly similar to a preset special item, the item to be identified will be classified as a special item, an abnormal alarm will be triggered, and the information of the identified special item will be sent to the abnormal alarm terminal.

5. The intelligent identification method for special items according to claim 1, characterized in that, The calculation process for the suspicious item identification impact index is as follows: S21. Acquire and analyze data on the temperature, humidity, and light intensity of the surrounding environment of the object being identified. S22. Calculate the identification impact index of suspicious items according to the following formula. : in, To identify the temperature of the surrounding environment for the object, Identify the upper limit of the suitable temperature for the preset items. Identify the lower limit of the suitable temperature for a preset item. To identify the humidity of the surrounding environment for objects. Identify the upper limit of suitable humidity for the preset items. Identify the lower limit of suitable humidity for preset items. The preset temperature weighting coefficient, The preset humidity weighting coefficient, To identify the ambient light intensity of the object, The standard illumination intensity is the preset standard for item recognition. The identification impact index of suspicious items is used to reflect the degree of influence of environmental factors on item recognition. The higher the value of the identification impact index, the higher the degree of influence of the surrounding environment on item recognition.

6. The intelligent identification method for special items according to claim 1, characterized in that, The calculation process for the confidence index of suspicious items is as follows: S31. Obtain the operating power of the sensor array and analyze and calculate it in conjunction with the classification level index and identification impact index of the suspicious items; S32. Calculate the identification confidence index of suspicious items according to the following formula. : in, The classification level index for suspicious items. The impact index on the identification of suspicious items. The weighting coefficients for the preset classification levels. The preset weighting coefficients for environmental impact. For the first The actual operating power of each sensor For the first The rated operating power of each sensor; S33. Obtain the preset recognition confidence threshold. Confidence index for identifying suspicious items Comparative analysis, when If the item is found to be suspicious, it indicates that the item is not trustworthy and staff need to open the box for inspection and identification. S34, when If the suspicious item is deemed to be of high trustworthiness, it does not require staff to open the box for inspection and should be reclassified as a regular item.

7. A special item intelligent identification system, applied to the special item intelligent identification method according to any one of claims 1-6, characterized in that, It includes a data acquisition unit, an item classification unit, an environmental analysis unit, a confidence analysis unit, and an anomaly alarm unit; The data acquisition unit is used to collect classification data, impact data and operating power of the sensor array during the object recognition process through the sensor array arranged on the object recognition channel. The classification data is sent to the object classification unit, the impact data is sent to the environmental analysis unit, and the operating power data of the sensor array is sent to the confidence analysis unit. The item classification unit is used to acquire classification data and perform analysis and calculation to obtain the classification level index of the item to be identified and to classify the item to be identified. The environmental analysis unit is used to acquire impact data and perform analysis and calculation to obtain the identification impact index of suspicious items. The identification impact index is used to reflect the impact of environmental factors on item identification. The confidence analysis unit is used to obtain the operating power of the sensor array and combine it with the classification level index and identification impact index of the suspicious item to perform analysis and calculation, thereby obtaining the identification confidence index of the suspicious item; The abnormal alarm unit is used to trigger an abnormal alarm when a special item is identified, and sends the information of the identified special item to the abnormal alarm terminal so that staff can keep track of the special item information in a timely manner.

Images