Cloud layer measuring device and cloud layer measuring method

Abstract

To collectively provide information about a cloud shape and a cloud height.SOLUTION: A cloud shape and cloud layer measurement device comprises: an in-image position identification unit which compares a cloud shape map showing a position of a cloud and a cloud shape in an image with a relative cloud height map showing a position of a cloud and a relative cloud height in an image to identify a positional relation between the cloud shape map and the relative cloud height map on the basis of the comparison result; and a cloud layer estimation unit which separates the cloud shape shown by the cloud shape map from the relative cloud height by using the positional relation to output cloud shape and cloud layer information in which the cloud shape is associated with the relative cloud height.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a cloud layer measuring device and a cloud layer measuring method. [Background technology]

[0002] Traditionally, an important observation item in aviation weather observation is what type of clouds exist at what altitude. Specifically, for groups of clouds that exist in the same altitude band, information summarizing the cloud shape and cloud base altitude is provided as cloud layer information. This information was previously recorded by observers through visual observation, but automating this process can reduce labor costs and contribute to helping observers who require a lot of experience become proficient.

[0003] For this reason, models have been developed that estimate cloud shapes through image recognition of visible light images. Furthermore, there is a technology for measuring cloud height described in JP 2004-170350 A (Patent Document 1). This publication states that "light from an observation target is received by light receiving means having a predetermined viewing angle from an arbitrary point, and temperature information of the observation target is calculated by a predetermined calculation means. In particular, when the light receiving means is an infrared radiation thermometer and has an optical filter with a main transmission band of 8 to 13 μm, the effects of H2O and CO2 in the air are significantly reduced, enabling highly accurate observation. Furthermore, it is also possible to calculate the altitude of the cloud base based on such temperature information and apply it to frost prediction and rainfall prediction." [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2004-170350 A Summary of the Invention [Problem to be solved by the invention]

[0005] However, even if a cloud shape map is created by image recognition of visible light images, it is not possible to obtain information on cloud height. For example, if two cloud layers are formed and both have the same cloud shape, all areas of the cloud shape map will be the same. On the other hand, although an infrared camera can obtain the relative cloud height of clouds, image features such as the cloud texture are lost, making it difficult to recognize the cloud shape. Therefore, an object of the present invention is to provide comprehensive information regarding cloud shape and cloud height. [Means for solving the problem]

[0006] In order to achieve the above-mentioned object, one of the representative cloud shape and cloud layer measurement devices of the present invention is characterized by comprising an image position identification unit that compares a cloud shape map indicating the position and cloud shape of clouds in an image with a relative cloud height map indicating the position and relative cloud height of clouds in an image, and identifies the positional relationship between the cloud shape map and the relative cloud height map based on the result of the comparison, and a cloud layer estimation unit that uses the positional relationship to separate the cloud shapes shown in the cloud shape map by the relative cloud height, and outputs cloud shape and cloud layer information that corresponds to the cloud shape and the relative cloud height. Furthermore, one of the representative cloud shape and cloud layer measurement methods of the present invention is characterized in that the device includes an in-image position identification step in which a cloud shape map indicating the position and cloud shape of clouds in an image is compared with a relative cloud height map indicating the position and relative cloud height of clouds in an image, and based on the result of the comparison, identifies the positional relationship between the cloud shape map and the relative cloud height map, and a cloud layer estimation step in which the device uses the positional relationship to separate the cloud shapes shown in the cloud shape map by the relative cloud height, and outputs cloud shape and cloud layer information that corresponds to the cloud shape and the relative cloud height. Effect of the Invention

[0007] According to the present invention, information on cloud shape and cloud height can be provided comprehensively. Problems, configurations and effects other than those described above will become apparent from the following description of the embodiments. [Brief description of the drawings]

[0008] [Figure 1] Functional configuration diagram of the cloud layer measurement device [Diagram 2] Hardware configuration of the cloud layer measurement device [Diagram 3] Illustration of cloud boundary based localization [Figure 4] Flowchart showing a process for identifying a positional relationship according to priority order [Diagram 5] Flowchart showing the process of merging cloud height [Figure 6] Illustration of location identification based on the circumscribing rectangle of a cloud [Figure 7] Illustration of cloud layer estimation from temperature histogram DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] Hereinafter, the embodiments will be described with reference to the drawings. EXAMPLES

[0010] FIG. 1 is a functional configuration diagram of a cloud layer measuring device according to an embodiment. The cloud layer measuring device 10 shown in Fig. 1 includes a visible light camera 11 and an infrared camera 12. The cloud layer measuring device 10 also includes functional units including a visible light image receiving unit 21, an infrared image receiving unit 22, a cloud shape estimating unit 23, a cloud base temperature calculating unit 24, an in-image position identifying unit 25, and a cloud layer estimating unit 26.

[0011] The visible light camera 11 and the infrared camera 12 are imaging devices that capture images of the sky, and their imaging ranges at least partially overlap. The visible light image receiving unit 21 acquires the visible light image captured by the visible light camera 11 and outputs it to the cloud shape estimating unit 23. The infrared image receiving unit 22 acquires the infrared image captured by the infrared camera 12 and outputs it to the cloud base temperature calculation unit 24 .

[0012] The cloud shape estimation unit 23 uses a cloud shape recognition model 31 created in advance by machine learning, for example, to estimate the cloud shape of the cloud in the visible light image, and generates a cloud shape map 32 indicating the position and shape of the cloud. Here, the cloud shape is a classification of clouds according to their shape. For example, in a classification called ten types of cloud shapes, cloud shapes are classified into any one of "cirrus, cirrocumulus, cirrostratus, altocumulus, altostratus, nimbostratus, stratocumulus, stratus, cumulonimbus".

[0013] The cloud base temperature calculation unit 24 generates a relative cloud height map 33 that indicates the positions of clouds and the relative cloud heights in the infrared image. It is known that the cloud base temperature corresponds to the cloud height. Since it is affected by atmospheric conditions, it is difficult to obtain the absolute height from the cloud base temperature, but when multiple cloud layers exist, it is possible to identify the relative heights between the cloud layers. Therefore, the cloud base temperature calculation unit 24 measures the temperature of the cloud base from the infrared image, calculates the relative cloud height of the clouds in the infrared image, and outputs a relative cloud height map 33.

[0014] The intra-image position identification unit 25 compares the cloud shape map 32 with the relative cloud height map 33, and identifies the positional relationship between the cloud shape map 32 and the relative cloud height map 33 based on the result of the comparison. For example, if mountain ridges, buildings, etc. are included in both the cloud shape map 32 and the relative cloud height map 33, these objects can be used as reference objects to identify the positional relationship between the cloud shape map 32 and the relative cloud height map 33. However, since the images to be processed are images of the sky, it cannot necessarily be expected that the cloud shape map 32 and the relative cloud height map 33 will include an image of the reference object. The intra-image position identification unit 25 can identify the positional relationship between the cloud shape map 32 and the relative cloud height map 33 without relying on a reference object.

[0015] Specifically, the in-image position identification unit 25 identifies the positional relationship based on the area occupied by the cloud. As an example, the in-image position identification unit 25 compares the cloud boundary in the cloud shape map 32 with the cloud boundary in the relative cloud height map 33, and aligns the cloud boundary so that the degree of coincidence exceeds a threshold.

[0016] Here, for example, when the same cloud shape exists at different cloud heights, the boundaries of the relative cloud height map 33 are generated along the boundaries of the cloud heights, but the boundaries of the cloud shape map 32 are generated along the cloud shapes. For this reason, a situation may occur where a boundary exists in the relative cloud height map 33, but a corresponding boundary does not exist in the cloud shape map 32, and in such a situation, the boundary comparison cannot be performed properly. Therefore, the intra-image position identification unit 25 performs a merging process to combine different cloud heights shown in the relative cloud height map 33 , and compares the cloud boundary obtained by the merging with the boundary of the cloud shape map 32 . Regarding cloud shape and cloud height, it is possible to make the assumptions that "only one cloud shape appears at one cloud height" and "the same cloud shape can appear at multiple cloud heights." The image position identification unit 25 determines combination patterns of cloud heights using these assumptions as constraints, and searches for combinations whose degree of match exceeds a threshold by comparing the boundaries of each pattern with the boundaries of the cloud shape map 32.

[0017] In addition, depending on the cloud shape, the intensity of infrared radiation may be insufficient and the cloud may not appear in the infrared image. In such a case, a boundary may exist in the cloud shape map 32, but the corresponding boundary may not exist in the relative cloud height map 33. In such a case, the boundary comparison cannot be performed properly. Therefore, the intra-image position identification unit 25 excludes certain types of cloud shapes from the cloud shapes shown in the cloud shape map 32 and compares it with the relative cloud height map 33.

[0018] The cloud layer estimation unit 26 separates the cloud shapes shown in the cloud shape map 32 by relative cloud height using the positional relationship determined by the in-image position identification unit 25, and outputs cloud shape / cloud layer information 34 that associates the cloud shapes with the relative cloud heights. That is, the cloud shape and cloud layer information 34 output from the cloud shape and cloud layer measuring device 10 indicates how many cloud layers exist, the cloud shape of each layer, and the relative relationship between the heights of each layer.

[0019] 2 is a hardware configuration diagram of the cloud layer measurement device according to the embodiment. As shown in FIG. 2, the cloud layer measurement device 10 includes a display unit 13, an input unit 14, a CPU (Central Processing Unit) 15, a main memory device 16, and an auxiliary memory device 17 in addition to the visible light camera 11 and the infrared camera 12 already described.

[0020] The display unit 13 is, for example, a liquid crystal display, and is used to display and output various information to the operator of the cloud type and cloud layer measuring device 10. The cloud type and cloud layer information 34 can also be output by displaying it on the display unit 13. The input unit 14 is, for example, a keyboard, and receives operational inputs from an operator.

[0021] The CPU 15 can implement various functions by loading various programs and data into the main storage device 16 and sequentially executing the programs. The main storage device 16 is, for example, a RAM (Random Access Memory). Specifically, the CPU 15 deploys and executes a cloud shape and cloud layer measurement program in the main memory device 16, thereby realizing the functions of a visible light image receiving unit 21, an infrared image receiving unit 22, a cloud shape estimation unit 23, a cloud base temperature calculation unit 24, an in-image position identification unit 25, and a cloud layer estimation unit 26.

[0022] The auxiliary storage device 17 is, for example, a hard disk drive, and stores various data. Specifically, the auxiliary storage device 17 stores a cloud shape recognition model 31, a cloud shape map 32, a relative cloud height map 33, and cloud shape / cloud layer information 34, etc.

[0023] 3 is an explanatory diagram of position identification based on cloud boundaries, showing a cloud shape map 32 and a relative cloud height map 33 created by capturing images of the same sky. The cloud shape map 32 includes cloud shapes T1, T2, and T3, while the relative cloud height map 33 shows cloud heights H1, H2, and H3.

[0024] Clouds with cloud heights H2 and H3 are the same cloud type T2. Therefore, the boundary between cloud heights H2 and H3 on the relative cloud height map does not exist on the cloud type map. Therefore, the cloud layer measuring device 10 searches for a combination pattern of cloud heights and merges the cloud heights to generate a relative cloud height map 33a. The search for a combination pattern will be described later. Figure 3 shows a relative cloud height map 33a in which the cloud heights H2 and H3 are merged.

[0025] Cloud type T3 in the cloud type map 32 is, for example, a gaseous stratus cloud, and due to insufficient infrared radiation intensity, is not included in the relative cloud height map 33. Clouds that are not expected to appear in the relative cloud height map 33 can be identified by their cloud shapes. Therefore, the cloud-form layer measuring device 10 sets a predetermined type of cloud form as an exclusion target in advance, and generates a relative cloud height map 33a by excluding the cloud form of the exclusion target type from the cloud form map 32. Figure 3 shows the cloud form map 32a excluding the cloud form T3.

[0026] The cloud shape and cloud layer measuring device 10 compares the cloud boundary lines in the relative cloud height map 33a after merging the cloud heights with the cloud boundary lines in the cloud shape map 32a after excluding a specific cloud shape, and performs alignment so that the degree of agreement exceeds a threshold value.

[0027] The cloud type cloud layer measurement device 10 can use a combination of the following methods: identification of the positional relationship based on a reference object, identification of the positional relationship based on the cloud boundary, merging of cloud heights, and exclusion of a certain type of cloud shape. The priority order of these identification methods of the positional relationship may be set.

[0028] 4 is a flowchart showing the process of identifying the positional relationship according to the priority order. First, the image position identifying unit 25 of the cloud layer measuring device 10 identifies the position of a building or a mountain ridge as a reference object (step S101). If the position identification by the reference object is performed, the process can be ended as it is.

[0029] If position identification using the reference object is not possible, the in-image position identification unit 25 extracts and compares boundary lines from the cloud shape map 32 and the relative cloud height map 33 to identify the position (step S102). If position identification using boundary lines can be performed without merging cloud heights or excluding a specific cloud shape, that is, if the degree of coincidence of the boundary lines exceeds a threshold, the process can be terminated.

[0030] If the degree of coincidence of the boundary line does not exceed the threshold in step S102, the intra-image position identification unit 25 merges the cloud height, compares the boundary line, and performs position identification (step S103). As a result, if the degree of coincidence of the boundary line exceeds the threshold, the process can be ended as it is.

[0031] If the degree of coincidence of the boundary lines does not exceed the threshold even after merging of the cloud heights, the intra-image position identifying unit 25 compares the boundary lines while excluding a predetermined type of cloud height, and performs position identification (step S104).

[0032] In Fig. 4, the cloud layer measurement device 10 prioritizes position identification using a reference object, because if a reference object exists, position identification can be reliably performed with high accuracy. When position identification by the reference object is not possible, the cloud type cloud layer measurement device 10 adopts position identification by the boundary line. Then, if the degree of coincidence does not exceed the threshold, it adopts cloud height merging and cloud type exclusion in that order. The reason for lowering the priority of cloud type exclusion is that even a certain type of cloud type may appear in the infrared image, and excluding the cloud type may lead to a decrease in the degree of coincidence. It should be noted that the priority order shown in FIG. 4 is merely an example, and the method for identifying the positional relationship can be selected arbitrarily.

[0033] 5 is a flowchart showing the procedure for merging cloud heights. The in-image position identification unit 25 first acquires the cloud shape map 32 and the relative cloud height map 33 (step S201). Next, the in-image position identification unit 25 extracts boundary lines from both maps (step S202).

[0034] The intra-image position identifying unit 25 calculates the degree of coincidence of the extracted boundary lines (step S203), and determines whether the degree of coincidence exceeds a threshold value (step S204).

[0035] If the degree of coincidence does not exceed the threshold in step S204 (step S204; No), the intra-image position identification unit 25 selects and merges a combination of cloud heights in the relative cloud height map 33 (step S205). After that, by returning to step S202, a boundary line is extracted from the relative cloud height map 33a and the cloud shape map 32 in which the cloud heights have been merged.

[0036] Each time the degree of coincidence does not exceed the threshold and the process proceeds to step S205, the in-image position identifying unit 25 selects a different combination. Then, if the degree of coincidence exceeds the threshold value (step S205; Yes), the intra-image position identifying unit 25 determines the position where both maps are superimposed (step S206), and ends the process.

[0037] The cloud layer estimation unit 26 overlays the cloud shape map 32 and the relative cloud height map 33 at the overlay position determined by the in-image position identification unit 25, separates the cloud shapes shown in the cloud shape map 32 by the relative cloud height, and outputs cloud shape / cloud layer information 34 that corresponds the cloud shapes to the relative cloud heights. The relative cloud height map 33 used for overlay is the relative cloud height map 33 before cloud height merging. Also, the cloud shape map 32 for separating cloud shapes is the cloud shape map 32 before excluding cloud shapes of a certain type. Therefore, cloud layers that are not associated with cloud heights may be included in the cloud shape / cloud layer information 34. As an example, the cloud shape and cloud layer information 34 obtained from the cloud shape map 32 and the relative cloud height map 33 in FIG. First layer: Cloud type T1, relative altitude H1 Second layer: Cloud type T2, relative altitude H2 3rd layer: Cloud shape T2, relative height H3 4th layer: Cloud type T3, relative altitude unknown (The relative altitude is H1 <H2<H3)

[0038] In the above explanation, the case where the positional relationship based on the area occupied by the cloud is identified by comparing the cloud boundary lines in the cloud shape map 32 and the relative cloud height map 33 has been exemplified, but the present invention is not limited to this. As another example, the positional relationship based on the area occupied by the cloud can also be identified by comparing a rectangle circumscribing the cloud in the cloud shape map 32 with a rectangle circumscribing the cloud in the relative cloud height map 33.

[0039] 6 is an explanatory diagram of position identification based on the circumscribing rectangle of a cloud. First, the cloud shape and cloud layer measurement device 10 performs cloud shape removal and cloud height merging in the same manner as in FIG. 3 to generate a cloud shape map 32a and a relative cloud height map 33a.

[0040] Thereafter, the cloud shape cloud layer measuring device 10 detects a rectangle circumscribing the cloud in the cloud shape map 32a and a rectangle circumscribing the cloud in the relative cloud height map 33a, and sets them as rectangular areas. FIG. 6 shows a cloud shape map 32b in which a rectangular region is set, and a relative cloud height map 33b in which a rectangular region is set. Rectangular regions A1 to A3 are set in the cloud shape map 32b, and rectangular regions B1 to B3 are set in the relative cloud height map 33b.

[0041] The cloud layer measuring device 10 compares the rectangular regions A1 to A3 with the rectangular regions B1 to B3, and performs positioning so that the degree of coincidence exceeds a threshold value. By comparing rectangular regions in this way, the processing load can be reduced compared to comparing boundary lines. When rectangular regions are used, the search for combinations in cloud height merging can also be performed by comparing rectangular regions.

[0042] Next, the creation of the relative cloud height map 33 will be described. The cloud base temperature calculation unit 24 determines areas of the same cloud height by grouping areas of the same temperature zone. If the infrared camera 12 has a high temperature resolution and the temperature is not constant even within the same cloud layer, a temperature histogram is created and the number of groups is determined based on the number of peaks that appear. General clustering may then be performed. The clustering may be a method of automatically determining the number of groups. Alternatively, the allowable difference from the most frequent temperature in each cloud layer may be manually determined.

[0043] Figure 7 is an explanatory diagram of estimating cloud layers from a temperature histogram. In Figure 7, the number of pixels is counted for each temperature to create a temperature histogram. The temperature histogram in Figure 7 has three peaks, and it can be estimated that there are three cloud layers L1 to L3 corresponding to the three peaks.

[0044] As described above, the disclosed cloud shape and cloud layer measuring device 10 is characterized by comprising an image position identification unit 25 that compares a cloud shape map 32 indicating the position and cloud shape of clouds in an image with a relative cloud height map 33 indicating the position and relative cloud height of clouds in an image, and identifies the positional relationship between the cloud shape map 32 and the relative cloud height map 33 based on the result of the comparison, and a cloud layer estimation unit 26 that uses the positional relationship to separate the cloud shapes shown in the cloud shape map 32 by the relative cloud height, and outputs cloud shape and cloud layer information 34 that corresponds the cloud shapes to the relative cloud heights. With such a configuration and operation, the cloud shape and cloud layer measuring device 10 can comprehensively provide information regarding cloud shape and cloud height.

[0045] In addition, the disclosed cloud shape cloud layer measurement device 10 further includes a cloud shape estimation unit 23 that estimates the cloud shape of a cloud in a visible light image using a previously created cloud shape recognition model 31 and outputs the cloud shape map 32, and a cloud base temperature calculation unit 24 that measures the cloud base temperature from the output of the infrared image captured by the infrared camera 12, calculates the relative cloud height, and outputs the relative cloud height map 33. Therefore, by using the cloud shape obtained from the cloud shape recognition model 31 and the temperature of the cloud ladder obtained from the external radiation image, the cloud shape and the relative cloud height can be output in correspondence with each other.

[0046] In addition, in the disclosed cloud shape layer measuring device 10, the in-image position identifying unit 25 compares the cloud boundary line in the cloud shape map 32 with the cloud boundary line in the relative cloud height map 33 to identify the positional relationship. Therefore, the positional relationship between the cloud shape map 32 and the relative cloud height map 33 can be identified regardless of the presence or absence of a reference object.

[0047] In addition, in the disclosed cloud shape layer measuring device 10, the image position identifying unit 25 combines clouds of different cloud heights shown in the relative cloud height map 33 and compares them with the cloud shape map. This makes it possible to make a comparison with the cloud shape map 32 that does not have cloud height information.

[0048] In addition, in the disclosed cloud shape and cloud layer measuring device 10, the image position identification unit 25 excludes a predetermined type of cloud shape from the cloud shapes shown in the cloud shape map 32 and compares it with the relative cloud height map 33. Therefore, even when a cloud shape that is not captured in an infrared image is included, the cloud shape map 32 and the relative cloud height map 33 can be compared.

[0049] In addition, in the disclosed cloud shape and cloud layer measuring device 10, the in-image position identifying unit 25 identifies the positional relationship by using an object that is commonly included in the cloud shape map 32 and the relative cloud height map 33 as a reference object. Therefore, when a reference object is present, alignment can be performed easily and with high accuracy.

[0050] In addition, in the disclosed cloud-shaped cloud layer measurement device 10, the image position identification unit 25 compares a rectangle circumscribing the cloud in the cloud shape map 32 with a rectangle circumscribing the cloud in the relative cloud height map 33 to identify the positional relationship. Therefore, the positional relationship between the cloud shape map 32 and the relative cloud height map 33 can be identified without depending on the presence or absence of a reference object, while suppressing the processing load.

[0051] In addition, in the disclosed cloud shape cloud layer measurement device 10, the image position identification unit 25 selects a position relationship identification method in the following order of priority: identification of position relationship based on objects commonly included in the cloud shape map 32 and the relative cloud height map 33, identification of position relationship based on the area occupied by the cloud, identification of position relationship by combining clouds with different relative cloud heights, and identification of position relationship by excluding a specified type of cloud shape. By setting priorities in this way, it is possible to select an identification method that is appropriate for the situation.

[0052] The present invention is not limited to the above-mentioned embodiment, and various modifications are included. For example, the above-mentioned embodiment is described in detail to easily explain the present invention, and is not necessarily limited to the embodiment having all the described configurations. Moreover, the present invention is not limited to the deletion of the configurations, and it is also possible to replace or add the configurations. [Explanation of symbols]

[0053] 10: cloud shape and cloud layer measurement device, 11: visible light camera, 12: infrared camera, 13: display unit, 14: input unit, 15: CPU, 16: main memory device, 17: auxiliary memory device, 21: visible light image receiving unit, 22: infrared image receiving unit, 23: cloud shape estimation unit, 24: cloud base temperature calculation unit, 25: image position identification unit, 26: cloud layer estimation unit, 31: cloud shape recognition model, 32: cloud shape map, 32a: cloud shape map, 32b: cloud shape map, 33: relative cloud height map, 33: cloud shape map, 33a: relative cloud height map, 33a: cloud shape map, 33b: relative cloud height map, 34: cloud shape and cloud layer information

Claims

1. An in-image position identification unit that compares a cloud shape map showing the position and shape of clouds in an image with a relative cloud height map showing the position and relative cloud height of clouds in an image, and identifies the positional relationship between the cloud shape map and the relative cloud height map based on the result of the comparison; A cloud layer estimation unit that separates the cloud shapes shown in the cloud shape map by the relative cloud height using the positional relationship and outputs cloud shape and cloud layer information in which the cloud shapes and the relative cloud heights are associated with each other; A cloud layer measuring device comprising:

2. The cloud layer measuring device according to claim 1, A cloud shape estimation unit that estimates the cloud shape of a cloud in a visible light image using a cloud shape recognition model created in advance and outputs the cloud shape map; a cloud base temperature calculation unit that measures a cloud base temperature from an output of an infrared image captured by an infrared camera, calculates a relative cloud height, and outputs the relative cloud height map; The cloud layer measuring device further comprises:

3. The cloud layer measuring device according to claim 1, The cloud shape layer measurement device is characterized in that the in-image position identification unit identifies the positional relationship by comparing the cloud boundary lines in the cloud shape map with the cloud boundary lines in the relative cloud height map.

4. The cloud layer measuring device according to claim 1, The cloud shape layer measurement device is characterized in that the in-image position identification unit combines clouds of different cloud heights shown in the relative cloud height map and compares them with the cloud shape map.

5. The cloud layer measuring device according to claim 1, The cloud shape and cloud layer measurement device is characterized in that the image position identification unit excludes a predetermined type of cloud shape from the cloud shapes shown in the cloud shape map and compares it with the relative cloud height map.

6. The cloud layer measuring device according to claim 1, The cloud shape layer measurement device is characterized in that the in-image position identification unit identifies the positional relationship by using an object that is commonly included in the cloud shape map and the relative cloud height map as a reference object.

7. The cloud layer measuring device according to claim 1, The cloud shape layer measurement device is characterized in that the in-image position identification unit identifies the positional relationship by comparing a rectangle circumscribing the cloud in the cloud shape map with a rectangle circumscribing the cloud in the relative cloud height map.

8. The cloud layer measuring device according to claim 1, The cloud shape cloud layer measurement device is characterized in that the in-image position identification unit selects a positional relationship identification method in the following order of priority: identification of positional relationships based on objects commonly included in the cloud shape map and the relative cloud height map, identification of positional relationships based on the area occupied by the clouds, identification of positional relationships by combining clouds with different relative cloud heights, and identification of positional relationships by excluding a specified type of cloud shape.

9. The device, An in-image position identification step of comparing a cloud shape map showing the position and shape of clouds in the image with a relative cloud height map showing the position and relative cloud height of clouds in the image, and identifying the positional relationship between the cloud shape map and the relative cloud height map based on the result of the comparison; A cloud layer estimation step of separating the cloud shapes shown in the cloud shape map by the relative cloud height using the positional relationship and outputting cloud shape and cloud layer information in which the cloud shapes and the relative cloud heights are associated with each other; A method for measuring a cloud layer comprising:

Images