KR101823994B1 - High-risk driver's behavior analysis system - Google Patents

Abstract

The present invention relates to a high-risk driver behavior analysis system which integrally analyzes the work and accident history of high-risk drivers of vehicles such as taxis, buses, and trucks, digital tachograph (DTG) information, biometric signal information, and operation information of an electronic control unit (ECU) to provide human resource management information to prevent accidents. The high-risk driver behavior analysis system includes: a human resource data obtaining unit which obtains raw data for human resource management; a noise removing unit for filtering noises of the obtained raw data; a data conversion unit which converts the personal information of drivers and the human resource management data to the human resource management data; a risk exposure degree reflecting index calculation unit which calculates a risk exposure degree reflecting index based on the human resource analysis data; a driver analysis unit which provides a driver monitoring result and a data quality management report by analyzing drivers based on the risk exposure degree reflecting index; and a service quality management unit which generates service quality management data based on the driver monitoring result and analyzed data quality management report to provide service quality management data as human resource management information.

Description

{High-risk driver's behavior analysis system}

The present invention relates to a high-risk driver's behavior analysis system, and more particularly, to a system and method for analyzing a high-risk driver's driving behavior analysis system, The present invention relates to a system for analyzing high-risk driver's driving behavior, which is capable of providing human resource management information for accident prevention by comprehensively analyzing driving information of an ECU.

Currently, accidents involving business vehicles (taxis, buses, and lorries) are increasing, and fatalities in business vehicle traffic accidents are also increasing. As a result of investigation of traffic accident of business vehicles, the quality of service management system and management system were poor, and safety education about traffic violation and dangerous operation was all the countermeasures.

Therefore, measures for the prevention of traffic accidents of business vehicles were needed at national level, and the measures were the introduction of service quality management system based on electronic control device data using bus operation qualification test management and DTG.

The DTG data is submitted from the transportation company, and the operation record is analyzed using the eTAS, and the management information of the driver is provided to the transportation operator based on the analysis result.

On the other hand, studies are being conducted to prevent accidents by analyzing the driving behavior of the driver. Also, in the following Patent Documents 1 to 4, Lt; / RTI >

The prior art disclosed in Patent Document 1 includes a lateral G-sensor installed on a traveling vehicle to detect a lateral acceleration of the traveling vehicle, a rotational angular velocity sensor installed on the traveling vehicle for detecting a rotational angular velocity of the traveling vehicle yaw rate sensor, a velocity sensor for detecting the running speed of a traveling vehicle, a lateral acceleration sensor, a rotational angular velocity sensor, and a velocity sensor, receives and analyzes signals sensed by the sensors, And a controller.

Through such a configuration, it is possible to automatically determine the driving behavior in the turning section by solving the problem of judging the cause of accident or dangerous driving by subjective and visual inspection by the analyst, .

The conventional technique disclosed in Patent Document 2 includes a step of acquiring travel information from a vehicle control unit, a step of acquiring positional information of the vehicle from the position determination device over time, a step of acquiring the position information and the travel information Analyzing the driver driving behavior of the vehicle and outputting a warning sound or warning indication according to the analyzed driving behavior through the vehicle data output device.

Through such a configuration, a method of analyzing the driver's driving behavior is provided.

In addition, the prior art disclosed in Patent Document 3 records and analyzes accumulated driving information of a driver through a device capable of extracting a vehicle state and vehicle driving information such as an OBD, derives information related to the driving type, The present invention provides a device and method for guiding the driving fuel consumption efficiency and the driving habit for preventing the drowsy driving according to the result of the weighted function formula and provides personalized and specialized good and bad driving habits suitable for the driver's fuel efficiency To improve the fuel efficiency of the driver's vehicle and reduce the fuel cost.

In addition, the prior art disclosed in Patent Document 4 includes a boarding recognition unit for recognizing boarding of a passenger, a traveling sensing unit for sensing a traveling state of the vehicle and outputting sensing data according to the traveling state, A control unit for determining whether the vehicle starts or stops suddenly or suddenly from the sensing data, and a communication unit for mounting the sensing data on a radio signal and transmitting the sensed data to the outside.

With this configuration, a vehicle driving behavior monitoring apparatus and a monitoring method using the same provide real-time monitoring of a sudden stop, sudden start, and sudden turn of a driver driving a bus or a taxi.

Korean Registered Patent No. 10-1077094 (Registered October 20, 2011) (Vehicle Danger Driving Judgment System and Method) Korean Registered Patent No. 10-1244316 (Registered on Mar. 3, 2013) (server for calculating the insurance money for the driver of the vehicle) Korean Registered Patent No. 10-1577415 (Registered on December 8, 2015) (Driving guide apparatus and method through analysis of driving type) Korean Patent Laid-Open Publication No. 10-2016-0111278 (published on Sep. 26, 2016) (Monitoring device of vehicle driving behavior, monitoring system and monitoring method using the same)

However, the above-described management methods of general business drivers and related arts are methods of analyzing the driving behavior of the driver based on vehicle information (for example, DTG information) mounted on the vehicle, There is a disadvantage that it is not possible to prevent a vehicle accident in advance due to lack of personnel management of the driver.

For example, according to the results of analysis of cause of highway death of business vehicles by the National Police Agency's data (2015), there were 72 negligent negligence, 65 drowsiness, 41 speeding, 5 unsecured distance, 17 driver negligence, , One vehicle defect, nine unauthorized walking, and ten others.

According to the results of the police agency analysis, it can be seen that mortality caused by human resources such as drowsiness driving and forward obsession occupies a large portion in relation to the total fatal accident. This is caused by excessive work by the driver of the business, And lack of resource management.

Accordingly, the present invention has been proposed in order to solve the above-mentioned general business driver management methods and the related art problems, and it is an object of the present invention to provide a system and method for managing a business driver and a digital driving recorder (DTG) of a high risk driver such as a taxi, The present invention provides a high-risk driver behavior analysis system that can provide efficient human resource management information for accident prevention by comprehensively analyzing the information of the digital tacho graph, the biological signal information, and the operation information of the electronic control unit (ECU) It has its purpose.

Another object of the present invention is to provide an efficient human resource management system for solving the problems caused by lack of human resources management such as drowsiness operation and neglect of forward operation due to excessive work, Thereby providing a driver's behavior analysis system.

According to an aspect of the present invention, there is provided a high-risk driver's behavior analysis system comprising: a human resource data acquisition unit for acquiring raw data for human resource management; A noise removing unit for filtering the noise of the raw data acquired by the human resource data acquiring unit; A data conversion unit for converting the human resource management data acquired by the noise removing unit and the driver's personal information into human resource analysis data; A risk exposure reflection index calculation unit for calculating a risk exposure reflection index based on the human resource analysis data provided by the data conversion unit; A driver analysis unit for performing a driver analysis based on the risk exposure amount reflection index and providing a data quality management report and a driver monitoring result; And a service quality management unit for generating service quality management data based on the data quality management report and the driver monitoring result analyzed by the driver analysis unit and providing the data as the human resource management information.

The noise removing unit removes the anomaly value of the raw data by applying a noise filter to the daily accumulated travel distance data, and removes the anomaly value of the raw data by applying a noise filter to the vehicle position GPS data.

The risk-exposure-reflectance index calculator is configured to calculate the integrated risk-of-drowsiness risk, the unintentional risk of carelessness, the mishandling warning rate, the safety-distance uncertainty warning rate, and the warning level based on the human resource analysis data as the risk- .

In the above, the driver analysis unit compares the driver-side index comparison evaluation based on the calculated risk exposure reflection index to find potential high-risk drivers.

In the above, the driver analysis unit analyzes the route section of the business vehicle by applying the Point to Curve matching algorithm to the DTG GPS data and the BIS route-based information.

In the above, the driver analysis department provides the results of the driver's risk driving behavior and past risk driving behavior as a result of the driver monitoring, and provides the results of the risk driving behavior evaluation for each route section by analyzing the GPS- .

The service quality management unit reads the pattern of the management standard and the risk driving behavior based on the data quality management report and the driver monitoring result provided by the driver analysis unit and informs the transportation company manager of the risk of the driver in advance, To reduce the likelihood of an accident.

In this case, the service quality management unit provides the SPC (Statistical Process Control) alarm management function, which is a proven quality control method, for the driver monitoring and the dangerous operation analysis of the dangerous driving behavior.

According to the present invention, comprehensive analysis of the working experience and accident history of high-risk drivers such as taxis, buses, and trucks, information of digital tacho graphs (DTGs), bio-signal information and driving information of electronic control units By providing human resource management information, it is possible to reduce the incidents of high risk groups.

Also, according to the present invention, by continuously and systematically managing the human resource management of the business driver, it is possible to minimize the occurrence of accidents due to the absence of human resource management such as drowsiness driving due to excessive work, There is an advantage to be able to help.

1 is a conceptual diagram of a high-risk driver's behavior analysis system according to the present invention,
FIG. 2 is a block diagram of an embodiment of a high-risk driver's behavior analysis system according to the present invention;
3 is a diagram illustrating an example of providing a multi-upload function of raw data in the present invention,
4 is a diagram illustrating an example of quality evaluation of upload data in the present invention,
5A is an example of raw data of daily cumulative travel distance data (raw data in units of 100 mu s)
5B shows a result of filtering the noise data by applying the noise filter to the daily cumulative travel distance data in units of 1 second,
6A is an example of raw data of vehicle position GPS data (raw data in units of 100 mu s)
6B shows a result of filtering the noise data by applying the noise filter to the vehicle position GPS data in units of 1 second,
Fig. 7 is an exemplary management indicator for driver monitoring,
8 is an exemplary map matching algorithm for BIS route-based information gathering and spatial analysis in the present invention,
FIG. 9 is an exemplary view of driver monitoring and dangerous operation analysis results,
FIG. 10 is an exemplary diagram for analyzing a GIS-based risk action section,
11 is a SPC chart of the traveling speed,
FIG. 12 is a graph showing an example of a risk management operation management standard and reading result using an SPC chart,
Figure 13 is an example of an operator editing function for the SPC chart function.

Hereinafter, a high-risk driver's behavior analysis system according to a preferred embodiment of the present invention will be described in detail with reference to the accompanying drawings.

1 is a conceptual diagram of a high-risk driver's behavior analysis system according to the present invention.

The high-risk driver behavior analysis system according to the present invention includes DTG information mounted on a vehicle, bio-signal information of a business driver, personal information (work experience, accident history) of a driver, electronic control device (S1) and uploads the vehicle information as raw data to the web server for analyzing the driving behavior of the high-risk driver (S2).

The web server removes the noise of the raw data which is acquired human resource management information (S3), converts the data according to the required format (S4), and evaluates according to the risk exposure amount (S5).

Then, the driver monitoring is performed based on the evaluation information according to the risk exposure amount to determine the presence or absence of a high-risk driver (S6). Finally, the service quality management data is generated based on the data monitoring report and the driver monitoring result And provides it as human resource management information to the transportation company manager (S7).

As a result, transportation company managers can systematically and continuously manage human resources as high-risk drivers, and can identify high-risk drivers in advance, and provide education and rest for high-risk groups. It can be prevented in advance.

FIG. 2 is a block diagram showing an embodiment of a high-risk driver's behavior analysis system according to the present invention.

The high risk driver behavior analysis system according to the present invention includes a human resource data acquisition unit 10, a noise remover 20, a data conversion unit 30, a risk exposure amount reflection index calculation unit 40, a driver analysis unit 50, And a service quality management unit 60.

In FIG. 2, reference numeral 100 denotes a transportation company terminal used by a transportation company manager, and can be implemented by a PC or a smart phone. The operator requests application of a service for human resource management of a business driver to a high-risk driver behavior analysis system, which is a web server of the present invention, and receives human resource management information on the service. To this end, the personal information of the business driver and the information of the corresponding driving vehicle are registered in the web server in the high-risk driver driving behavior analysis system.

The data acquiring unit 10 acquires the driver personal information (work experience, accident history), the DTG information of the vehicle operated by the driver, the bio-signal information of the driver, and the vehicle information through the ECU of the vehicle from the transportation company It receives raw data for human resource management and acquires human resource information.

Here, the DTG information is information acquired using the DTG mounted on the vehicle, and the ECU can upload the vehicle information including the vehicle information and the lane departure information acquired through various sensors of the vehicle to the CAN (CXN) do.

In addition, the bio-signal information of the driver measures the driver's health information in the vehicle, and includes a blood pressure meter for measuring blood pressure, a pulse meter for measuring pulse, a thermometer for measuring temperature, a pressure sensor for measuring pressure, And a GSR (Galvanic Skin Reflux) sensor for detecting an electric signal generated by the driver.

The noise removing unit 20 filters the noise of the raw data acquired by the human resource data obtaining unit 10.

For example, the noise removing unit 20 may apply a noise filter to daily accumulated travel distance data to remove anomalies in raw data, and apply a noise filter to vehicle location GPS data to remove anomalies in raw data.

The data conversion unit 30 converts the human resource management data acquired by the noise removing unit 20 and personal information of the driver into human resource analysis data in a form capable of analyzing human resources, Extracts and stores them.

In addition, the risk exposure amount reflection index calculation unit 40 calculates a risk exposure amount reflection index based on the human resource analysis data provided by the data conversion unit 30.

For example, the risk-exposure-reflectance index calculator 40 may calculate the integrated risk-of-drowsiness risk, the integrated risk of carelessness, the safe driving distance warning ratio, the safety distance uncertainty warning rate, and the warning level as a risk- .

In addition, the driver analysis unit 50 performs a driver analysis based on the risk exposure amount reflection index, and provides a data quality management report and a driver monitoring result.

For example, the driver analysis unit 50 can compare the driver-to-driver index based on the calculated risk exposure reflectance index to find potential high-risk drivers. It is also possible to analyze the line segments of business vehicles by applying Point to Curve matching algorithm to DTG GPS data and BIS route based information. Also, for the high - risk analysis, the results of the driver 's risk driving behavior and past risk driving behavior are provided as a result of the driver monitoring, and the results of the risk driving behavior evaluation for each service route section can be provided through the analysis of the GPS based risk action.

In addition, the service quality management unit 60 generates service quality management data based on the data quality management report analyzed by the driver analysis unit 50 and the driver monitoring result, and provides the service quality management data as human resource management information.

For example, the service quality management unit 60 reads the patterns of the management criteria and the risk driving behavior based on the data quality management report and the driver monitoring result provided by the driver analysis unit 50, It is desirable to inform the risk and lower the possibility of an accident in the risk driving behavior.

In addition, the service quality management unit 60 preferably provides the SPC (Statistical Process Control) alarm management function, which is a proven quality management method, for the driver monitoring and the dangerous operation analysis of the dangerous driving behavior.

The operation of the high-risk driver's behavior analysis system according to the present invention will now be described in detail with reference to the accompanying drawings.

First, the transportation company manager manages human resources efficiently for high risk drivers who drive business vehicles such as buses, cabs, and trucks, and in order to reduce traffic accidents by pre - And uploads necessary human resource analysis information to the Web server.

The data acquisition unit 10 of the high-risk driver behavior analysis system, which is a web server, analyzes human resource analysis information provided by the transportation company, that is, driver personal information (work experience, accident history), DTG information of a vehicle driven by the driver, The bio-signal information of the driver and the vehicle information through the ECU of the vehicle are loaded in the raw data, and the human resource information is acquired.

As shown in FIG. 3, the data acquisition unit 10 provides a multi-upload function for user's convenience so as to facilitate the calculation of the main management index. The data acquisition unit 10 evaluates the quality (integrity verification) do. The data quality evaluation checks the data for redundancy and integrity, and informs the transportation company manager of the results. In addition, a data quality history management function may be applied to improve the quality of data.

The obtained human resource analysis information is transferred to the data converting unit 30 after the noise is removed by the noise removing unit 20 as raw data.

For example, the noise removing unit 20 removes the anomaly value of the raw data by applying a noise filter to daily cumulative travel distance data (raw data of 100 mu s) as shown in FIG. 5A. In this case, the noise filter is based on the following: when the daily travel distance is negative or the data increase width is 2 or more, the cumulative travel distance is smaller than the previous data, the increase width is 2 or more, and the GPS reference average speed is 33 m / If the change is more than 90 ° per second. FIG. 5B is a result of filtering the noise data by applying a noise filter to daily cumulative travel distance data in units of 1 second.

In addition, the noise filter is applied to the vehicle position GPS data (raw data of 100 mu s unit) as shown in Fig. 6A to remove the anomaly value of the raw data. 6B is a result of filtering the noise data by applying a noise filter to the vehicle position GPS data in units of 1 second.

The noise-filtered raw data is transmitted to the data converting unit 30. The data converting unit 30 analyzes the human resource management data acquired by the noise removing unit 20 and the personal information of the driver, To human resource analysis data of a form that can be made. And extracts only the data necessary for human resource analysis from the converted data and stores it. The stored human resource analysis data is transmitted to the risk exposure reflection index calculation unit 40.

The risk exposure amount reflection index calculation unit 40 calculates a risk exposure amount reflection index based on the human resource analysis data provided by the data conversion unit 30 as shown in FIG.

Here, the risk-exposure-reflecting-index calculating unit 40 calculates the risk of integrated drowsiness, unintentional carelessness, meandering driving warning rate, safety distance uncertainty warning rate, and warning level based on the human resource analysis data.

The index calculation method calculates the number of alarm occurrences per 1 hour and the number of alarm occurrences per 1 km as shown in the following [Equation 1] and data that does not change the accumulated travel distance at the start of the first turn, Data without a change in cumulative operating distance at the end of the turn operation, and data without a change in the cumulative operating distance for 30 minutes between the first turn and the last turn are excluded.

That is, the number of alarm occurrences per hour is calculated by converting a value obtained by dividing the sum of alarm occurrence numbers by the total number of data into a time.

Using this, the alarm occurrence number is substituted into the occurrence number of the drowsiness risk to calculate the number of drowsiness occurrences per hour, the number of peripheral drowsiness per hour is calculated by calculating the number of peripheral drowsiness per hour, , And calculates the number of unsecured safety distances per hour by substituting the number of unsuccessful safety distances.

Here, the node-to-node length is calculated by converting 1m unit into 1km unit using the route shape data of BIS bus route information. At this time, the item of the calculation index item uses the same item as the above-mentioned equation (1).

In this case, the integrated judgment drowsiness risk, unified judgment and carelessness risk, and the jogging driving warning ratio are calculated based on the occurrence rate of each level and the occurrence rate of the level 4 or more, and the warning level is calculated for each level and the occurrence rate for level 2 or more.

The calculated risk exposure reflection index is provided to the driver analysis unit 50. The driver analysis unit 50 analyzes the driver based on the risk exposure reflection index and provides a data quality management report and a driver monitoring result.

For example, the driver analysis unit 50 finds potential high-risk drivers by comparing and evaluating inter-driver indicators based on the risk exposure amount reflection index as shown in FIG.

In FIG. 7, ZigZag Warning represents a meandering driving warning indicator, Safe Distance Warning represents a safety distance uncertainty warning indicator, Fatigability is a biometric data based fatigue recognition result index, Carelessness is an inattention recognition result index, DTG1 is a rapid acceleration index, DTG3 represents a rapid deceleration index, and DTG4 represents a rapid stop index.

In addition, as shown in FIG. 8, the driver analysis unit 50 analyzes the route section of the business vehicle by applying the Point to Curve matching algorithm to the DTG GPS data and the BIS route-based information. It collects information on bus stops and service routes using the BIS Open API on a daily basis and analyzes the route sections of business vehicles by map matching using the operating speed and azimuth based on the point to curve matching.

In addition, for the analysis of high-risk groups, the results of the driver's dangerous driving behaviors and past dangerous driving behaviors as shown in FIG. Here, we provide the results of dashboard analysis for driver monitoring and provide analysis information on time, day of week, and period of dangerous driving behavior.

Also, as shown in FIG. 10, the driver analysis unit 50 also provides the results of the evaluation of the dangerous driving behavior according to the operation route section through the analysis of the GPS-based dangerous action interval.

The driver analysis information thus analyzed is transmitted to the service quality management unit 60. The service quality management unit 60 performs service quality management based on the data quality management report analyzed by the driver analysis unit 50 and the driver monitoring result, Data is generated and provided as human resource management information.

For example, the service quality management unit 60 reads the patterns of the management criteria and the risk driving behavior based on the data quality management report and the driver monitoring result provided by the driver analysis unit 50, The risk should be known to reduce the likelihood of an accident in a dangerous driving behavior.

It is preferable that the service quality management unit 60 provides a statistical process control (SPC) alarm management function, which is a proven quality control method, for driver monitoring and dangerous operation analysis of a dangerous driving behavior.

11 is an example of an SPC chart that provides for service quality management. Management upper limit line 51.5 and the lower limit line 48 to provide a management standard and read the information in the SPC 7 alarm rule for the dangerous driving behavior and transmit it to the transportation company terminal 100. Here SPC is a general quality control method that operates the process efficiently by statistical method while returning Plan-Do-Check-Act (PDCA) cycle to achieve the quality or productivity target required in the process. With the help of statistical analysis method, it is a kind of quality control method that keeps the cause of fluctuation of process quality and ability status of process, and manages the continuous improvement of quality by returning the PDCA cycle so that a given quality goal can be achieved.

FIG. 12 is an example of providing a dangerous driving behavior management standard and a reading result using the SPC chart, and FIG. 13 is an example of an editing function by the operator for the SPC chart function.

Provides SPC alarm management for dangerous driving behavior, provides operator editing function for automatically calculated management limits from data, and provides operator editing function for key management items.

The transportation company manager looks at the human resource management information provided above, finds potential high-risk drivers, and provides education or sufficient rest for safe driving, thereby reducing the possibility of traffic accidents.

The present invention has been described that the high risk driver's operation analysis system is implemented as a web server and the transportation company terminal is connected to provide the human resource management information, the high risk driver analysis is performed and the result is provided to the transportation company terminal. But it is also possible to provide the human resource management information in real time by uploading only the human resource management information with the high risk driver operating behavior analysis system directly by the transportation company.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, It is obvious to those who have.

The present invention is applied to a technology for effectively analyzing the driving behavior of a high-risk driver driving a commercial vehicle such as a taxi, a friend or a lorry, and preventing a traffic accident using the analysis result.

10: Human resource data acquisition unit
20: Noise elimination
30: Data conversion unit
40: Risk index reflection index calculation unit
50: Driver Analysis Section
60: Service Quality Management

Claims (8)

A system for analyzing the driving behavior of high risk drivers,
A human resource data acquisition unit for acquiring raw data for human resource management;
A noise removing unit for filtering the noise of the raw data acquired by the human resource data acquiring unit;
A data conversion unit for converting the human resource management data acquired by the noise removing unit and the driver's personal information into human resource analysis data;
A risk exposure reflection index calculation unit for calculating a risk exposure reflection index based on the human resource analysis data provided by the data conversion unit;
A driver analysis unit for performing a driver analysis based on the risk exposure amount reflection index and providing a data quality management report and a driver monitoring result; And
And a service quality management unit for generating service quality management data based on the data quality management report and the driver monitoring result analyzed by the driver analysis unit and providing the service quality management data as human resource management information,
The noise removing unit removes the anomaly value of the raw data by applying a noise filter to the daily accumulated travel distance data, removes the anomaly value of the raw data by applying the noise filter to the vehicle location GPS data,
If the daily cumulative travel distance is smaller than the previous data or the increase width is 2 or more, the noise filter is applied to the daily accumulated travel distance data in which the azimuth is changed by 90 ° or more in 1 second when the GPS reference average velocity is 33 m / The noise data is filtered to remove the anomaly value of the raw data,
The service quality management unit reads the pattern of the management standard and the risk driving behavior based on the data quality management report and the driver monitoring result provided by the driver analysis unit and informs the transportation company manager of the risk of the driver in advance, And high-risk driver behavior analysis, which provides driver monitoring results with SPC (Statistical Process Control) alarm management function, which is a proven quality control method for driver monitoring and risk operation analysis of dangerous driving behavior. system.
delete The dangerous-energy-re- duction index calculator calculates the risk-re- sulting index based on the human-resource analysis data by using the integrated risk-of-drowsiness risk, the combined judgment neglect risk, the safe driving distance warning ratio, A high - risk driver behavior analysis system.
The system of claim 1, wherein the driver analysis unit evaluates a driver's inter-driver index based on the calculated risk exposure metric, and identifies potential high-risk driver.
The system according to claim 1, wherein the driver analysis unit analyzes a route section of a business vehicle by applying a Point to Curve matching algorithm to DTG GPS data and BIS route-based information.
The driver analysis section provides the results of the driver's dangerous driving behavior and past dangerous driving behavior as a result of the driver monitoring for the analysis of high risk and analyzes the results of the risk driving behavior evaluation A high risk driver's behavior analysis system.



delete delete

Images