CN-121982499-A - Panoramic sensing monitoring method and system based on AI big data

CN121982499ACN 121982499 ACN121982499 ACN 121982499ACN-121982499-A

Abstract

The invention relates to the technical field of artificial intelligence application, and particularly discloses a panoramic perception monitoring method and system based on AI big data, wherein the method comprises the steps of collecting three-dimensional coordinates of passengers and suspected baggage, calculating space-time coincidence degree through Euclidean distance and time synchronism, and identifying actual affiliated baggage; dividing the luggage to be separated and associated with the luggage by the separation contribution value, and constructing a main and auxiliary judgment matrix to identify the main luggage based on the volume ratio coefficient, the historical track synergy degree and the like; obtaining a blocking coefficient for the core temporary luggage by fusing the area occupation coefficient and the passenger flow interference ratio, and calculating a retention risk coefficient by coupling the real-time retention time length, classifying risk grades according to the blocking coefficient and starting hierarchical intervention; the system comprises an accessory identification module, a separation judgment module, a main attachment judgment module, a retention evaluation module and a hierarchical intervention module.

Inventors

YE YILONG
MO YIJING
ZHANG FEIFEI
XU HUI

Assignees

杭州智科飞创信息科技有限公司

Dates

Publication Date: 20260505
Application Date: 20260126

Claims (10)

1. A panoramic sensing monitoring method based on AI big data is characterized by comprising the following steps: Acquiring the three-dimensional coordinates of the center of gravity of the passenger and the three-dimensional coordinates of the geometric center of the suspected auxiliary luggage and calculating the space-time coincidence ratio of the passenger and the suspected auxiliary luggage; Passenger luggage separation analysis is carried out on the basis of the actual auxiliary luggage to obtain a separation contribution value, and the actual auxiliary luggage is divided into to-be-taken separated luggage and carry-on associated luggage according to the separation contribution value; Calculating to obtain a volume ratio coefficient, a historical track synergy degree and a historical correlation strength according to the to-be-taken separated luggage and the carry-on correlation luggage, constructing a main auxiliary judgment matrix to identify the main luggage, and judging whether the to-be-taken separated luggage is the main luggage or not; If the to-be-taken separated luggage is the main luggage and marked as the core temporary luggage, carrying out retention analysis on the core temporary luggage to obtain retention risk coefficients of the core temporary luggage; And if the forgetting retention event is triggered, performing real-time grading intervention processing according to the retention risk coefficient of the forgetting retention event.
2. The panoramic sensing and monitoring method based on AI big data as recited in claim 1, wherein the calculating of the space-time coincidence ratio of the passenger and the suspected auxiliary luggage comprises the following steps: acquiring three-dimensional coordinates of passengers and coordinates of suspected auxiliary luggage, and calculating Euclidean distances at different acquisition time points; Obtaining a boundary bit distance coefficient, carrying out ratio processing on the Euclidean distance and the boundary bit distance coefficient, and carrying out difference value calculation on a reference unit value and a ratio processing result to obtain a space coincidence ratio; acquiring a judging time window and performing common coverage analysis to obtain time synchronization; and (5) carrying out product calculation on the time synchronization degree and the space coincidence degree to obtain the space-time coincidence degree.
3. The panoramic sensing and monitoring method based on AI big data as set forth in claim 2, wherein the process of performing the common coverage analysis is: and calculating the ratio of the total frame number of the common coverage frame number and the total frame number of the judging time window to obtain the time synchronization.
4. The panoramic sensing and monitoring method based on AI big data as set forth in claim 1, wherein the manner of obtaining the separation contribution value is: Acquiring a separation judgment time window, and calculating a real-time Euclidean distance between a three-dimensional coordinate of a passenger at an acquisition time point and an actual auxiliary luggage coordinate; comparing the real-time Euclidean distance with the boundary bit distance coefficient to identify a separation risk time point in the acquisition time point; and analyzing according to the separation risk time point to obtain a separation time value and a separation degree value, and performing product calculation on the separation time value and the separation degree value to obtain a separation contribution value.
5. The AI big data based panoramic sensing and monitoring method as set forth in claim 4, wherein the separation time value is obtained by: Counting the number of the separation risk time points in the separation judgment time window, and carrying out ratio calculation on the number of the separation risk time points and the total acquisition time points in the separation judgment time window to obtain a separation time value.
6. The AI big data based panoramic sensing and monitoring method as set forth in claim 4, wherein the separation degree value is obtained by: Based on the separation risk time point, acquiring a real-time Euclidean distance between a passenger three-dimensional coordinate at the separation risk time point and an actual auxiliary luggage coordinate; and obtaining absolute distance deviation of all the separation risk time points and carrying out averaging calculation to obtain a separation degree value.
7. The panoramic sensing and monitoring method based on AI big data as set forth in claim 1, wherein the method for constructing the main and auxiliary decision matrix to identify the main baggage is as follows: obtaining the volume ratio coefficient of the actual auxiliary luggage, the historical track synergy degree and the historical association strength, and performing product calculation to obtain the main auxiliary comprehensive coefficient of each actual auxiliary luggage; And arranging the main auxiliary comprehensive coefficients of all the actual auxiliary baggage according to the reverse order, and marking the actual auxiliary baggage corresponding to the maximum main auxiliary comprehensive coefficient as the main baggage.
8. The panoramic sensing and monitoring method based on AI big data as set forth in claim 1, wherein the process of obtaining the retention risk coefficient of the core temporary luggage is as follows: Acquiring an effective passing area of an effective passing area, and marking a projection area of the core temporary luggage in the effective passing area as a luggage occupation area; calculating the ratio of the occupied area of the luggage to the effective passing area to obtain an area occupation coefficient; Obtaining a passenger flow interference ratio, and carrying out product calculation on the area occupation coefficient and the passenger flow interference ratio to obtain a blocking coefficient; and obtaining the real-time retention time length, and performing product calculation on the blocking coefficient and the real-time retention time length to obtain the retention risk coefficient.
9. The AI big data based panoramic sensing monitoring method of claim 8, wherein the manner of obtaining the real-time retention time length is: And obtaining a current acquisition time point in the retention analysis time window, and calculating the difference value between the current acquisition time point and an initial acquisition time point of the retention analysis time window to obtain the real-time retention time.
10. The AI-big-data-based panoramic sensing monitoring system of claim 1, configured to implement the AI-big-data-based panoramic sensing monitoring method of any one of claims 1 to 9, comprising: the auxiliary identification module is used for acquiring the three-dimensional coordinates of the center of gravity of the passenger and the three-dimensional coordinates of the geometric center of the suspected auxiliary luggage and calculating the space-time coincidence ratio of the passenger and the suspected auxiliary luggage; The separation judgment module is used for carrying out passenger luggage separation analysis based on the actual auxiliary luggage to obtain a separation contribution value, and dividing the actual auxiliary luggage into to-be-taken separation luggage and carry-on associated luggage according to the separation contribution value; The main and auxiliary judging module is used for calculating the volume ratio coefficient, the historical track synergy degree and the historical association strength according to the to-be-taken separated luggage and the carry-on associated luggage, constructing a main and auxiliary judging matrix to identify the main luggage, and judging whether the to-be-taken separated luggage is the main luggage or not; The retention evaluation module is used for marking the to-be-taken separated luggage as the main luggage as the core temporary luggage, carrying out retention analysis on the core temporary luggage to obtain retention risk coefficients of the core temporary luggage, and judging whether the core temporary luggage triggers a forgetting retention event or not based on the retention risk coefficients; and the grading intervention module is used for carrying out real-time grading intervention processing according to the retention risk coefficient of the forgetting retention event if the forgetting retention event is triggered.

Description

Panoramic sensing monitoring method and system based on AI big data Technical Field The invention relates to the technical field of artificial intelligence application, in particular to a panoramic sensing monitoring method and system based on AI big data. Background The panoramic perception monitoring is to arrange a multidimensional data acquisition array to form a sensing network with global coverage instead of relying on local visual angle acquisition of a single camera for a high-dynamic, high-shielding and multi-target people flow dense scene such as a high-speed rail station escalator entrance, along with the rapid development of the high-speed rail network, the high-speed rail station becomes one of traffic hubs with most frequent personnel flow, and the escalator entrance is used as a core channel for passengers to enter and exit a waiting hall and a platform, so that the traffic efficiency and safety control directly relate to the operation order of the whole station. In order to ensure the passing safety of the escalator entrance area and reduce the problems of passenger flow blocking, potential safety hazards and the like caused by luggage retention, the panoramic sensing monitoring system is widely applied to abnormal behavior identification of the area. The intelligent monitoring scheme of the elevator landing entrance of the current high-speed rail station relies on static image recognition or simple two-dimensional motion trail analysis technology to carry out target association judgment, so that the boundary between temporary placement and forgetting retention of baggage cannot be distinguished, the attribution relation between the baggage and passengers in a high shielding scene cannot be effectively identified, the technical short board with the missing association judgment of main baggage and auxiliary articles in a multi-line baggage scene exists, misjudgment and missed judgment easily occurs in a scene of the elevator landing entrance where people are dense and frequent shielding, the passing order and safety management and control efficiency are further affected, and the panoramic perception technology based on AI big data is used for integrating massive data such as three-dimensional space-time and mechanical parameters acquired by multi-source equipment and combining a quantitative calculation model with definite physical meaning. Therefore, the invention provides a panoramic sensing monitoring method and a panoramic sensing monitoring system based on AI big data. Disclosure of Invention The invention aims to provide a panoramic sensing monitoring method and a panoramic sensing monitoring system based on AI big data, so as to solve the background problem. The invention aims at realizing the panoramic sensing monitoring method based on the AI big data, which comprises the following steps: Acquiring the three-dimensional coordinates of the center of gravity of the passenger and the three-dimensional coordinates of the geometric center of the suspected auxiliary luggage and calculating the space-time coincidence ratio of the passenger and the suspected auxiliary luggage; Passenger luggage separation analysis is carried out on the basis of the actual auxiliary luggage to obtain a separation contribution value, and the actual auxiliary luggage is divided into to-be-taken separated luggage and carry-on associated luggage according to the separation contribution value; Calculating to obtain a volume ratio coefficient, a historical track synergy degree and a historical correlation strength according to the to-be-taken separated luggage and the carry-on correlation luggage, constructing a main auxiliary judgment matrix to identify the main luggage, and judging whether the to-be-taken separated luggage is the main luggage or not; If the to-be-taken separated luggage is the main luggage and marked as the core temporary luggage, carrying out retention analysis on the core temporary luggage to obtain retention risk coefficients of the core temporary luggage; And if the forgetting retention event is triggered, performing real-time grading intervention processing according to the retention risk coefficient of the forgetting retention event. Further, the process of calculating the space-time coincidence ratio of the passenger and the suspected auxiliary luggage is as follows: acquiring three-dimensional coordinates of passengers and coordinates of suspected auxiliary luggage, and calculating Euclidean distances at different acquisition time points; Obtaining a boundary bit distance coefficient, carrying out ratio processing on the Euclidean distance and the boundary bit distance coefficient, and carrying out difference value calculation on a reference unit value and a ratio processing result to obtain a space coincidence ratio; acquiring a judging time window and performing common coverage analysis to obtain time synchronization; and (5) carrying out product calculation on the time synchronization