Cashier loss prevention video monitoring method and system, terminal and storage medium
The integration of multi-dimensional data analysis with a pre-trained model in cashier systems improves detection and verification of abnormalities, enhancing loss prevention by ensuring timely and accurate identification of cashier irregularities.
Patent Information
- Application Number
- CN202411205000.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-30
- Publication Date
- 2024-12-24
AI Technical Summary
Existing cashier systems lack real-time linkage and intelligent analysis with video surveillance, leading to inefficiencies in detecting and verifying cashier abnormalities, resulting in economic losses and compromised customer experience.
A cashier loss prevention method that integrates multi-dimensional data acquisition (video, audio, motion, and temperature) with a pre-trained recognition model for real-time analysis, correlating abnormal behaviors with transaction data to generate alarms and reports.
Enhances the accuracy and timeliness of abnormal behavior detection, reducing false alarms and improving loss prevention by comprehensively monitoring cashier activities and verifying credibility through integrated data analysis.
Smart Images
Figure CN119181055A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of information security, and in particular to a cashier loss prevention video monitoring method, system, terminal and storage medium. Background Art
[0002] At present, with the rapid development of the commodity retail industry and the continuous innovation of marketing models, supermarkets, convenience stores and other retail stores are facing more and more loss prevention challenges in their daily operations. Among them, the loss problem in the cashier link is particularly prominent, mainly including abnormal behaviors such as missed settlement and wrong scanning of goods. These behaviors not only directly lead to economic losses, but also affect the customer's shopping experience and the reputation of the store. The cashier loss prevention methods of the existing cashier system mainly rely on manual monitoring and post-verification, but manual monitoring is prone to missed detection or misjudgment due to factors such as fatigue and negligence. Although the cashier system can record transaction data, it has not been effectively combined with the video surveillance system, resulting in difficulty in quickly locating and verifying problems when cashier abnormalities occur.
[0003] The above-mentioned existing technical solutions have the following defects: the existing traditional video surveillance and cash register systems are independent of each other, lack real-time linkage and intelligent analysis, and cannot effectively prevent and timely discover cash register losses, so there is room for improvement. Summary of the invention
[0004] In order to improve the accuracy of cashier loss prevention, the present application provides a cashier loss prevention video monitoring method, system, terminal and storage medium.
[0005] The above-mentioned invention objective of the present application is achieved through the following technical solutions: A cashier loss prevention video monitoring method, the cashier loss prevention video monitoring method comprising: Acquire multi-dimensional data of the cashier during the cashiering process, wherein the multi-dimensional data includes video data, audio data, cashier's motion data and temperature data; Preprocessing the multidimensional data, and transmitting the preprocessed multidimensional data to a pre-trained recognition model for comparison and analysis to obtain an analysis result of the bank receipt behavior; Correlate and compare the abnormal behavior in the analysis result of the bank collection behavior with the real-time transaction data of the cash register system to verify the credibility of the abnormal behavior, and calculate the risk level of the bank collection behavior according to the analysis result of the bank collection behavior; When the credibility of the abnormal behavior reaches a preset credibility threshold and the risk level of the bank-receiving behavior reaches a preset risk threshold, a corresponding alarm signal is generated according to the risk level of the bank-receiving behavior; When the alarm signal is generated, the corresponding video data, audio recording, cashier data and sensor data are acquired and integrated, and an analysis report is generated and stored in the cloud.
[0006] By adopting the above technical solution, by obtaining multi-dimensional data of the cashier during the cashiering process, including video data, audio data, cashier's motion data and temperature data, it is possible to comprehensively cover various information of the cashier during the work process, thereby ensuring the diversity and integrity of the data, which is conducive to subsequent comparison and analysis; by preprocessing the multi-dimensional data, it is possible to effectively remove noise and unnecessary interference information, thereby improving the quality of the data; by transmitting the preprocessed multi-dimensional data to a pre-trained recognition model for comparison and analysis, it is possible to quickly and accurately identify abnormal behaviors in the cashiering process, thereby improving the accuracy and timeliness of recognition; finally, by correlating and comparing these analysis results with the real-time transaction data of the cashier system, and combining the analysis results to calculate the risk level of the cashier behavior, it is possible to effectively verify the credibility of the abnormal behavior, and issue an alarm in time when necessary to prevent potential financial losses and improve the loss prevention effect of the system.
[0007] In a preferred example, the present application can be further configured as follows: the multi-dimensional data obtained during the cashier's cashiering process, wherein the multi-dimensional data includes video data, audio data, cashier's motion data and temperature data, includes: The multi-camera system is used to obtain video data from all angles of the cash register to ensure monitoring without blind spots; Acquiring ambient audio data around the cash register through a microphone array, and being able to separate and identify key sound signals; Using a motion capture system to capture the cashier's body movements and hand movements in real time, and summarizing the cashier's motion data; The temperature sensor monitors the temperature changes in the cash register environment in real time to identify potential abnormal operating environments.
[0008] By adopting the above technical solution, video data of the cashier counter from all angles is obtained through a multi-camera system to ensure blind spot monitoring, avoid blind spots in monitoring, and ensure that all actions of the cashier can be clearly recorded, thereby improving the comprehensiveness of monitoring; the ambient audio data around the cashier counter is obtained through a microphone array, and key sound signals can be separated and identified, and abnormal sounds in the cashier environment can be accurately captured and identified, thereby enhancing the accuracy of monitoring; the cashier's body movements and hand movements are captured in real time by using a motion capture system, and the cashier's motion data is summarized, so that every subtle movement of the cashier can be accurately captured, thereby improving the accuracy and detail of the motion data; the temperature changes of the cashier environment are monitored in real time by a temperature sensor to identify potential abnormal operating environments, which can effectively identify abnormal operating environments that may exist during the cashier process, improve the abnormal recognition ability of the monitoring system, and thus improve the overall loss prevention effect of the monitoring system.
[0009] In a preferred example, the present application may be further configured as follows: before preprocessing the multidimensional data and transmitting the preprocessed multidimensional data to a pre-trained recognition model for comparison and analysis to obtain an analysis result of the cashier behavior, the cashier loss prevention video monitoring method may further include: Acquire historical cashier data, and preprocess the historical cashier data to obtain a training set; A recognition model is constructed based on a convolutional neural network and a recurrent neural network, the recognition model is trained using the training set, and the parameters of the trained recognition model are optimized using Bayesian optimization to obtain the pre-trained recognition model.
[0010] By adopting the above technical solution, by obtaining historical cashier data and preprocessing the historical cashier data, effective information can be extracted from a large amount of historical data, noise and irrelevant data can be removed, thereby improving the availability of data; by constructing a recognition model based on a convolutional neural network and a recurrent neural network, the image recognition ability of the convolutional neural network and the time series processing ability of the recurrent neural network can be fully utilized to improve the model's recognition effect on complex behavior patterns; by using the training set to train the recognition model and using Bayesian optimization to optimize the parameters of the trained recognition model, the model can be made more robust after training and the overfitting problem can be reduced, thereby further improving the accuracy and generalization ability of the recognition model in practical applications and improving the overall loss prevention effect of the system.
[0011] In a preferred example, the present application can be further configured as follows: the multi-dimensional data is pre-processed, and the pre-processed multi-dimensional data is transmitted to a pre-trained recognition model for comparison and analysis, and the analysis results of the bank collection behavior include: Processing the audio data in the multidimensional data using a denoising algorithm to obtain a key audio signal in the audio data; Using image enhancement technology to process the video data in the multidimensional data to improve image quality and obtain processed video image data; Performing feature extraction on the action data in the multidimensional data, extracting key points and action trajectories, and obtaining a feature set of the action data; The key audio signal, the processed video image data and the feature set of the action data are transmitted to the pre-trained recognition model for comparison and analysis to obtain the analysis result of the bank collection behavior.
[0012] By adopting the above technical solution, by using a denoising algorithm to process the audio data in the multidimensional data, the background noise in the audio can be effectively removed, the key audio signal can be extracted, and the clarity and recognition effect of the audio data can be improved; by using image enhancement technology to process the video data in the multidimensional data, the clarity and contrast of the video image can be improved, so that the recognition model can more accurately analyze the details in the video data and improve the accuracy of the video analysis; by performing feature extraction on the motion data in the multidimensional data, extracting key points and motion trajectories, the motion features of the cashier can be effectively captured, and a representative feature set can be generated, thereby improving the analysis effect of the motion data; by transmitting the key audio signal, the processed video image data and the feature set of the motion data to a pre-trained recognition model for comparison and analysis, the multidimensional data can be comprehensively analyzed to generate a more accurate analysis result of the cashier's behavior, thereby improving the system's detection capability and accuracy for abnormal behavior, thereby further improving the loss prevention effect of the monitoring system.
[0013] In a preferred example, the present application can be further configured as follows: the abnormal behavior in the analysis result of the bank collection behavior is correlated and compared with the real-time transaction data of the cash register system to verify the credibility of the abnormal behavior, and the risk level of the bank collection behavior is calculated according to the analysis result of the bank collection behavior, including: Comparing and analyzing the abnormal behavior in the analysis result of the receiving bank with the standard operating procedures, and identifying the credibility score of the abnormal behavior; By using preset cashier rules, the correlation between the credibility score of the abnormal behavior and the real-time transaction data of the cashier system is analyzed to verify the credibility of the abnormal behavior, and the risk level of the cashier behavior is calculated based on the analysis results of the cashier behavior.
[0014] By adopting the above technical solution, by comparing and analyzing the abnormal behavior in the analysis results of cash register behavior with the standard operating procedures, it is possible to effectively identify abnormal behavior and generate a credibility score, reduce the false alarm rate, and improve the reliability of the identification results; by utilizing preset cash register rules to analyze the correlation between the credibility score of abnormal behavior and the real-time transaction data of the cash register system, it is possible to verify the credibility of the abnormal behavior and ensure the accuracy of the alarm signal, thereby reducing the interference caused by false alarms, improving the system's risk assessment ability for abnormal behavior, and ensuring the loss prevention effect of the monitoring system.
[0015] In a preferred example, the present application can be further configured as follows: the cashier loss prevention video monitoring method further includes: Acquire cashier recognition data in real time, and adjust the pre-trained recognition model according to the cashier recognition data within a preset time to improve the recognition accuracy of the model for abnormal behavior; After abnormal behavior occurs, the corresponding cashier is subjected to behavioral pattern analysis and performance evaluation, and the evaluation results are generated and fed back to the user end.
[0016] By adopting the above technical solution, by acquiring cashier recognition data in real time and adjusting the pre-trained recognition model according to the cashier recognition data within a preset time, the model can be updated and optimized in time, and the dynamic adaptability of the recognition model can be improved, thereby improving the model's recognition accuracy of abnormal behavior; by conducting behavioral pattern analysis and performance evaluation on the corresponding cashier after abnormal behavior occurs, generating evaluation results and feeding back to the user end, it can provide managers with detailed analysis reports on employee behavior and performance, facilitating managers to make reasonable decisions, thereby optimizing employee management and work processes, and further improving the loss prevention effect of the monitoring system.
[0017] The second object of the invention is achieved by the following technical solutions: A cashier loss prevention video monitoring system, the cashier loss prevention video monitoring system comprising: A data acquisition module is used to acquire multi-dimensional data during the cashier's cashiering process, wherein the multi-dimensional data includes video data, audio data, cashier's motion data, and temperature data; A model analysis module is used to pre-process the multi-dimensional data and transmit the pre-processed multi-dimensional data to a pre-trained recognition model for comparison and analysis to obtain an analysis result of the bank's behavior; A verification module, used to associate and compare the abnormal behavior in the analysis result of the bank collection behavior with the real-time transaction data of the cash register system, verify the credibility of the abnormal behavior, and calculate the risk level of the bank collection behavior according to the analysis result of the bank collection behavior; An alarm module, configured to generate a corresponding alarm signal according to the risk level of the bank-receiving behavior when the credibility of the abnormal behavior reaches a preset credibility threshold and the risk level of the bank-receiving behavior reaches a preset risk threshold; The storage module is used to obtain and integrate the corresponding video data, audio recordings, cashier data and sensor data when the alarm signal is generated, and generate an analysis report and store it in the cloud.
[0018] By adopting the above technical solution, by obtaining multi-dimensional data of the cashier during the cashiering process, including video data, audio data, cashier's motion data and temperature data, it is possible to comprehensively cover various information of the cashier during the work process, thereby ensuring the diversity and integrity of the data, which is conducive to subsequent comparison and analysis; by preprocessing the multi-dimensional data, it is possible to effectively remove noise and unnecessary interference information, thereby improving the quality of the data; by transmitting the preprocessed multi-dimensional data to a pre-trained recognition model for comparison and analysis, it is possible to quickly and accurately identify abnormal behaviors in the cashiering process, thereby improving the accuracy and timeliness of recognition; finally, by correlating and comparing these analysis results with the real-time transaction data of the cashier system, and combining the analysis results to calculate the risk level of the cashier behavior, it is possible to effectively verify the credibility of the abnormal behavior, and issue an alarm in time when necessary to prevent potential financial losses and improve the loss prevention effect of the system.
[0019] The third objective of the present application is achieved through the following technical solutions: A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the cashier loss prevention video monitoring method are implemented.
[0020] The fourth objective of the present application is achieved through the following technical solutions: A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the cashier loss prevention video monitoring method are implemented.
[0021] In summary, the present application includes at least one of the following beneficial technical effects: 1. By acquiring multi-dimensional data during the cashier's checkout process, including video data, audio data, cashier's motion data, and temperature data, it can comprehensively cover a variety of information during the cashier's work process, thereby ensuring the diversity and integrity of the data, which is conducive to subsequent comparison and analysis; by pre-processing the multi-dimensional data, it can effectively remove noise and unnecessary interference information, thereby improving the quality of the data; 2. By transmitting the pre-processed multi-dimensional data to the pre-trained recognition model for comparison and analysis, abnormal behaviors in the checkout process can be identified quickly and accurately, thereby improving the accuracy and timeliness of recognition; finally, by correlating and comparing these analysis results with the real-time transaction data of the cash register system, and combining the analysis results to calculate the risk level of the cash register behavior, the credibility of the abnormal behavior can be effectively verified, and an alarm can be issued in time when necessary to prevent potential financial losses and improve the system's loss prevention effect. BRIEF DESCRIPTION OF THE DRAWINGS
[0022] Figure 1 is a flow chart of a cashier loss prevention video monitoring method in an embodiment of the present application; Figure 2 This is a flowchart for implementing step S10 in a cashier loss prevention video monitoring method in one embodiment of the present application; Figure 3 This is a flowchart for implementing step S20 in a cashier loss prevention video monitoring method in an embodiment of the present application; Figure 4 This is a flowchart for implementing step S20 in a cashier loss prevention video monitoring method in one embodiment of the present application; Figure 5 This is a flowchart for implementing step S30 in a cashier loss prevention video monitoring method in one embodiment of the present application; Figure 6 This is a flow chart of an implementation of a cashier loss prevention video monitoring method in an embodiment of the present application; Figure 7 This is a principle block diagram of a cashier loss prevention video monitoring system in one embodiment of the present application; Figure 8 It is a schematic diagram of a device in an embodiment of the present application. DETAILED DESCRIPTION
[0023] The present application is further described in detail below in conjunction with the accompanying drawings.
[0024] In one embodiment, if Figure 1 As shown, the present application discloses a cashier loss prevention video monitoring method, which specifically includes the following steps: S10: Acquire multi-dimensional data during the cashier's checkout process, where the multi-dimensional data includes video data, audio data, cashier's motion data, and temperature data.
[0025] Specifically, the system collects multi-dimensional data generated by the cashier during the checkout process, including video, audio, motion and temperature data. It collects multi-dimensional data related to the cashier, captures visual information of the checkout area, records sound information of the checkout area, monitors the cashier's body movements through sensors or other technologies, and is used to monitor temperature changes in equipment or the environment.
[0026] S20: pre-process the multi-dimensional data, and transmit the pre-processed multi-dimensional data to a pre-trained recognition model for comparison and analysis to obtain an analysis result of the bank collection behavior.
[0027] Specifically, the collected multidimensional data is preprocessed through cleaning, standardization, format conversion and other steps to facilitate model analysis, and the preprocessed data is input into a pre-trained recognition model. This model may use machine learning or deep learning technology to identify normal bank checkout behaviors and abnormal behaviors. The analysis results will include an evaluation of the bank checkout behaviors, especially the abnormal behaviors, which may be the cashier's erroneous operations, suspicious behaviors or other abnormal patterns.
[0028] S30: Correlate and compare the abnormal behavior in the analysis result of the bank collection behavior with the real-time transaction data of the cash register system to verify the credibility of the abnormal behavior, and calculate the risk level of the bank collection behavior according to the analysis result of the bank collection behavior.
[0029] Specifically, the detected abnormal behaviors are compared with the real-time transaction data of the cash register system to verify whether these abnormal behaviors may affect the accuracy and security of transactions. Based on the analysis results of the abnormal behaviors, the system will calculate the risk level of the cash register behavior based on the frequency, severity or other relevant factors of the behavior.
[0030] S40: When the credibility of the abnormal behavior reaches a preset credibility threshold and the risk level of the receiving bank's behavior reaches a preset risk threshold, a corresponding alarm signal is generated according to the risk level of the receiving bank's behavior.
[0031] Specifically, the system is set with preset trust thresholds and risk thresholds. Only when the credibility and risk level of the abnormal behavior reach these thresholds at the same time, the system will consider that further action is needed. If the credibility and risk level of the abnormal behavior meet the preset conditions, the system will generate an alarm signal to alert managers or security systems to possible problems.
[0032] S50: When an alarm signal is generated, corresponding video data, audio recordings, cash register data, and sensor data are acquired and integrated, and an analysis report is generated and stored in the cloud.
[0033] Specifically, when an alarm signal is generated, the system will collect and integrate all data related to the alarm, including video, audio, cashier data, and sensor data. The system will then generate a detailed analysis report and store it in the cloud. Based on the alarm signal and analysis report, managers can take corresponding actions, such as further investigation, adjustment of security measures, or employee training.
[0034] By adopting the above technical solution, by obtaining multi-dimensional data of the cashier during the cashiering process, including video data, audio data, cashier's motion data and temperature data, it is possible to comprehensively cover various information of the cashier during the work process, thereby ensuring the diversity and integrity of the data, which is conducive to subsequent comparison and analysis; by preprocessing the multi-dimensional data, it is possible to effectively remove noise and unnecessary interference information, thereby improving the quality of the data; by transmitting the preprocessed multi-dimensional data to a pre-trained recognition model for comparison and analysis, it is possible to quickly and accurately identify abnormal behaviors in the cashiering process, thereby improving the accuracy and timeliness of recognition; finally, by correlating and comparing these analysis results with the real-time transaction data of the cashier system, and combining the analysis results to calculate the risk level of the cashier behavior, it is possible to effectively verify the credibility of the abnormal behavior, and issue an alarm in time when necessary to prevent potential financial losses and improve the loss prevention effect of the system.
[0035] In one embodiment, if Figure 2 As shown, in step S10, multi-dimensional data of the cashier during the cashiering process is obtained, and the multi-dimensional data includes video data, audio data, cashier's motion data and temperature data, specifically including: S11: The multi-camera system is used to obtain video data from all angles of the cash register to ensure monitoring without blind spots.
[0036] Specifically, all-round monitoring helps capture any possible abnormal behavior or incidents, using multiple cameras to capture real-time video of the checkout counter from different angles to ensure that there are no blind spots.
[0037] S12: The ambient audio data around the cash register is obtained through a microphone array, and key sound signals can be separated and identified.
[0038] Specifically, the ambient audio data around the checkout counter is acquired through a microphone array. Through advanced audio processing technology, the system is able to separate and identify key sound signals, such as alarms, abnormal noises or specific instructions.
[0039] S13: Use a motion capture system to capture the cashier's body movements and hand movements in real time, and summarize the cashier's motion data.
[0040] Specifically, a motion capture system is used to capture the cashier's body and hand movements in real time. The cashier's motion data is obtained by using motion capture technology. This data can be used to analyze the cashier's behavior patterns and identify any abnormal or non-standard movements. The captured cashier's motion data is summarized to form a complete data set for further analysis.
[0041] S14: Monitor the temperature change of the cash register environment in real time through a temperature sensor to identify potential abnormal operating environment.
[0042] Specifically, through the temperature sensors installed in the checkout area, the system can monitor the ambient temperature in real time. Abnormal changes in temperature may indicate potential problems, such as equipment overheating or environmental control failure, which may affect the normal progress of checkout operations. By monitoring temperature and other environmental factors, the system can identify abnormal operating environments that may affect the normal work of cashiers.
[0043] In one embodiment, if Figure 3 As shown, before step S20, that is, before preprocessing the multidimensional data and transmitting the preprocessed multidimensional data to a pre-trained recognition model for comparison and analysis to obtain the analysis result of the cashier behavior, the cashier loss prevention video monitoring method further includes: S201: Obtain historical cashier data, and pre-process the historical cashier data to obtain a training set.
[0044] Specifically, collect cashier data from the past period of time, which may include transaction records, cashier behavior records, customer interaction data, etc., and preprocess the acquired historical data to facilitate subsequent analysis and model training. Preprocessing includes removing noise, filling missing values, standardizing data, feature selection and other steps. The preprocessed data is organized into a training set, which will be used to train the machine learning model so that it can identify normal cashier behavior and abnormal behavior.
[0045] S202: construct a recognition model based on a convolutional neural network and a recurrent neural network, train the recognition model using a training set, and optimize parameters of the trained recognition model using Bayesian optimization to obtain a pre-trained recognition model.
[0046] Specifically, convolutional neural networks are usually used to process image data and can capture visual features, which are suitable for the recognition of cashier actions in video data. Recurrent neural networks can process sequence data and are suitable for behavioral pattern recognition of audio data or time series. Therefore, a recognition model is constructed based on convolutional neural networks and recurrent neural networks, and the constructed recognition model is trained using the prepared training set. During the training process, the model learns how to extract features from the input data and make predictions based on these features. Bayesian optimization is used to optimize the parameters of the trained recognition model, and the hyperparameters of the model, such as the learning rate, the number of network layers, etc., are adjusted to improve the performance of the model, and finally a pre-trained recognition model is obtained.
[0047] In one embodiment, if Figure 4As shown, in step S20, the multidimensional data is preprocessed, and the preprocessed multidimensional data is transmitted to a pre-trained recognition model for comparison and analysis to obtain the analysis results of the bank's behavior, which specifically include: S21: Process the audio data in the multidimensional data using a denoising algorithm to obtain a key audio signal in the audio data.
[0048] Specifically, the denoising algorithm is used to reduce or eliminate background noise and other interference in audio data to improve the quality of the audio signal. After processing, key audio signals, such as the sound of the cash register, the conversation between the customer and the cashier, etc., will be clearer, which will help subsequent analysis.
[0049] S22: Use image enhancement technology to process the video data in the multidimensional data to improve image quality and obtain processed video image data.
[0050] Specifically, image enhancement technology is used to improve the quality of video images, including adjusting brightness, contrast, sharpness, or using more advanced technologies such as deblurring, super-resolution, etc. Through these technologies, the cashier's actions and transaction processes in the video data will be more clearly visible, providing more accurate visual information for behavior recognition.
[0051] S23: Perform feature extraction on the action data in the multidimensional data, extract key points and action trajectories, and obtain a feature set of the action data.
[0052] Specifically, feature extraction is performed on the motion data in the multidimensional data. Feature extraction is a key step in data analysis, which is used to identify key information that helps pattern recognition from the original data. Here, the motion data will be analyzed to extract key points and motion trajectories, such as hand position, body posture, etc. and motion trajectories to form a feature set.
[0053] S24: The key audio signal, the processed video image data and the feature set of the motion data are transmitted to a pre-trained recognition model for comparison and analysis to obtain an analysis result of the bank behavior.
[0054] Specifically, the processed audio signal, enhanced video image data, and extracted motion data feature set are provided as input to a pre-trained recognition model. The recognition model will use these input data for comparative analysis to identify normal patterns and potential abnormal patterns in cashier behavior. The model uses deep learning techniques, such as convolutional neural networks and recurrent neural networks, to analyze visual and audio data, as well as the features of motion data. After model analysis, the analysis results of the cashier behavior will be obtained. These results include information such as the classification of cashier behavior, the type of abnormal behavior, the time of occurrence, etc.
[0055] In one embodiment, if Figure 5 As shown, in step S30, the abnormal behavior in the analysis result of the bank collection behavior is correlated and compared with the real-time transaction data of the cash register system to verify the credibility of the abnormal behavior, and the risk level of the bank collection behavior is calculated according to the analysis result of the bank collection behavior, which specifically includes: S31: Compare and analyze the abnormal behaviors in the analysis results of the receiving bank with the standard operating procedures, and identify the credibility score of the abnormal behaviors.
[0056] Specifically, the system compares the abnormal behavior obtained through the recognition model with the known standard operating procedures. This comparative analysis can help the system identify deviations in the cashier's behavior and thus evaluate whether these abnormal behaviors may indicate a problem. Based on the results of the comparative analysis, the system assigns a credibility score to each abnormal behavior. This score reflects the degree of deviation of the abnormal behavior from the standard operating procedure and the possibility of the abnormal behavior being confirmed as a real problem.
[0057] S32: Analyze the correlation between the credibility score of the abnormal behavior and the real-time transaction data of the cash register system based on the preset cash register rules, verify the credibility of the abnormal behavior, and calculate the risk level of the cash register behavior based on the analysis results of the cash register behavior.
[0058] Specifically, the system uses a set of predefined cashier rules to further analyze the credibility score of abnormal behavior. These rules include transaction frequency, amount, transaction pattern, etc., which are used to evaluate whether the abnormal behavior matches known fraud or error patterns. By analyzing the correlation between the credibility score of the abnormal behavior and the real-time transaction data, the system can verify whether the abnormal behavior is credible. If there is an obvious inconsistency between the abnormal behavior and the real-time transaction data or it does not comply with the preset rules, its credibility may be low; conversely, if there is consistency between the two, the credibility of the abnormal behavior is higher. Based on the analysis results of the abnormal behavior, the system calculates the risk level of each behavior based on multiple factors, including the frequency, severity, credibility score of the abnormal behavior and the correlation with the real-time transaction data.
[0059] In one embodiment, if Figure 6 As shown, the cashier loss prevention video monitoring method also includes: S61: Acquire cashier recognition data in real time, and adjust the pre-trained recognition model according to the cashier recognition data within a preset time to improve the recognition accuracy of the model for abnormal behavior.
[0060] Specifically, the system continuously collects real-time data from the cash register, including video, audio, transaction records, etc., for real-time monitoring and analysis. Based on the collected real-time data, the system adjusts the pre-trained recognition model at preset time intervals, for example, every hour or every day. This adjustment may include retraining some layers of the model, updating the model's parameters, or fine-tuning the model with new data. Through regular adjustments, the system can adapt to new or changing cash register behavior patterns and improve its ability to identify abnormal behaviors. This continuous learning and adjustment helps the model capture new abnormal features and reduce false positives and negatives.
[0061] S62: After an abnormal behavior occurs, the corresponding cashier is subjected to behavioral pattern analysis and performance evaluation, and the evaluation results are generated and fed back to the user end.
[0062] Specifically, when the system identifies abnormal behavior, it will conduct an in-depth analysis of the behavioral patterns of the cashiers involved, including the time, frequency, duration, and differences from normal behavior. The system not only analyzes abnormal behavior, but also conducts performance evaluations based on the cashier's overall performance. The performance evaluation may consider transaction processing speed, accuracy, customer satisfaction, and compliance with operating procedures. Based on the behavioral pattern analysis and performance evaluation, the system generates a detailed evaluation report, which includes the cashier's strengths and areas for improvement. The evaluation results will be fed back to the user end, which may be management or the cashier himself. After receiving this information, the user end can take corresponding actions based on the evaluation results, such as providing training, adjusting work processes, or strengthening supervision.
[0063] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.
[0064] In one embodiment, a cashier loss prevention video monitoring system is provided, and the cashier loss prevention video monitoring system corresponds one-to-one to the cashier loss prevention video monitoring method in the above embodiment. Figure 7 As shown, the cashier loss prevention video monitoring system includes a data acquisition module, a model analysis module, a verification module, an alarm module and a storage module. The detailed description of each functional module is as follows: The data acquisition module is used to acquire multi-dimensional data during the cashier's checkout process, including video data, audio data, cashier's motion data, and temperature data; The model analysis module is used to pre-process the multi-dimensional data and transmit the pre-processed multi-dimensional data to the pre-trained recognition model for comparison and analysis to obtain the analysis results of the bank's behavior; The verification module is used to correlate and compare the abnormal behavior in the analysis results of the bank collection behavior with the real-time transaction data of the cash register system, verify the credibility of the abnormal behavior, and calculate the risk level of the bank collection behavior based on the analysis results of the bank collection behavior; An alarm module is used to generate a corresponding alarm signal according to the risk level of the bank-receiving behavior when the credibility of the abnormal behavior reaches a preset credibility threshold and the risk level of the bank-receiving behavior reaches a preset risk threshold; The storage module is used to obtain and integrate the corresponding video data, audio recordings, cashier data and sensor data when an alarm signal is generated, and generate an analysis report and store it in the cloud.
[0065] Optionally, the cashier loss prevention video surveillance system further includes: Model optimization module, which is used to obtain cashier recognition data in real time and adjust the pre-trained recognition model according to the cashier recognition data within a preset time to improve the model's recognition accuracy of abnormal behavior; The feedback module is used to analyze the behavior patterns and evaluate the performance of the corresponding cashiers after abnormal behavior occurs, and generate evaluation results to feedback to the user end.
[0066] A module for obtaining historical cashier data is used to obtain historical cashier data and pre-process the historical cashier data to obtain a training set; The model training module is used to build a recognition model based on a convolutional neural network and a recurrent neural network, train the recognition model using a training set, and use Bayesian optimization to optimize the parameters of the trained recognition model to obtain a pre-trained recognition model.
[0067] Optionally, the data acquisition module includes: The video acquisition submodule is used to acquire video data from all angles of the cash register through a multi-camera system to ensure monitoring without blind spots; The audio acquisition submodule is used to acquire the ambient audio data around the cash register through a microphone array, and can separate and identify key sound signals; The motion collection submodule is used to use the motion capture system to capture the cashier's body and hand movements in real time and summarize the cashier's motion data; The temperature acquisition submodule is used to monitor the temperature changes of the cashier environment in real time through a temperature sensor to identify potential abnormal operating environments.
[0068] Optionally, the model analysis module includes: A denoising submodule, used to process the audio data in the multidimensional data using a denoising algorithm to obtain a key audio signal in the audio data; The image enhancement submodule is used to process the video data in the multidimensional data using the image enhancement technology to improve the image quality and obtain the processed video image data; The feature extraction submodule is used to extract features of the action data in the multidimensional data, extract key points and action trajectories, and obtain a feature set of the action data; The comparison and analysis submodule is used to transmit the feature set of key audio signals, processed video image data and action data to the pre-trained recognition model for comparison and analysis to obtain the analysis results of the collected bank behavior.
[0069] Optionally, the verification module includes: The credibility identification submodule is used to compare and analyze the abnormal behavior in the analysis results of the receiving bank with the standard operating procedures and identify the credibility score of the abnormal behavior; The verification submodule is used to analyze the correlation between the credibility score of abnormal behavior and the real-time transaction data of the cash register system based on preset cash register rules, verify the credibility of abnormal behavior, and calculate the risk level of the cash register behavior based on the analysis results of the cash register behavior.
[0070] For the specific definition of a cashier loss prevention video monitoring system, please refer to the definition of a cashier loss prevention video monitoring method above, which will not be repeated here. Each module in the above-mentioned cashier loss prevention video monitoring system can be implemented in whole or in part by software, hardware and their combination. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.
[0071] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 8 As shown. The computer device includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a cashier loss prevention video monitoring method is implemented.
[0072] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program: Acquire multi-dimensional data of the cashier during the cashiering process, including video data, audio data, cashier's motion data, and temperature data; Preprocess the multidimensional data, and transmit the preprocessed multidimensional data to the pre-trained recognition model for comparison and analysis to obtain the analysis results of the bank collection behavior; The abnormal behavior in the analysis results of the bank collection behavior is correlated and compared with the real-time transaction data of the cash register system to verify the credibility of the abnormal behavior, and the risk level of the bank collection behavior is calculated based on the analysis results of the bank collection behavior; When the credibility of the abnormal behavior reaches a preset credibility threshold and the risk level of the bank's behavior reaches a preset risk threshold, a corresponding alarm signal is generated according to the risk level of the bank's behavior; When an alarm signal is generated, the corresponding video data, audio recordings, cashier data and sensor data are acquired and integrated, and an analysis report is generated and stored in the cloud.
[0073] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented: Acquire multi-dimensional data of the cashier during the cashiering process, including video data, audio data, cashier's motion data, and temperature data; Preprocess the multidimensional data, and transmit the preprocessed multidimensional data to the pre-trained recognition model for comparison and analysis to obtain the analysis results of the bank collection behavior; The abnormal behavior in the analysis results of the bank collection behavior is correlated and compared with the real-time transaction data of the cash register system to verify the credibility of the abnormal behavior, and the risk level of the bank collection behavior is calculated based on the analysis results of the bank collection behavior; When the credibility of the abnormal behavior reaches a preset credibility threshold and the risk level of the bank's behavior reaches a preset risk threshold, a corresponding alarm signal is generated according to the risk level of the bank's behavior; When an alarm signal is generated, the corresponding video data, audio recordings, cashier data and sensor data are acquired and integrated, and an analysis report is generated and stored in the cloud.
[0074] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
[0075] Those skilled in the art will clearly understand that for the sake of convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above.
[0076] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.
Claims
1. A cashier loss prevention video monitoring method, characterized in that: The cashier loss prevention video monitoring method comprises: Acquire multi-dimensional data of the cashier during the cashiering process, wherein the multi-dimensional data includes video data, audio data, cashier's motion data and temperature data; Preprocessing the multidimensional data, and transmitting the preprocessed multidimensional data to a pre-trained recognition model for comparison and analysis to obtain an analysis result of the bank receipt behavior; Correlate and compare the abnormal behavior in the analysis result of the bank collection behavior with the real-time transaction data of the cash register system to verify the credibility of the abnormal behavior, and calculate the risk level of the bank collection behavior based on the analysis result of the bank collection behavior; When the credibility of the abnormal behavior reaches a preset credibility threshold and the risk level of the bank-receiving behavior reaches a preset risk threshold, a corresponding alarm signal is generated according to the risk level of the bank-receiving behavior; When the alarm signal is generated, the corresponding video data, audio recording, cashier data and sensor data are acquired and integrated, and an analysis report is generated and stored in the cloud.
2. A cashier loss prevention video monitoring method according to claim 1, characterized in that: The multi-dimensional data obtained during the cashier's cashiering process, wherein the multi-dimensional data includes video data, audio data, cashier's motion data and temperature data, includes: The multi-camera system is used to obtain video data from all angles of the cash register to ensure monitoring without blind spots; Acquiring ambient audio data around the cash register through a microphone array, and being able to separate and identify key sound signals; Using a motion capture system to capture the cashier's body movements and hand movements in real time, and summarizing the cashier's motion data; The temperature sensor monitors the temperature changes in the cash register environment in real time to identify potential abnormal operating environments.
3. A cashier loss prevention video monitoring method according to claim 1, characterized in that: Before preprocessing the multidimensional data and transmitting the preprocessed multidimensional data to a pre-trained recognition model for comparison and analysis to obtain an analysis result of the cashier behavior, the cashier loss prevention video monitoring method further includes: Acquire historical cashier data, and preprocess the historical cashier data to obtain a training set; A recognition model is constructed based on a convolutional neural network and a recurrent neural network, the recognition model is trained using the training set, and the parameters of the trained recognition model are optimized using Bayesian optimization to obtain the pre-trained recognition model.
4. A cashier loss prevention video monitoring method according to claim 3, characterized in that: The multi-dimensional data is preprocessed, and the preprocessed multi-dimensional data is transmitted to a pre-trained recognition model for comparison and analysis, and the analysis results of the bank receipt are obtained, including: Processing the audio data in the multidimensional data using a denoising algorithm to obtain a key audio signal in the audio data; Using image enhancement technology to process the video data in the multidimensional data to improve image quality and obtain processed video image data; Performing feature extraction on the action data in the multidimensional data, extracting key points and action trajectories, and obtaining a feature set of the action data; The key audio signal, the processed video image data and the feature set of the action data are transmitted to the pre-trained recognition model for comparison and analysis to obtain the analysis result of the bank collection behavior.
5. A cashier loss prevention video monitoring method according to claim 4, characterized in that: The step of correlating and comparing the abnormal behavior in the analysis result of the bank collection behavior with the real-time transaction data of the cash register system to verify the credibility of the abnormal behavior, and calculating the risk level of the bank collection behavior according to the analysis result of the bank collection behavior includes: Comparing and analyzing the abnormal behavior in the analysis result of the receiving bank with the standard operating procedures, and identifying the credibility score of the abnormal behavior; By using preset cashier rules, the correlation between the credibility score of the abnormal behavior and the real-time transaction data of the cashier system is analyzed to verify the credibility of the abnormal behavior, and the risk level of the cashier behavior is calculated based on the analysis results of the cashier behavior.
6. A cashier loss prevention video monitoring method according to claim 1, characterized in that: The cashier loss prevention video monitoring method further includes: Acquire cashier recognition data in real time, and adjust the pre-trained recognition model according to the cashier recognition data within a preset time to improve the recognition accuracy of the model for abnormal behavior; After abnormal behavior occurs, the corresponding cashier is subjected to behavioral pattern analysis and performance evaluation, and the evaluation results are generated and fed back to the user end.
7. A cashier loss prevention video monitoring system, characterized in that: The cashier loss prevention video monitoring system comprises: A data acquisition module is used to acquire multi-dimensional data during the cashier's cashiering process, wherein the multi-dimensional data includes video data, audio data, cashier's motion data, and temperature data; A model analysis module is used to pre-process the multi-dimensional data and transmit the pre-processed multi-dimensional data to a pre-trained recognition model for comparison and analysis to obtain an analysis result of the bank's behavior; A verification module, used to associate and compare the abnormal behavior in the analysis result of the bank collection behavior with the real-time transaction data of the cash register system, verify the credibility of the abnormal behavior, and calculate the risk level of the bank collection behavior according to the analysis result of the bank collection behavior; An alarm module, configured to generate a corresponding alarm signal according to the risk level of the bank-receiving behavior when the credibility of the abnormal behavior reaches a preset credibility threshold and the risk level of the bank-receiving behavior reaches a preset risk threshold; The storage module is used to obtain and integrate the corresponding video data, audio recordings, cashier data and sensor data when the alarm signal is generated, and generate an analysis report and store it in the cloud.
8. The cashier loss prevention video monitoring system according to claim 7, characterized in that: The cashier loss prevention video monitoring system also includes: A model optimization module, which is used to obtain cashier recognition data in real time, and adjust the pre-trained recognition model according to the cashier recognition data within a preset time to improve the recognition accuracy of the model for abnormal behavior; The feedback module is used to analyze the behavior patterns and evaluate the performance of the corresponding cashiers after abnormal behavior occurs, and generate evaluation results to feedback to the user end.
9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the cashier loss prevention video monitoring method as described in any one of claims 1 to 6 are implemented.
10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of a cashier loss prevention video monitoring method as described in any one of claims 1 to 6 are implemented.