Method and apparatus for measuring target visibility

Abstract

The application provides a target visibility measurement method and device, and relates to the field of target measurement. The method comprises the following steps: performing segmentation processing on a target image according to material quality to obtain a plurality of segmented regions of different material quality; obtaining a first relationship between transmittance and visibility of segmented regions of the same material quality at different distances and a second relationship between transmittance and visibility of segmented regions of different material quality at the same distance; generating a transmittance and visibility relationship library based on the first relationship and the second relationship; obtaining attribute information of a target region, wherein the attribute information comprises transmittance corresponding to the material quality of the target region; obtaining distance information of the target region, wherein the distance information is a distance parameter of the target region from an observation point; obtaining a transmittance and visibility relationship corresponding to the target region from the transmittance and visibility relationship library based on the attribute information and the distance information; and calculating a visibility value of the target region according to the transmittance and visibility relationship.

Description

Technical Field

[0001] This invention relates to the field of target measurement, and more specifically, to a method and apparatus for measuring target visibility. Background Technology

[0002] Currently, visibility plays a crucial role in transportation and aircraft takeoff and landing. The commonly used visibility detection equipment is the infrared scattering visibility detector, but this equipment is expensive and can only measure visibility in the vicinity of the instrument, which cannot meet the requirements for large-scale visibility measurement.

[0003] Compared to the methods mentioned above, using cameras to measure visibility has significant advantages. Cameras offer a wide field of view, allowing for the measurement of visibility values ​​over a large area, and their low cost makes them suitable for widespread deployment. The main algorithms for camera-based visibility measurement include contrast measurement, luminance measurement, and neural network measurement. Contrast and luminance methods are susceptible to various factors, resulting in unstable measurement results. When using neural network methods, different observation targets have a significant impact on the visibility measurement results. Therefore, traditional visibility measurement methods suffer from low stability and are easily affected by surrounding environmental factors. Summary of the Invention

[0004] (a) Technical problems to be solved

[0005] To address the aforementioned problems, this invention provides a method and apparatus for measuring target visibility. By analyzing the visibility measurement parameters of observed targets of different materials and at different distances, the relationship between transmittance and visibility is estimated based on the target material and target distance, thereby obtaining an accurate visibility calculation relationship to calculate the target's visibility value.

[0006] (II) Technical Solution

[0007] One aspect of this invention provides a method for measuring target visibility, comprising: segmenting a target image according to its material to obtain multiple segmented regions of different materials; obtaining a first relationship between transmittance and visibility for segmented regions of the same material at different distances, and a second relationship between transmittance and visibility for segmented regions of different materials at the same distance; generating a transmittance-visibility relationship database based on the first and second relationships; obtaining attribute information of the target region, wherein the attribute information includes the transmittance corresponding to the material of the target region; obtaining distance information of the target region, wherein the distance information is a distance parameter between the target region and the observation point; obtaining the transmittance-visibility relationship corresponding to the target region from the transmittance-visibility relationship database based on the attribute information and the distance information; and calculating the visibility value of the target region based on the transmittance-visibility relationship.

[0008] In one embodiment of the present invention, obtaining distance information of a target area includes: using a full-energy sensor to obtain distance information of the target area, wherein using a full-energy sensor to obtain distance information of the target area includes: using a UAV to scan a digital model of the target area; performing virtual-real fusion processing on the digital model; obtaining distance parameters of all pixels in the target area based on the digital model after virtual-real fusion processing; calculating the average value of the distance parameters of all pixels in the target area to obtain the distance information of the target area.

[0009] In one embodiment of the present invention, segmenting a target image according to its material to obtain multiple segmented regions of different materials includes: using a neural network to segment the target image according to its material to obtain multiple segmented regions of different materials, wherein the neural network includes a Mask R-CNN neural network, a segNet neural network, or a target detection neural network.

[0010] In one embodiment of the present invention, obtaining the attribute information of the target region includes: obtaining the attribute information of the target region using a convolutional neural network, wherein the structure of the convolutional neural network includes an input layer, a hidden layer and an output layer; the input layer is used to normalize the target region; the hidden layer is used to classify and identify the normalized target region to obtain a classification and identification result; and the output layer is used to output the attribute information corresponding to the material of the target region according to the classification and identification result.

[0011] In one embodiment of the present invention, the hidden layer includes a first hidden layer and a second hidden layer, wherein the first hidden layer and the second hidden layer respectively include a convolutional layer, a pooling layer and a fully connected layer; the convolutional layer is used to extract features from the target region; the pooling layer is used to perform average pooling processing on the extracted features; and the fully connected layer is used to perform classification and recognition processing on the average pooling features to obtain classification and recognition results.

[0012] In one embodiment of the present invention, obtaining a first relationship between transmittance and visibility of segmented regions of the same material at different distances, and a second relationship between transmittance and visibility of segmented regions of different materials at the same distance, includes: obtaining the first relationship between transmittance and visibility of segmented regions of the same material at different distances and the second relationship between transmittance and visibility of segmented regions of different materials at the same distance through manual observation, wherein the duration of manual observation is at least six months.

[0013] In one embodiment of the present invention, obtaining the transmittance and visibility relationship of a target area from a transmittance and visibility relationship database based on attribute information and distance information includes: selecting the closest transmittance and visibility relationship corresponding to the target area from the transmittance and visibility relationship database based on attribute information and distance information as the transmittance and visibility relationship of the target area; or, selecting multiple closest transmittance and visibility relationships corresponding to the target area and calculating the average value as the transmittance and visibility relationship of the target area.

[0014] In one embodiment of the present invention, the visibility value of the target area is calculated based on the dark channel algorithm according to the relationship between transmittance and visibility.

[0015] In one embodiment of the present invention, before segmenting the target image according to the material to obtain multiple segmented regions of different materials, the method further includes: acquiring the target image using a visible light camera, an infrared camera, or a camera array.

[0016] Another aspect of this invention provides a target visibility measurement device, comprising: a segmentation module for segmenting a target image according to material to obtain multiple segmented regions of different materials; a first acquisition module for acquiring a first relationship between transmittance and visibility of segmented regions of the same material at different distances, and a second relationship between transmittance and visibility of segmented regions of different materials at the same distance; a generation module for generating a transmittance and visibility relationship library based on the first and second relationships; a second acquisition module for acquiring attribute information of a target region, wherein the attribute information includes the transmittance corresponding to the material of the target region; a third acquisition module for acquiring distance information of the target region, wherein the distance information is a distance parameter between the target region and the observation point; a fourth acquisition module for acquiring the transmittance and visibility relationship corresponding to the target region from the transmittance and visibility relationship library based on the attribute information and the distance information; and a calculation module for calculating the visibility value of the target region based on the transmittance and visibility relationship.

[0017] (III) Beneficial Effects

[0018] The target visibility measurement method and apparatus provided in this invention have at least the following beneficial effects:

[0019] (1) The target visibility measurement method and device provided in the embodiments of the present invention analyze the visibility measurement parameters of the observed targets of different materials and different distances, use a convolutional neural network to classify and identify the transmittance of the target material, use a high-energy sensor to obtain the distance parameters of the target area, estimate the relationship between transmittance and visibility based on the transmittance of the target material and the target distance parameters, thereby obtaining an accurate visibility calculation relationship to calculate the visibility value of the target. This measurement method has high stability and is not easily affected by the surrounding environmental factors.

[0020] (2) The target visibility measurement method and device provided in the embodiments of the present invention can be applied to any scenario since the target's material information and distance information can be identified, thus greatly expanding the scope of application.

[0021] (3) The target visibility measurement method and device provided in the embodiments of the present invention rely only on optical cameras. Optical cameras are inexpensive and easy to deploy over a wide area. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0023] Figure 1 The flowchart illustrating the target visibility measurement method provided by an embodiment of the present invention is shown schematically.

[0024] Figure 2 The diagram illustrates the segmentation process of a target image in the target visibility measurement method provided by an embodiment of the present invention.

[0025] Figure 3 The diagram illustrates the structure of the convolutional neural network in the target visibility measurement method provided by an embodiment of the present invention.

[0026] Figure 4 The diagram illustrates the structure of a target visibility measuring device provided in an embodiment of the present invention. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments and the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention. The terminology used herein is merely for describing specific embodiments and is not intended to limit the invention. The terms "comprising," "including," etc., used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0028] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection, an electrical connection, or a connection that allows communication between them; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0029] In the description of this invention, it should be understood that the terms "longitudinal", "length", "circumferential", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the subsystem or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0030] Throughout the accompanying drawings, identical elements are represented by the same or similar reference numerals. Conventional structures or configurations may be omitted where they might cause confusion in understanding the invention. Furthermore, the shapes, sizes, and positional relationships of the components in the drawings do not reflect actual size, scale, or actual positional relationships. Additionally, any reference numerals placed between parentheses in the claims should not be construed as limiting the claims.

[0031] Similarly, to simplify the invention and aid in understanding one or more aspects of the invention, in the above description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together in a single embodiment, figure, or description thereof. The use of terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples" indicates that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0032] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0033] Figure 1 The flowchart illustrating the target visibility measurement method provided by an embodiment of the present invention is shown schematically.

[0034] like Figure 1 As shown, the target visibility measurement method provided in this embodiment of the invention may include:

[0035] S1, segment the target image according to the material to obtain multiple segmented regions with different materials.

[0036] Before segmenting the target image according to its material to obtain multiple segmented regions of different materials, the process also includes: acquiring the target image using a visible light camera, an infrared camera, or a camera array.

[0037] S2, obtain the first relationship between transmittance and visibility of segmented regions of the same material at different distances, and the second relationship between transmittance and visibility of segmented regions of different materials at the same distance.

[0038] The first relationship between transmittance and visibility of segmented regions of the same material at different distances, and the second relationship between transmittance and visibility of segmented regions of different materials at the same distance, include:

[0039] The first relationship between transmittance and visibility of segmented regions of the same material at different distances is obtained by manual observation, and the second relationship between transmittance and visibility of segmented regions of different materials at the same distance is obtained, wherein the duration of the manual observation is at least six months.

[0040] S3 generates a database of transmittance and visibility relationships based on the first and second relationships.

[0041] S4, obtain the attribute information of the target area, including the transmittance of the material of the target area.

[0042] The acquisition of attribute information of the target region includes using a convolutional neural network (CNN). The CNN structure comprises an input layer, hidden layers, and an output layer. The input layer normalizes the target region; the hidden layers classify the normalized target region to obtain the classification result; and the output layer outputs the material attribute information of the target region based on the classification result.

[0043] S5, obtain the distance information of the target area, where the distance information is the distance parameter between the target area and the observation point.

[0044] Among them, obtaining distance information of the target area includes: using a full-energy sensor to obtain distance information of the target area.

[0045] Among them, the distance information of the target area obtained by using a rich-energy sensor includes:

[0046] The S50 1 uses drones to scan and obtain all digital models of the target area.

[0047] S502 performs virtual-real fusion processing on the digital model to ensure that the images captured by the camera system and the images captured by the virtual camera system are consistent.

[0048] S503 obtains the distance parameters of all pixels in the target area based on the digital model after virtual-real fusion processing.

[0049] S504 calculates the average distance parameter of all pixels in the target area to obtain the distance information of the target area.

[0050] S6. Based on attribute information and distance information, obtain the transmittance and visibility relationship corresponding to the target area from the transmittance and visibility relationship database.

[0051] Among them, obtaining the transmittance and visibility relationship of the target area from the transmittance and visibility relationship database based on attribute information and distance information includes:

[0052] Based on attribute and distance information, the closest transmittance and visibility relationship corresponding to the target area is selected from the transmittance and visibility relationship database as the transmittance and visibility relationship of the target area; or, multiple closest transmittance and visibility relationships corresponding to the target area are selected and the average value is calculated as the transmittance and visibility relationship of the target area.

[0053] S7 calculates the visibility value of the target area based on the relationship between transmittance and visibility.

[0054] The visibility values ​​of the target area calculated based on the relationship between transmittance and visibility include:

[0055] The visibility value of the target area is calculated based on the relationship between transmittance and visibility using the dark channel algorithm.

[0056] The target visibility measurement method provided in this invention analyzes the visibility measurement parameters of observed targets of different materials and at different distances. It uses a convolutional neural network to classify and identify the transmittance of the target material, uses a high-energy sensor to obtain the distance parameters of the target area, and estimates the relationship between transmittance and visibility based on the transmittance of the target material and the target distance parameters. This results in an accurate visibility calculation relationship to calculate the target's visibility value. This measurement method has high stability and is not easily affected by surrounding environmental factors.

[0057] The target visibility measurement method provided in this embodiment of the invention can be applied to any scenario since the target's material information and distance information can be identified, thus greatly expanding the scope of application.

[0058] The target visibility measurement method provided in this embodiment of the invention relies solely on an optical camera, which is inexpensive and easy to deploy over a wide area.

[0059] Figure 2 The diagram illustrates the segmentation process of a target image in the target visibility measurement method provided by an embodiment of the present invention.

[0060] like Figure 2 As shown, the target image segmentation process in the target visibility measurement method provided in this embodiment of the invention includes:

[0061] The target image is segmented based on material using a neural network to obtain multiple segmented regions with different materials. The neural network includes Mask R-CNN neural network, segNet neural network, or FPN (object detection) neural network.

[0062] Figure 3 The diagram illustrates the structure of the convolutional neural network in the target visibility measurement method provided by an embodiment of the present invention.

[0063] like Figure 3 As shown, the structure of a convolutional neural network includes an input layer, hidden layers, and an output layer.

[0064] The input layer is used to normalize the target region.

[0065] The hidden layer is used to classify and identify the normalized target region to obtain the classification and identification results.

[0066] The output layer is used to output the attribute information corresponding to the material of the target area based on the classification and recognition results.

[0067] The hidden layer includes a first hidden layer and a second hidden layer.

[0068] The first and second hidden layers each consist of a convolutional layer, a pooling layer, and a fully connected layer, respectively.

[0069] Convolutional layers are used to extract features from target regions.

[0070] Pooling layers are used to perform average pooling on the extracted features.

[0071] The fully connected layer is used to classify and recognize the features after average pooling to obtain the classification and recognition results.

[0072] The input to a convolutional neural network is a segmented image of the material region, and the output is attribute information, namely material parameters, which reflect the strength of the material's reflectivity.

[0073] Figure 4 The diagram illustrates the structure of a target visibility measuring device provided in an embodiment of the present invention.

[0074] like Figure 4 As shown, the target visibility measurement device provided in this embodiment of the invention may include: a segmentation module 401, a first acquisition module 402, a generation module 403, a second acquisition module 404, a third acquisition module 405, a fourth acquisition module 406, and a calculation module 407.

[0075] The segmentation module 401 is used to segment the target image according to its material to obtain multiple segmentation regions with different materials.

[0076] The first acquisition module 402 is used to acquire a first relationship between transmittance and visibility of segmented regions of the same material at different distances, and a second relationship between transmittance and visibility of segmented regions of different materials at the same distance.

[0077] The generation module 403 is used to generate a library of transmittance and visibility relationships based on the first and second relationships.

[0078] The second acquisition module 404 is used to acquire attribute information of the target area, wherein the attribute information includes the transmittance corresponding to the material of the target area.

[0079] The third acquisition module 405 is used to acquire distance information of the target area, wherein the distance information is the distance parameter between the target area and the observation point.

[0080] The fourth acquisition module 406 is used to acquire the transmittance and visibility relationship corresponding to the target area from the transmittance and visibility relationship database based on attribute information and distance information.

[0081] The calculation module 407 is used to calculate the visibility value of the target area based on the relationship between transmittance and visibility.

[0082] According to embodiments of the present invention, any plurality of modules among the segmentation module 401, the first acquisition module 402, the generation module 403, the second acquisition module 404, the third acquisition module 405, the fourth acquisition module 406, and the calculation module 407 can be merged into one module, or any one of these modules can be split into multiple modules, or at least part of the functionality of one or more of these modules can be combined with at least part of the functionality of other modules and implemented in one module. According to embodiments of the present disclosure, at least one of the segmentation module 401, the first acquisition module 402, the generation module 403, the second acquisition module 404, the third acquisition module 405, the fourth acquisition module 406, and the calculation module 407 can be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or implemented in hardware or firmware by any other reasonable means of integrating or packaging the circuitry, or implemented in any one of software, hardware, and firmware methods, or in a suitable combination of any of these. Alternatively, at least one of the preprocessing module, extraction module, generation module, detection module, and measurement module can be at least partially implemented as a computer program module, which can perform corresponding functions when the computer program module is run.

[0083] It should be noted that the target visibility measurement device in the embodiments of the present invention corresponds to the target visibility measurement method in the embodiments of the present invention, and their specific implementation details and the resulting technical effects are the same, which will not be repeated here.

[0084] Although the invention has been illustrated and described in detail in the accompanying drawings and the foregoing description, such illustrations and descriptions should be considered illustrative or exemplary rather than restrictive.

[0085] Those skilled in the art will understand that the features described in the various embodiments and / or claims of the present invention can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in the present invention. In particular, the features described in the various embodiments and / or claims of the present invention can be combined and / or combined in various ways without departing from the spirit and teachings of the present invention. All such combinations and / or combinations fall within the scope of the present invention.

[0086] Although the invention has been shown and described with reference to specific exemplary embodiments thereof, those skilled in the art will understand that various changes in form and detail may be made to the invention without departing from the spirit and scope of the invention as defined by the appended claims and their equivalents. Therefore, the scope of the invention should not be limited to the above embodiments, but should be determined not only by the appended claims but also by their equivalents.

Claims

1. A method of measuring the visibility of a target, characterized in that, The method comprises the following steps: segmenting a target image according to materials to obtain segmented regions of different materials; obtaining a first relationship between transmittance and visibility of segmented regions of the same material at different distances, and a second relationship between transmittance and visibility of segmented regions of different materials at the same distance; generating a transmittance-visibility relationship library based on the first relationship and the second relationship; obtaining attribute information of a target region by using a convolutional neural network, wherein the structure of the convolutional neural network comprises an input layer, a hidden layer and an output layer; the input layer is used for normalizing the target region; the hidden layer is used for classifying and identifying the normalized target region to obtain a classification and identification result; the output layer is used for outputting attribute information corresponding to the material of the target region according to the classification and identification result; the attribute information comprises transmittance corresponding to the material of the target region; obtaining distance information of the target region by using a rich energy sensor, wherein the distance information is a distance parameter of the target region from an observation point; obtaining a transmittance-visibility relationship corresponding to the target region from the transmittance-visibility relationship library based on the attribute information and the distance information; calculating a visibility value of the target region according to the transmittance-visibility relationship.

2. The method of measuring the target visibility according to claim 1, characterized in that, The method for obtaining distance information of the target region by using a rich energy sensor comprises the following steps: scanning a digital model of the target region by using a drone; performing virtual-real fusion processing on the digital model; obtaining distance parameters of all pixels in the target region according to the virtual-real fusion processed digital model; calculating an average value of the distance parameters of all pixels in the target region to obtain distance information of the target region.

3. The method of measuring the target visibility according to claim 1, wherein, The method for segmenting a target image according to materials to obtain segmented regions of different materials comprises the following steps: segmenting a target image according to materials by using a neural network to obtain segmented regions of different materials, wherein the neural network comprises a Mask R-CNN neural network, a segNet neural network or a target detection neural network.

4. The method of measuring the target visibility according to claim 1, wherein, The hidden layer comprises a first hidden layer and a second hidden layer, wherein the first hidden layer and the second hidden layer each comprise a convolutional layer, a pooling layer and a fully connected layer; the convolutional layer is used for feature extraction of the target region; the pooling layer is used for average pooling processing of the extracted features; the fully connected layer is used for classification and identification processing of the average pooling processed features to obtain a classification and identification result.

5. The method of measuring the target visibility according to claim 1, wherein, The method for obtaining a first relationship between transmittance and visibility of segmented regions of the same material at different distances, and a second relationship between transmittance and visibility of segmented regions of different materials at the same distance comprises the following steps: obtaining a first relationship between transmittance and visibility of segmented regions of the same material at different distances, and a second relationship between transmittance and visibility of segmented regions of different materials at the same distance by using artificial observation, wherein the artificial observation lasts at least half a year.

6. The method of measuring the target visibility according to claim 1, wherein, The method for obtaining a transmittance-visibility relationship of the target region from the transmittance-visibility relationship library based on the attribute information and the distance information comprises the following steps: The attribute information and the distance information are used to select a closest transmittance-visibility relationship corresponding to the target region from the transmittance-visibility relationship library as the transmittance-visibility relationship of the target region, or to select multiple closest transmittance-visibility relationships corresponding to the target region and obtain an average value as the transmittance-visibility relationship of the target region.

7. The method of measuring the target visibility according to claim 1, wherein, The transmittance-visibility relationship is used to calculate a visibility value of the target region. The transmittance-visibility relationship is used to calculate a visibility value of the target region based on a dark channel algorithm.

8. The method of measuring the target visibility according to claim 1, wherein, Before the target image is segmented according to the material to obtain multiple segmented regions of different materials, the method further includes: The target image is acquired by using a visible light camera, an infrared camera or a camera array.

9. A device for measuring the visibility of a target, characterized in that The method includes: segmenting the target image according to the material to obtain multiple segmented regions of different materials; a first acquisition module is configured to acquire a first relationship between transmittance and visibility of a segmented region of the same material at different distances and a second relationship between transmittance and visibility of a segmented region of different materials at the same distance; a generation module is configured to generate a transmittance-visibility relationship library based on the first relationship and the second relationship; a second acquisition module is configured to acquire attribute information of a target region by using a convolutional neural network, where a structure of the convolutional neural network includes an input layer, a hidden layer and an output layer; the input layer is configured to perform normalization processing on the target region; the hidden layer is configured to perform classification and recognition processing on the normalized target region to obtain a classification and recognition result; the output layer is configured to output attribute information corresponding to a material of the target region according to the classification and recognition result; the attribute information includes transmittance corresponding to the material of the target region; a third acquisition module is configured to acquire distance information of the target region by using a rich energy sensor, where the distance information is a distance parameter of the target region from an observation point; a fourth acquisition module is configured to acquire a transmittance-visibility relationship corresponding to the target region from the transmittance-visibility relationship library based on the attribute information and the distance information; a calculation module is configured to calculate a visibility value of the target region according to the transmittance-visibility relationship.

Images