Monitoring system for fire fighting

Abstract

The invention discloses a monitoring system for fire fighting, which comprises a ground sensing network, an unmanned aerial vehicle inspection module, a data integration module, a fire early warning module and a monitoring enhancement module, and relates to the technical field of fire fighting. Ground fixed point location monitoring and unmanned aerial vehicle inspection are combined, global coverage of a monitoring range is realized, two types of data are fused and calibrated by means of dynamic weight adjustment, data accuracy and relevance are improved, a reliable basis is provided for fire risk assessment, and the fire risk assessment accuracy is improved by quantitatively calculating the fire occurrence probability and possible occurrence time. Support is provided for graded early warning, an early warning mechanism is finer, false alarm and missing alarm are reduced, a monitoring enhancement module is used for automatically triggering high-frequency sampling adjustment after warning, a continuous tracking closed loop for an abnormal area is formed, efficient linkage and transmission of key information are achieved in combination with a dynamic two-dimensional code label, the emergency response process is accelerated, and the early warning efficiency is improved. And a more reliable and intelligent guarantee is provided for fire safety in different scenes.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of fire fighting technology, in particular to a fire monitoring system. BACKGROUND

[0002] Fire monitoring generally refers to a system that uses sensors, devices and technologies such as smoke detectors, temperature sensors, flame detectors, etc. to monitor fire hazards in a building or area in real time and alarm in time when an anomaly is found.

[0003] The traditional fire monitoring system has a single monitoring method, relies on fixed-point sensors for data collection, has a limited coverage range, and is prone to form monitoring dead angles due to uneven distribution of equipment. In particular, in large factory areas, complex buildings or open areas, it is difficult to achieve full-area and dead-angle-free monitoring. For example, the fire monitoring system based on unmanned aerial vehicles disclosed in CN105243627A introduces unmanned aerial vehicles for fire monitoring to make up for dead-angle fire monitoring, but still has the following defects:

[0004] 1) Poor data collection accuracy: unmanned aerial vehicle data collection is prone to environmental interference, and the data collected by the sensors carried during movement is not accurate enough. The fire judgment based solely on image recognition is limited by the development of technology and has poor judgment accuracy. In addition, the fixed-point sensors are also affected by distance interference and have inaccurate monitoring data, which seriously affects the prediction accuracy of fire risks. Moreover, since the data of fixed-point sensors and the data of unmanned aerial vehicle inspection are stored and analyzed independently, there is no effective dynamic integration mechanism, resulting in poor data correlation and difficulty in comprehensive judgment of fire risks.

[0005] 2) Extensive early warning mechanism: the traditional fire alarm uses a single threshold to trigger the alarm, lacks quantitative analysis of the probability of fire occurrence and development trend, and is prone to false positives and false negatives, affecting the accuracy of emergency response.

[0006] 3) Weak emergency linkage: after the alarm, it is difficult to quickly link sensors and unmanned monitoring equipment for intensive tracking, and the transmission of key information such as on-site fire equipment information and emergency contacts is not timely, delaying the disposal opportunity.

[0007] Therefore, a fire monitoring system is proposed, which integrates data collected by unmanned aerial vehicles and data collected by ground sensors to provide accurate data support for fire warning, and refines the warning mechanism by calculating the probability of fire occurrence and the time of possible fire occurrence, and sets up dynamic two-dimensional code labels to reduce the control intensity of the cloud while achieving efficient emergency linkage. SUMMARY

[0008] In view of the deficiencies of the prior art, the fire monitoring system is provided, which solves the problems of poor data collection accuracy, rough early warning mechanism and weak emergency linkage of the current fire monitoring mode.

[0009] To achieve the above object, the fire monitoring system is disclosed, comprising:

[0010] The ground sensing network is composed of a plurality of fire data monitoring modules arranged at fixed points, wherein the fire data monitoring module is used for collecting fixed-point fire data, including temperature information, smoke concentration information and gas concentration information, and the monitoring radius of the fire data monitoring module;

[0011] The unmanned aerial vehicle inspection module is used for inspecting each monitoring area after the monitoring area is divided according to the ground sensing network, and obtaining inspection fire data, including temperature information, smoke concentration information and gas concentration information, position information, video information, and the distance between the inspection fire data collection point and the nearest fixed point in the current monitoring area;

[0012] The data integration module is based on the dynamic calibration of the fixed-point fire data and the inspection fire data to obtain calibrated fire data, and the method comprises:

[0013]

[0014] W s =1-W w

[0015] In the formula, W w is the weight of the inspection fire data, W s is the weight of the fixed-point fire data, W0 is the initial weight of the inspection fire data, K is the distance influence adjustment factor, D is the distance between the inspection fire data collection point and the nearest fixed point in the current monitoring area, R is the monitoring radius of the fire data monitoring module, K1 is the environmental correction coefficient, and K1∈[0, 0.2];

[0016] After the fixed-point fire data and the inspection fire data are weighted and summed based on the weight, the calibrated temperature information, smoke concentration information and gas concentration information are obtained;

[0017] The fire warning module is based on the calibrated fire data to perform safety warning, comprising:

[0018] Fire occurrence probability calculation:

[0019]

[0020] In the formula, P is a fire occurrence probability, T is calibration temperature information, T0 is a current regional fire critical temperature threshold, S is calibration smoke concentration information, S0 is a current regional fire critical smoke concentration threshold, n is a total number of gas types, G i is the i-th calibration gas concentration information, G i，0 is a fire critical threshold of the i-th calibration gas concentration in the current region, alpha, beta and gamma i are weight coefficients;

[0021] Fire possible occurrence time calculation:

[0022]

[0023] In the formula, t is a fire possible occurrence time, K2 is a safety factor of the current region, and V is a data change rate.

[0024] When the fire occurrence probability exceeds a preset safety threshold, a hidden danger alarm is issued; and when the fire possible occurrence time is lower than a preset safety threshold, a fire alarm is issued.

[0025] A monitoring enhancement module controls the ground perception network and the unmanned aerial vehicle inspection module to perform high-frequency sampling adjustment based on the safety alarm information.

[0026] The application further provides that a dynamic two-dimensional code label is deployed in the monitoring region, and the dynamic two-dimensional code label stores unmanned aerial vehicle inspection routes, fire emergency equipment and state information, emergency contact persons and communication protocols.

[0027] The application further provides that the abnormal monitoring region includes safety alarm point coordinate, the safety alarm point coordinate is written into the unmanned aerial vehicle inspection route of the dynamic two-dimensional code label, and safety warning information of the safety alarm point coordinate in adjacent inspection periods is recorded; when repeated safety warning occurs, communication is performed with the emergency contact person, and video information of the safety alarm point coordinate is transmitted to the emergency contact person.

[0028] The application provides a fire monitoring system.

[0029] The application combines ground fixed point monitoring with unmanned aerial vehicle inspection to achieve full coverage of the monitoring range, eliminate monitoring dead angles, fuse and calibrate the two types of data by means of dynamic weight adjustment, improve the accuracy and relevance of the data, provide reliable basis for fire risk assessment, support hierarchical early warning by quantitatively calculating the fire occurrence probability and the possible occurrence time, make the early warning mechanism more fine, reduce false positives and false negatives, automatically trigger high-frequency sampling adjustment by the monitoring enhancement module after the alarm, form a continuous tracking closed loop for the abnormal region, realize efficient linkage and transmission of key information by combining the dynamic two-dimensional code label, speed up the emergency response process, and provide more reliable and intelligent protection for fire safety in different scenarios. Attached Figure Description

[0030] Figure 1 This is a schematic diagram of the system architecture of the present invention. Detailed Implementation

[0031] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0032] Please see Figure 1 The embodiments of the present invention provide the following technical solutions.

[0033] Example 1

[0034] In the field of fire safety monitoring, the accuracy of data is a crucial guarantee for fire early warning. Currently, in fire safety monitoring, whether it is data collected by drone inspections or data collected by fixed-point sensors, there are environmental interferences. In particular, data collected by drones is easily affected by environmental factors such as airflow and lighting, resulting in instantaneous deviations. Data collected by fixed-point sensors is limited by distance. Although they can collect data within an effective range, under environmental interference, the data becomes less accurate the further away from the sensor, the less accurate the data becomes.

[0035] Therefore, this embodiment provides a fire monitoring system, including a ground sensing network, an unmanned aerial vehicle (UAV) inspection module, and a data integration module. The ground sensing network consists of several fire data monitoring modules deployed at fixed locations. Each fire data monitoring module includes fiber optic sensors, smoke sensors, and gas sensors, which collect temperature, smoke concentration, and gas concentration information as fixed-location fire data. The gas concentration information includes CO, CH4, and O2 gas concentrations. For the fiber optic sensors, smoke sensors, and gas sensors, the smallest effective working radius is used as the monitoring radius of the fire data monitoring module. After obtaining the effective monitoring range, the effective monitoring range of the fire data monitoring modules deployed at the fixed location is taken as the ground monitoring area. The fixed-location fire data output format is: [timestamp, location ID, temperature, smoke concentration, CO concentration, CH4 concentration, O2 concentration, monitoring radius]. It is important to note that the deployment density of fire data monitoring modules should be adjusted according to the environment. For example, one module should be installed every 50m in open areas, and one module should be installed every 30m in complex buildings such as workshops and warehouses. The modules should also be kept at least 5m away from strong electromagnetic interference sources such as transformers and motors.

[0036] In order to ensure the comprehensiveness of fire monitoring, and to ensure the coverage of the fire data monitoring module as much as possible, and to eliminate the monitoring dead angle, there is at least a 5% overlapping area between the edges of two adjacent ground monitoring areas, and then the monitoring area is divided in the area where the ground sensing network is located, and the monitoring area contains at least three complete ground monitoring areas.

[0037] The unmanned aerial vehicle inspection module is used to realize the guarantee of regional comprehensive fire monitoring, and to conduct inspection in each monitoring area, and to obtain inspection fire data from the air. The unmanned aerial vehicle selects a four-rotor unmanned aerial vehicle, carries a data transmission module, and establishes a connection with the ground sensing network in the monitoring area through the MQTT protocol communication protocol. The four-rotor unmanned aerial vehicle is equipped with an infrared thermal imager, a smoke sensor, a gas sensor, a GPS positioning module and a high-definition camera. The temperature, smoke concentration, gas concentration information, position information and video information are collected as inspection fire data. The gas concentration information also includes CO, CH4 and O2 gas concentration.

[0038] An edge data processing station is arranged in the monitoring area. When the unmanned aerial vehicle inspection module transmits the inspection fire data to the edge data processing station, the distance between the current position and each fixed point in the current monitoring area is calculated according to the position information in the inspection fire data. The shortest distance of one fixed point is selected to correspond to the fire data monitoring module. After aligning the time characteristics, the temperature, smoke concentration and gas concentration information in the fixed position fire data and the inspection fire data are dynamically calibrated through the data integration module to obtain calibrated fire data. The dynamic calibration mode includes:

[0039]

[0040] W s =1-W w

[0041] In the formula, W w is the weight of the inspection fire data, W s is the weight of the fixed position fire data, W0 is the initial weight of the inspection fire data, K is the distance influence adjustment factor, D is the distance between the inspection fire data collection point and the nearest fixed point in the current monitoring area, R is the monitoring radius of the fire data monitoring module, K1 is the environmental correction coefficient, K1∈[0, 0.2], and the environmental correction coefficient is considered to be disturbed by environmental factors. For example, in the case of no wind and sunny weather, the environmental correction coefficient is 0, in the case of rainy weather, the environmental correction coefficient is 0.1, and in the case of strong wind, the environmental correction coefficient is 0.2. The stronger the environmental disturbance, the larger the value.

[0042] The calibration temperature information, the smoke concentration information and the gas concentration information are obtained by respectively calculating calibration values of the temperature, the smoke concentration, and the gas concentration in the fixed-point fire-fighting data and the inspection fire-fighting data based on the weights and then performing weighted summation. The storage format of the calibration fire-fighting data is: [timestamp, area ID, calibration temperature, calibration smoke concentration, calibration CO concentration, calibration CH4 concentration, calibration O2 concentration]. The calibration fire-fighting data is temporarily stored in the edge data processing station and is deleted after 30 days.

[0043] It should be noted that when the distances between the current position and each fixed point in the current monitoring area all exceed the effective monitoring range of the fire-fighting data monitoring module deployed on the corresponding fixed point, the inspection fire-fighting data is used as the criterion.

[0044] In this embodiment, the fixed-point fire-fighting data and the inspection fire-fighting data are integrated based on the weights, so that the comprehensive and effective collection of the fire-fighting related data is realized.

[0045] Embodiment 2

[0046] The traditional fire alarm uses a single threshold to trigger the alarm, lacks quantitative analysis of the fire occurrence probability and development trend, and is prone to false positives and false negatives, which affects the accuracy of emergency response.

[0047] Therefore, the embodiment provides a fire monitoring system, which comprises a fire warning module. The fire warning module performs safety warning based on the calibration fire-fighting data obtained in the embodiment 1. The safety warning is performed in the following manner:

[0048] Fire occurrence probability calculation:

[0049]

[0050] In the formula, P is the fire occurrence probability, T is the calibration temperature information, T0 is the current area fire critical temperature threshold, S is the calibration smoke concentration information, S0 is the current area fire critical smoke concentration threshold, n is the total number of gas types, G i is the i-th calibration gas concentration information, G i，0 is the i-th calibration gas concentration of the current area, and a, β and γ i are weight coefficients.

[0051] Fire possible occurrence time calculation:

[0052]

[0053] In the formula, t is the fire possible occurrence time, K2 is the safety coefficient of the current area, and the calculation formula is:

[0054]

[0055] In the formula, F is the current monitoring area fire-fighting facility intact rate, taking value 0-1, F0 is the standard intact rate, taking value 1, M is the current monitoring area combustible quantity coefficient, being set according to combustible types and quantity, taking value 0.5-2, M0 is the standard combustible quantity coefficient, taking value 1, C is the current monitoring area fire-fighting passage unobstructed rate, taking value 0-1, C0 is the standard unobstructed rate, taking value 1, w1, w2, w3 are weight coefficients;

[0056] V is the data change rate, after the change amount of temperature, smoke concentration, and each gas concentration per unit time is obtained, weighted summation is carried out, and the result is taken as the data change rate, such as:

[0057]

[0058] In the formula, Δt is the calculation time interval, is the change amount of temperature per unit time, is the change amount of smoke concentration per unit time, is the change amount of the i-th gas concentration per unit time, δ1, δ2, δ i+2 is the weight coefficient.

[0059] When the fire occurrence probability exceeds the preset safety threshold, a hidden danger alarm is issued; when the fire possible occurrence time is lower than the preset safety threshold, a fire alarm is issued.

[0060] After the fire warning and the quantitative analysis of the fire warning are implemented, a more refined warning mechanism can be formulated according to the fire occurrence probability and the fire possible occurrence time, and the safety alarm information is as follows:

[0061] The first-level alarm represents a low-risk prompt, and the triggering condition is P≤20% and t>30min;

[0062] The second-level alarm represents a general hidden danger warning, and the triggering condition is 20%<P≤40% or 15min≤t<30min;

[0063] The third-level alarm represents a high-risk warning, and the triggering condition is 40%<P≤60% or 5min≤t<15min;

[0064] The fourth-level alarm represents an emergency fire warning, and the triggering condition is P>60% or t<5min.

[0065] In the embodiment, the fire warning is carried out in the manner of quantifying the fire development trend, the warning mechanism is refined, the false alarm and missed alarm risks are reduced, and the alarm accuracy of the emergency fire condition is improved.

[0066] Embodiment 3

[0067] After the fire safety alarm, the response mainly relies on manual processing, including but not limited to contacting the safety person in charge, manually deploying fire emergency equipment, and analyzing the situation at the fire site, resulting in a delay in fire response.

[0068] To this end, the embodiment provides a fire monitoring system, which comprises a monitoring enhancement module, a ground sensing network and a UAV inspection module controlled based on safety alarm information for high-frequency sampling adjustment. In order to facilitate the control of the UAV, a dynamic two-dimensional code label is deployed inside the monitoring area, and the dynamic two-dimensional code label stores the UAV inspection route, fire emergency equipment and state information, emergency contact person and communication protocol, so as to accurately locate the person in charge of the current monitoring area and the nearest fire emergency equipment, shorten the disposal error, and further determine the accuracy of the fire alarm. When a hidden danger alarm and / or a fire alarm occurs, the corresponding monitoring area is recorded as an abnormal monitoring area, the sampling frequency of the fire data monitoring module in the abnormal monitoring area is increased to a set frequency, and the inspection cycle of the UAV inspection module is shortened to a set inspection cycle, avoiding the interference of the failure of the UAV inspection module. The abnormal monitoring area includes the safety alarm point coordinate, i.e. the corresponding position information in the inspection fire data, and the data format is: [abnormal ID, area ID, safety alarm point coordinate, alarm type, generation time]. At the same time, the safety alarm point coordinate is written into the UAV inspection route of the dynamic two-dimensional code label, and the UAV will fly to this coordinate point in the next inspection cycle. Through the record analysis of the safety warning information of the safety alarm point coordinate in the adjacent inspection cycle, when repeated safety warning occurs, communication is carried out with the emergency contact person, and video information of the safety alarm point coordinate is transmitted to the emergency contact person.

[0069] There are several large warehouses in the warehouse logistics park, which store various goods, and the overall fire monitoring of the warehouse and the surrounding area needs to be realized. When the monitoring system proposed by the application is applied in the warehouse logistics park, the following deployment is made:

[0070] I. In terms of ground sensing network, fixed-point fire data monitoring modules are arranged every 50m in the four corners of each warehouse and inside the warehouse. The fire data monitoring module is composed of a distributed optical fiber sensor, a smoke sensor and a gas sensor. The distributed optical fiber sensor is laid along the warehouse shelves and cable lines, with a monitoring radius of 30m, used for collecting temperature information; the intelligent smoke sensor and gas sensor are used to collect smoke concentration information and CO, CH4, O2 gas concentration information, and record the monitoring radius data at the same time.

[0071] II. The UAV inspection module divides the monitoring area according to the ground sensing network, each monitoring area contains three complete ground monitoring areas, the adjacent ground monitoring area edge overlaps by 5%, the UAV selects a multi-rotor UAV with a endurance time of 60 minutes, inspects according to the divided monitoring area, inspects once at 9 am and 3 pm each day, obtains inspection fire data, including temperature, smoke concentration, gas concentration information, and video information of the warehouse periphery and interior, records the distance between the inspection data collection point and the nearest fixed point, and sets a dynamic two-dimensional code label at the warehouse entrance and main channel, stores the UAV inspection route of the warehouse, the position and state information of the fire emergency equipment such as fire extinguishers and fire hydrants, the warehouse manager as an emergency contact and the corresponding communication protocol;

[0072] III. The data integration module performs dynamic calibration based on the fixed position fire data and the inspection fire data, the initial weight is set to 0.4, the distance influence adjustment factor is set to 0.3, and the environment correction coefficient is set to 0 when it is sunny. When the UAV inspection data collection point is 15m away from the nearest fixed point, the inspection fire data weight is 0.55, and the fixed position fire data weight is 0.45. The calibrated fire data is obtained by weighted summation;

[0073] IV. In the fire warning module, the fire critical temperature threshold is set to 70℃, the fire critical smoke concentration threshold is set to 0.2mg / m 3 , the fire critical thresholds of CO, CH4 and O2 gas concentrations are set to 50ppm, 1% and 19% respectively, and the corresponding weight coefficients of the fire critical temperature threshold, the fire critical smoke concentration threshold, the fire critical thresholds of CO, CH4 and O2 gas concentrations are set to 0.4, 0.3, 0.15, 0.1 and 0.05 respectively. When the calculated fire occurrence probability exceeds 30%, a hidden danger warning is issued. When the fire possible occurrence time t is less than 15 minutes, a fire warning is issued.

[0074] V. The monitoring enhancement module increases the sampling frequency of the fire data monitoring module in the abnormal monitoring area from 1 time / minute to 1 time / 5 seconds after receiving the warning information, shortens the UAV inspection period from 12h to 2h, and communicates with the emergency contact when a repeated safety warning occurs, and transmits the video information of the safety warning point coordinate.

[0075] Some data in the above formula is calculated by removing the dimension and taking its numerical value, and the formula is obtained by software simulation of a large amount of collected data to obtain a formula closest to the real situation; the preset parameters and preset thresholds in the formula are set by a person skilled in the art according to the actual situation or obtained by a large amount of data simulation.

[0076] The above examples are only used to illustrate the technical method of the present application but not limit the present application. Although the present application is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present application can be modified or equivalently replaced without departing from the spirit and scope of the technical method of the present application.

Claims

1. A fire monitoring system, characterized in that, include: The ground-based sensing network consists of several fire data monitoring modules deployed at fixed locations, with the fire data monitoring modules used to collect fire data from these fixed locations. The drone inspection module is used to inspect each monitoring area after dividing the monitoring area according to the ground sensing network and to obtain inspection fire data. The data integration module performs dynamic calibration based on fixed-position fire protection data and patrol fire protection data to obtain calibrated fire protection data; The fire early warning module issues safety alerts based on calibrated fire protection data; The monitoring enhancement module controls the ground perception network and the UAV inspection module to perform high-frequency sampling and adjustment based on security alarm information.

2. The fire monitoring system according to claim 1, characterized in that, The methods for dividing the monitoring area include: Centered on a fixed location, the effective monitoring range of the fire data monitoring module deployed on it is taken as the ground monitoring area; After obtaining the area where the ground sensing network is located, the monitoring area is divided. The monitoring area includes at least three complete ground monitoring areas, and there is at least 5% overlap between the edges of two adjacent ground monitoring areas.

3. A fire monitoring system according to claim 1, characterized in that, Both the fixed-position fire protection data and the patrol fire protection data include temperature information, smoke concentration information, and gas concentration information, and the gas concentration information includes the concentrations of CO, CH4, and O2 gases.

4. A fire monitoring system according to claim 3, characterized in that, The fixed-position fire protection data also includes the monitoring radius of the fire protection data monitoring module; The fire inspection data also includes location information, video information, and the distance between the fire inspection data collection point and the nearest fixed point within the current monitoring area.

5. A fire monitoring system according to claim 4, characterized in that, The method of dynamic calibration based on fixed-position fire protection data and patrol fire protection data includes: IN s =1-W w In the formula, W w To weight the fire inspection data, W s W0 is the initial weight of the fire protection data, K is the distance influence adjustment factor, D is the distance between the fire protection data collection point and the nearest fixed point in the current monitoring area, R is the monitoring radius of the fire protection data monitoring module, and K1 is the environmental correction coefficient, K1∈[0,0.2]. After weighting and summing the fixed-position fire protection data and the patrol fire protection data, calibration temperature information, smoke concentration information and gas concentration information are obtained.

6. A fire monitoring system according to claim 5, characterized in that, The method of generating safety alarms based on calibrated fire protection data includes: Fire probability calculation: In the formula, P represents the probability of fire occurrence, T represents the calibration temperature information, T0 represents the current area's critical fire temperature threshold, S represents the calibration smoke concentration information, S0 represents the current area's critical smoke concentration threshold, n represents the total number of gas types, and G... i For the i-th type of calibration gas concentration information, G i，0 Let α, β, and γ be the fire critical threshold for the i-th calibration gas concentration in the current region. i All are weighting coefficients; Calculation of possible fire time: In the formula, t is the time when the fire may occur, K2 is the safety factor of the current area, and V is the data change rate; When the probability of a fire exceeds its preset safety threshold, a hazard alarm is issued; when the time when a fire may occur is less than its preset safety threshold, a fire alarm is issued.

7. A fire monitoring system according to claim 6, characterized in that, The method of controlling the ground perception network and UAV inspection module to perform high-frequency sampling adjustment based on security alarm information includes: When a hazard alarm and / or fire alarm occurs, the corresponding monitoring area is designated as an abnormal monitoring area. The sampling frequency of the fire data monitoring module in the abnormal monitoring area is increased to the set frequency, and the inspection cycle of the drone inspection module is shortened to the set inspection cycle.

8. A fire monitoring system according to claim 7, characterized in that, Dynamic QR code tags are deployed within the monitored area. These tags store information such as drone inspection routes, fire emergency equipment and status information, emergency contacts, and communication protocols.

9. A fire monitoring system according to claim 8, characterized in that, The abnormal monitoring area includes the coordinates of safety alarm points. The coordinates of the safety alarm points are written into the drone inspection route of the dynamic QR code label, and the safety warning information of the safety alarm point coordinates in adjacent inspection cycles is recorded. When repeated safety warnings occur, communication is made with the emergency contact person and video information of the safety alarm point coordinates is transmitted to them.

Images