Building condition recording system

Abstract

When a field worker or a mobile device configured to move within a building patrols the building, taking photographs of the building's interior with a recording device and recording the images along with their location, the system aims to improve the accuracy of location while simplifying the configuration and facilitating operation. [Solution] The building condition recording system 1, which records the state inside a building, comprises: a shooting and recording device 2 configured to be held or attached to a field worker or mobile body, which photographs the inside of the building and acquires captured images; a movement information acquisition means 22 provided on the shooting and recording device 2 for acquiring movement information of a field worker or mobile body associated with the captured images; a shooting position estimation unit 32 which estimates the shooting position where each captured image was taken based on the movement information associated with the captured image; a detection unit 33 which detects a location-identifiable object or identification tag installed inside the building that can identify its installation location from the captured image; and a shooting position correction unit 34 which corrects the shooting position based on the installation location of the detected location-identifiable object or identification tag.

Description

Technical Field

[0003] , , ,

[0001] The present invention relates to a building state recording system for recording the state inside a building.

Background Art

[0002] When constructing a building, it is essential to grasp the state of the construction site in order to manage quality, processes, etc. Also, even after the building is completed and put into operation, it is important to grasp the state of the building from the perspectives of building facility management and facility management. When grasping the state of a building as described above, information may be shared among relevant parties using photos and videos. Here, in order to appropriately utilize the taken photos and videos, the position where the photos and videos were taken needs to be accurately provided as information with respect to the photos and videos, so as to clarify which part inside the building the scene of the photos and videos was taken.

[0003] Regarding this, for example, Patent Document 1 discloses a configuration of a position management system including a position acquisition unit that acquires the positions of people and construction equipment at a construction site, and a display control unit that displays the positions of people and construction equipment acquired by the position acquisition unit. In this configuration, the position acquisition unit includes a transmitter attached to a person or construction equipment and transmitting radio waves, and a receiver attached to the building and receiving radio waves transmitted from the transmitter. Also, Patent Document 2 discloses a configuration of a site management system that installs a plurality of receiving antennas indoors to receive radio waves transmitted from wireless tags attached to on-site workers and on-site construction equipment, and analyzes the radio waves from the wireless tags received by the receiving antennas to detect the positions of the wireless tags. Applying the configurations described in Patent Documents 1 and 2 above, for example, it is conceivable to equip on-site workers with transmitters that emit radio waves, install receivers in the building to receive the radio waves, and analyze the received radio waves to identify the location of the image and add it to the photograph or video. In this case, the accuracy of location identification may be higher. However, in this case, receivers would need to be installed inside the building to cover all the spaces within the building. In particular, when targeting buildings under construction, there is a high possibility that the receivers will be damaged or lost due to construction. Therefore, operation may not be easy.

[0004] In contrast, Patent Document 3 discloses a configuration for a location management system that detects the location of a construction worker within a construction site, comprising a pedestrian autonomous navigation means carried by the construction worker to acquire information on the construction worker's movement, a photographic recording device for photographing the construction site, and identification tags placed at the construction site. In this configuration, the location of the construction worker is detected based on equipment information and / or identification information from the identification tag obtained using the photographic recording device, and the construction worker's movement information obtained from the pedestrian autonomous navigation means. A smartphone, tablet terminal, etc., can be used as the pedestrian autonomous navigation means, a digital camera, digital video camera, or omnidirectional camera (360-degree camera) can be used as the photographic recording device, and a two-dimensional barcode, AR marker, etc. can be used as the identification tag.

[0005] In the configuration of Patent Document 3, there is no need to install receivers inside the building to cover all the spaces within the building, making it easier to operate compared to Patent Documents 1 and 2. However, if, for example, a smartphone is used as a means of pedestrian autonomous navigation and an omnidirectional camera is used as a recording device, then when a site worker patrols the construction site, both of these must be carried and operated individually by the site worker. In this respect, the configuration of Patent Document 3 is considered to have the potential to make operation even easier. When field workers or mobile units configured to move within a building patrol the building, taking photographs of the building's interior with a recording device and recording the images along with their locations, there is a need to improve the accuracy of the location while implementing a simpler configuration and making operation easier. [Prior art documents] [Patent Documents]

[0006] [Patent Document 1] Japanese Patent Publication No. 2020-16466 [Patent Document 2] Japanese Patent Publication No. 2019-138785 [Patent Document 3] Patent No. 7562398 [Overview of the Initiative] [Problems that the invention aims to solve]

[0007] The problem that this invention aims to solve is to provide a building condition recording system that allows on-site workers or a mobile device configured to move within a building to patrol the building, take photographs of the building's interior with a photographic recording device, and record the captured images along with their location, while improving the accuracy of the location, and enabling a simpler configuration that facilitates operation. [Means for solving the problem]

[0008] The inventors of this invention focused on the fact that in a building condition recording system, by applying image recognition AI to captured images and recognizing specific, immovable objects or shapes, such as location-identifiable objects or identification tags, it is possible to automatically acquire accurate location information of on-site workers and moving objects, which led to the present invention. To solve the above problems, the present invention employs the following means. That is, the present invention provides a building condition recording system for recording the state inside a building, comprising: a shooting recording device configured to be held or attached to a field worker or a mobile body configured to be movable inside the building, and which photographs the inside of the building and acquires captured images; a movement information acquisition means provided inside the shooting recording device for acquiring movement information of the field worker or the mobile body associated with the captured images; a shooting position estimation unit for estimating the shooting position where each of the captured images was taken, based on the movement information associated with the captured image; a detection unit for detecting a location-identifiable object or identification tag installed inside the building that can identify its installation location, from the captured images; and a shooting position correction unit for correcting the shooting position based on the installation location of the detected location-identifiable object or identification tag. With the above configuration, when a field worker or a mobile device configured to move within the building patrols the building, the interior of the building is photographed at any time by a photographic recording device configured to be held or attached to the field worker or mobile device, and the captured images are acquired. At the same time, movement information of the field worker or mobile device is acquired at any time by a movement information acquisition means provided inside the photographic recording device. This movement information is associated with the captured images, and the shooting position estimation unit estimates the shooting position where each captured image was taken based on the movement information associated with that image. Therefore, by referring to these shooting positions and captured images, it is possible to understand what kind of images were taken at each shooting position within the building during the patrol. In this way, the state of the building can be recorded. Here, the detection unit detects a location-identifiable object or identification tag from the image captured by the image recording device. Since the location-identifiable object or identification tag has a specific installation location within the building, if a location-identifiable object or identification tag is detected from the image, the shooting location where the image was taken can be accurately estimated from the installation location of the detected location-identifiable object or identification tag. Therefore, even if the accuracy of the shooting location estimated by the shooting location estimation unit based on movement information is not high, the accuracy of the shooting location can be improved by correcting the shooting location based on the installation location of the location-identifiable object or identification tag detected as described above. In this way, the accuracy of the shooting location where the image was taken can be further improved. Furthermore, since the movement information acquisition mechanism is located inside the recording device, for example, when a site worker patrols the construction site, there is no need for the site worker to individually carry and operate the movement information acquisition mechanism and the recording device. In this way, the building condition recording system has a simple configuration and can be easily operated. In this way, when a field worker or a mobile device configured to move within the building patrols the building, takes pictures of the building's interior with a recording device, and records the captured images along with their location, it becomes possible to improve the accuracy of the location while implementing a simpler configuration and making operation easier.

[0009] In one embodiment of the present invention, at least one of the locatable objects is provided on each floor of the building. With the configuration described above, the exact location where the captured image was taken can be accurately determined at each level.

[0010] In another embodiment of the present invention, the motion information acquisition means is an acceleration sensor and an angular velocity sensor, the motion information is acceleration data and angular velocity data, and the motion information is either embedded in the captured image or, if the motion information and the captured image are separated, the system includes an information integration unit that associates the motion information and the captured image over time. With the configuration described above, movement information can be accurately associated with images captured by the image recording device. [Effects of the Invention]

[0011] According to the present invention, when a field worker or a mobile device configured to move within a building patrols the building, takes photographs of the building's interior with a photographic recording device, and records the captured images along with their location, it is possible to provide a building condition recording system that can be implemented with a simpler configuration while improving the accuracy of the location, and that is easy to operate. [Brief explanation of the drawing]

[0012] [Figure 1] This is a block diagram of a building condition recording system and a construction progress management system using the building condition recording system, according to a first embodiment of the present invention. [Figure 2] This figure shows an example of a construction site related to a building whose condition is recorded by a building condition recording system. [Figure 3] This figure shows an example of a field worker carrying a recording device attached to a support rod. [Figure 4] This is an explanatory diagram regarding the relative shooting position estimated by the shooting position estimation unit and the correction of the shooting position. [Figure 5] This is an explanatory diagram showing an example of display on a display device in a building condition recording system. [Figure 6] This figure shows the estimated location and extent of wall construction by the construction scope estimation unit. [Figure 7] This figure shows the estimated location and scope of ceiling work by the construction scope estimation unit. [Figure 8] This flowchart shows the flow of the building condition recording method performed using the building condition recording system of the first embodiment described above. [Figure 9] This is a flowchart showing the flow of the construction progress management method in the construction progress management system of the first embodiment described above. [Figure 10]Block diagram of a building condition recording system and a construction progress management system using the building condition recording system according to the first modification of the first embodiment. [Figure 11] Block diagram of a building condition recording system and an equipment / material management system using the building condition recording system according to the second modification of the first embodiment. [Figure 12] It is a figure which shows an example of the photographed image which detected the on-site equipment / material with the detection part.

Modes for Carrying Out the Invention

[0013] The present invention is a building condition recording system that estimates the on-site situation, construction progress situation, or location of construction equipment / materials from photographed images for the inside of a structure at a construction site. Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. (First Embodiment) FIG. 1 is a block diagram of a building condition recording system and a construction progress management system using the building condition recording system in the first embodiment of the present invention. FIG. 2 is a diagram showing an example of a construction site related to a building whose condition is recorded by the building condition recording system. As shown in FIG. 1, the building condition recording system 1 includes a photographing and recording device 2. When the building condition recording system 1 moves inside the building while the photographing and recording device 2 is carried by the on-site worker Q or a moving body configured to be movable inside the building, the photographing and recording device 2 photographs the inside of the building to obtain a photographed image, thereby recording the condition inside the building. The photographing and recording device 2 includes a movement information acquisition means 22 inside it. The movement information acquisition means 22 acquires movement information of the on-site worker Q or the moving body associated with the photographed image. Thereby, when the on-site worker Q or the moving body moves inside the building, the photographed image and the movement information are associated and stored. Based on such a photographed image and movement information, it is possible to know which location inside the building the photographed image was taken of, so by viewing the saved record, information regarding the inside of the building can be confirmed or shared among interested parties.

[0014] Such a building condition recording system 1 can be used for facility management and operational purposes of a building after its completion. Furthermore, the building condition recording system 1 can also be used to understand the condition of the construction site G during building construction and to manage quality, progress, etc. This embodiment will primarily describe the latter case, but the former case can also be described similarly. This embodiment will also describe a case where the construction progress management system 40, which will be described later, uses the records stored in the building condition recording system 1 to manage the progress of construction work, including wall and ceiling work, within the building construction site G. In this embodiment, construction site G is a site where construction work is carried out on a building having multiple floors, such as a high-rise building.

[0015] The recording device 2 is carried by the site worker Q and takes pictures inside the building to acquire images. As described above, the recording device 2 takes pictures of the construction site G (building) to acquire images. In particular, in this embodiment, the recording device 2 is a so-called omnidirectional camera (360-degree camera) that can capture images in all directions of 360 degrees. The recording device 2 may be a portable digital video camera or the like. Figure 3 shows an example of a field worker carrying a recording device attached to a support rod. The recording device 2 is configured to be held or worn by the field worker Q. Especially when an omnidirectional camera is used as the recording device 2, it is desirable that the recording device 2 be positioned above the field worker Q in order to avoid as much as possible the field worker Q appearing in the image and obstructing the recording. Therefore, as shown in Figure 3, it is desirable that the recording device 2 be fixed to the tip 100t of the support rod 100, and that the field worker Q carry it by gripping the lower end 100a of the support rod so that the recording device 2 is positioned above the field worker Q's head. Alternatively, the recording device 2 may be configured to be fixed or worn on the field worker Q, for example, on their helmet. The recording device 2 may be configured to be held or attached to a mobile body such as a drone or robot that is configured to move along a predetermined route within the building, instead of being carried by the on-site worker Q. Hereafter, even if the explanation is given to the effect that the recording device 2 is carried by the on-site worker Q, that explanation shall also include the case in which the recording device 2 is carried by a mobile body.

[0016] The recording device 2 comprises a recording means 21, a movement information acquisition means 22, a storage unit 23, and a transmission unit 24. The imaging means 21 is equipped with an image sensor using a CCD (Charge Coupled Device), CMOS (Complementary Metal Oxide Semiconductor), etc. The shooting means 21 generates multiple images by continuously taking pictures of the construction site G at predetermined time intervals, for example, when a site worker Q is patrolling the building construction site G. The shooting means 21 may also generate multiple images by taking still images at predetermined short intervals, such as 0.5 seconds. Alternatively, the shooting means 21 may generate multiple images by shooting a video. In this case, the shooting means 21 can be considered to generate multiple images taken at predetermined intervals by treating each frame taken at predetermined intervals, such as 1 second to several seconds, within the video as a single still image. When the shooting means 21 generates multiple images by shooting a video, the process of actually extracting and generating still images from the video may be performed by the shooting recording device 2 or by the system main body 3, which will be described later. In the recording device 2, recording by the recording means 21 can be started and stopped at any arbitrary timing. Recording by the recording means 21 may be started and stopped at the same time that on-site worker Q starts and finishes patrolling the construction site G.

[0017] In this embodiment, the motion information acquisition means 22 consists of an acceleration sensor and an angular velocity sensor. The motion information acquisition means 22 is a so-called 6-axis sensor. The motion information acquisition means 22 acquires acceleration data in each of the three directions: the first horizontal direction, the second horizontal direction, and the vertical direction, which are mutually orthogonal in the horizontal plane, using the acceleration sensor. The motion information acquisition means 22 also acquires angular velocity data in each of the first horizontal direction, the second horizontal direction, and the vertical direction using the angular velocity sensor. The movement information acquisition means 22 is located inside the housing of the recording device 2 and acquires acceleration data and angular velocity data. When the recording device 2 is carried by the worker Q and the worker Q patrols the construction site G, the movement information acquisition means 22 detects changes in speed and angle caused by the worker Q's movement at predetermined short intervals after the start of positioning, acquiring acceleration data (the difference in speed) and angular velocity data (the difference in angle). Based on the acceleration data, the amount of movement of the worker Q (from the time the previous acceleration data was acquired) can be determined. Similarly, based on the angular velocity data, the amount of displacement in the direction the worker Q is facing (from the time the previous angular velocity data was acquired) can be determined. Therefore, by accumulating the acceleration data and angular velocity data over time, the movement path of the worker Q after the start of positioning can be determined. From this perspective, the acceleration data and angular velocity data can be considered movement information, relating to the path the worker Q took within the construction site G.

[0018] When the shooting means 21, for example, shoots a video, depending on the configuration of the shooting recording device 2, the above-mentioned movement information can be embedded in the video. That is, in this case, the movement information can be considered to be embedded in each captured image, such that movement information is provided for each captured image. In this way, in this case, the captured image and the movement information are associated with each other at the time when the captured image is acquired by the shooting means 21 and the movement information is acquired by the movement information acquisition means 22. Unlike the above, the processing in cases where the captured image and the movement information are not directly related to each other at the time the captured image is acquired by the shooting means 21 and the movement information is acquired by the movement information acquisition means 22 will be described later as the first modified example of this embodiment.

[0019] The captured images obtained by the imaging means 21 are stored in the storage unit 23. Similarly, the movement information obtained by the movement information acquisition means 22 is also stored in the storage unit 23, associated with the captured images.

[0020] The transmitting unit 24 transmits the associated captured images and movement information stored in the storage unit 23 to the system main unit 3. In this embodiment, the captured images and movement information are transmitted to the system main unit 3 all at once after the field worker Q has finished his patrol. In this case, the transmitting unit 24 can transmit the captured images and movement information using methods such as data transfer via a connecting cable or various portable memory devices. Alternatively, the recording device 2 may be configured such that when captured images and movement information are acquired, the transmission unit 24 transmits them sequentially and in real time to the system main unit 3. In this case, the transmission unit 24 transmits the captured images and movement information to the system main unit 3 via data transfer, for example, through a wireless LAN (Local Area Network) such as Wi-Fi or Bluetooth®, or a mobile phone communication network.

[0021] The main system unit 3 consists of computer terminals such as a server and a personal computer, and performs the required functions by executing a pre-configured program. Functionally, the main system unit 3 is equipped with a data input receiving unit 31, a shooting position estimation unit 32, a detection unit 33, a shooting position correction unit 34, a database 35, and an information display unit 36. The data input receiving unit 31 acquires the associated captured images and movement information transmitted from the transmission unit 24 of the shooting recording device 2 and stores them in the database 35.

[0022] Database 35 stores the associated captured images and movement information, as well as the data used and generated in the processes described later. Database 35 stores the design drawing data for construction site G. The drawing data for each floor within construction site G includes CAD (Computer-Aided Design) data and BIM (Building Information Modeling) data for the walls, ceilings, and other parts of each floor.

[0023] The shooting position estimation unit 32 estimates the shooting position of each image within the construction site G based on the movement information associated with that image. Specifically, it uses acceleration data and angular velocity data recorded in the shooting recording device 2 to calculate the amount and direction of movement using pedestrian autonomous navigation (PDR). For example, consider a scenario where, at position P0 in Figure 2, associated images and movement information are acquired, and then, at position P1, associated images and movement information are acquired. In this case, the movement information acquired at position P1 includes acceleration data and angular velocity data as a history of the movement of the field worker Q from position P0 to position P1. Using this acceleration data and angular velocity data, it is possible to determine in which direction and distance the field worker Q actually moved from position P0 to position P1. Therefore, by applying the direction and distance calculated as described above to position P0, position P1 at the construction site G can be calculated as position Pr1 relative to position P0. Figure 2 shows the calculated relative shooting positions Pr1 to Pr4, starting from shooting position P0, obtained by repeating the calculations described above. By performing this process, it becomes possible to determine at which shooting position P1 to P4 (more precisely, at which relative shooting position Pr1 to Pr4) within the construction site G the captured image was taken by the image recording device 2.

[0024] However, as described above, the coordinates of shooting position Pr1, calculated by applying the direction and distance traveled by the site worker Q to the coordinates of shooting position P0 in the coordinate system of construction site G, are relative values ​​to the coordinates of shooting position P0. Therefore, if there is an error in the calculation of the direction and distance traveled by the site worker Q, this error will accumulate as the relative position calculation is repeated, and the error between the shooting positions Pr1 to Pr4 calculated as relative positions and the coordinates of the actual shooting positions P1 to P4 may become large. Figure 4 is an explanatory diagram regarding the relative shooting position estimated by the shooting position estimation unit and the correction of the shooting position. In Figure 4, the relative shooting position Pr5 estimated by the shooting position estimation unit 32 is different from the correct position P5 where the image was actually taken due to an error. Similarly, when estimating the relative shooting positions where the images were taken, as relative shooting positions Pr6 to Pr9, the error between the relative shooting positions Pr6 to Pr9 and the positions P6 to P9 where the images were actually taken gradually increases.

[0025] The shooting position correction unit 34, which will be described later, corrects the relative shooting position Pr, which may contain errors and was estimated by the shooting position estimation unit 32, so that it becomes a value close to the actual shooting position where the captured image was taken. For this correction, the detection unit 33 detects a location-identifiable object M installed inside the building from the captured image. A locatable object M is an object whose location on each floor of the building construction site G can be uniquely identified based on its external shape. In other words, each locatable object M has a unique shape, and there are no more than two objects with the same external shape on each floor. A locatable object M is not installed in the building solely for the purpose of performing each process in the building state recording system 1 of this embodiment, particularly the process related to correcting the shooting position in the shooting position correction unit 34 described later. Rather, it is a component of the building, provided for use in the building even when the building state recording system 1 is not applied to the building. Examples of such locatable objects M include, for example, electrical distribution boards that include installation location information, equipment such as elevators and stairs that include floor numbering, and installations and decorations such as signs and fixtures. Examples of such installations and decorations include signs that indicate the floor number and location within the floor, such as "15F N Zone," and grid line numbers such as "X1-Y5" printed on pillars.

[0026] In Figure 4, the locatable object M1 is an elevator that includes a floor indicator showing the current status of the elevator's stopping position, of which there is only one per floor. In Figure 4, the locatable object M1 is positioned to face downwards, and when an image is taken from a direction opposite to the locatable object M1, that is, from a shooting position below the locatable object M1 in Figure 4, including the area above, the locatable object M1 is captured in the image. Furthermore, signage installed on the pillars is used as location-identifiable objects M2, M3, M4, and M5. Location-identifiable objects M2 and M5 are positioned to face left in Figure 4, and when an image is taken from a direction opposite to location-identifiable objects M2 and M5, that is, from a shooting position to the left of location-identifiable objects M2 and M5 in Figure 4, including the right side, location-identifiable objects M2 and M5 are captured in the captured image. Location-identifiable object M3 is positioned to face right in Figure 4, and when an image is taken from a direction opposite to location-identifiable object M3, that is, from a shooting position to the right of location-identifiable object M3 in Figure 4, including the left side, location-identifiable object M3 is captured in the captured image. In Figure 4, the locatable object M4 is positioned to face downwards, and when an image is captured from a direction opposite to the locatable object M4, that is, from a shooting position below the locatable object M4 in Figure 4, including the area above, the locatable object M4 is captured in the image. Such locatable objects M are provided at least once on each level of the construction site G.

[0027] Information regarding the locatable objects M described above is stored in database 35. Specifically, the database 35 stores information about each locatable object M, such as its identification number, its appearance when photographed from the front, its horizontal and vertical position (level) within the construction site G, and the orientation (azimuth or angle) in which it is installed. In particular, database 35 stores the absolute coordinates on the coordinate system represented in the design drawing data as the horizontal position of the locatable object M within the construction site G.

[0028] The detection unit 33 can be implemented to detect a localizable object M using a trained model that has been deep-trained using training input images of localizable objects M and the identification numbers of localizable objects M captured in those training input images as training data. In this case, the trained model is preferably composed of a convolutional neural network (CNN). The trained model is deep-trained to identify a localizable object M when an image is input, by outputting an identification number corresponding to that localizable object M if such an object M is captured in the image. Thus, the trained model is a program module that is part of artificial intelligence software and has learned appropriate training parameters. The detection unit 33 executes this trained model as a program on a CPU or GPU, for example, to identify a localizable object M when a captured image is input.

[0029] The detection unit 33 inputs each captured image into a trained model configured as described above, and when a location-identifiable object M is captured, it outputs the identification number of the location-identifiable object M, thereby estimating the location-identifiable object M captured in the captured image. If the detection unit 33 detects a location-identifiable object M from the captured image, it stores the identification number of the detected location-identifiable object M in the database 35, associating it with the captured image in which the location-identifiable object M was detected.

[0030] The detection unit 33 further identifies the range in the captured image in which the localizable object M is captured, that is, the outer shape of the localizable object M within the captured image. This identification of the range in which the localizable object M is captured may be achieved by a known image processing algorithm, or by using a machine learning model configured with semantic segmentation or the like. The detection unit 33, for example, obtains appearance information from the database 35 that is associated with the identification number of the locatable object M detected from the captured image, and compares this appearance information with the outer shape of the locatable object M obtained from the captured image. If the locatable object M is photographed from an angle different from the front, the locatable object M should appear distorted in the photograph, unlike when it is photographed from the front. Therefore, the detection unit 33 calculates the angle from which the image was taken relative to the front of the locatable object M by comparing, for example, the outer shape of the locatable object M obtained from the shadow image with the appearance information obtained from the database 35 that shows the locatable object M when it is photographed from the front. Furthermore, the detection unit 33 calculates the size (for example, the area within the image) of the locatable object M captured in the image. Based on the size of the locatable object M within the image, the detection unit 33 calculates the distance from the shooting location where the image was taken to the locatable object M.

[0031] When a locatable object M is detected, the detection unit 33 stores in the database 35, in association with the angle at which the image was taken from the front of the locatable object M, and the distance from the shooting position where the image was taken, both calculated as described above, for the image in which the locatable object M was detected.

[0032] The shooting position correction unit 34 corrects the shooting position based on the installation position of the detected locatable object M. More specifically, the shooting position correction unit 34 corrects the relative shooting position estimated by the shooting position estimation unit 32 based on the installation position of the locatable object M. When a location-identifiable object M is detected in the captured image, the shooting position correction unit 34 obtains the angle at which the image was taken from the front direction of the location-identifiable object M, and the distance from the shooting position where the image was taken, which are stored in the database 35 in association with the captured image. The shooting position correction unit 34 then obtains the horizontal position of the location-identifiable object M within the construction site G, i.e., its absolute position in the construction site G, and the orientation in which the location-identifiable object M is located, which are stored in the database 35 in association with the captured image. Based on this absolute position and orientation, the distance from the location-identifiable object M, and the angle from the front direction of the location-identifiable object M, the shooting position of the image in which the location-identifiable object M was photographed is calculated as an absolute coordinate value in the construction site G.

[0033] In this way, the shooting position correction unit 34 calculates the absolute shooting position of the captured image in which the identifiable object M is photographed, and determines this absolute shooting position as the shooting position in which the image was taken. The shooting position correction unit 34 then adjusts the shooting position of each image based on the absolute shooting position of the image in which the identifiable object M is captured. More specifically, the shooting position correction unit 34 corrects the relative shooting position estimated by the shooting position estimation unit 32 for images in which the identifiable object M is not captured, based on the absolute shooting position of the image in which the identifiable object M is captured.

[0034] In Figures 2, 4, and Figure 5 (which will be used in a later explanation), the location of an identifiable object M is indicated by a black circle in the captured image, and by a white circle in the captured image where an identifiable object M is not captured. For example, consider the case shown in Figure 4, where a site worker Q begins inspecting a certain floor immediately after exiting the elevator on that floor. In this case, the first image taken on that floor includes the elevator with the floor number displayed. Therefore, the detection unit 33 detects the elevator as a location-identifiable object M1, and the shooting location where the image was taken is calculated as the absolute shooting location P0 at the construction site G.

[0035] For the four images taken following the image corresponding to shooting position P0, the shooting position estimation unit 32 calculates these shooting positions as Pr1 to Pr4 relative to the shooting position P0 of the last image in which the locatable object M1 was captured. Here, the images corresponding to shooting positions Pr1 to Pr3 do not contain any location-identifiable object M, while the image corresponding to shooting position Pr4 contains a sign on a pillar, which is captured as location-identifiable object M2. Therefore, for the image corresponding to shooting position Pr4, the relative shooting position Pr4 is calculated as described above, and the shooting position correction unit 34 calculates the absolute shooting position P4 where the image was taken based on the installation position of the detected location-identifiable object M2. The shooting position correction unit 34 then determines the absolute shooting position P4 as the final calculated shooting position P4.

[0036] The shooting position correction unit 34 calculates the distance between the relative shooting position Pr4 and the absolute shooting position P4 and compares it with a threshold. In this case, the relative shooting position Pr4 and the absolute shooting position P4 are roughly the same, and the calculated distance is below the threshold. In such a case, it is considered that the error in the relative shooting positions Pr1 to Pr3 estimated by the shooting position estimation unit 32, which are located on the path R between the shooting position P0 of the image in which the location-identifiable object M was detected immediately before, and the relative shooting position Pr4, is small. Therefore, the shooting position correction unit 34 determines that the relative shooting positions Pr1 to Pr3 are the final calculated shooting positions P1 to P3.

[0037] Furthermore, for the five images taken after the image corresponding to shooting position P4, the shooting position estimation unit 32 calculates these shooting positions as Pr5 to Pr9 relative to the shooting position P4 of the last image in which the locatable object M2 was captured. Here, the images corresponding to shooting positions Pr5 to Pr8 do not contain any location-identifiable object M, while the image corresponding to shooting position Pr9 contains a sign on a pillar, which is captured as location-identifiable object M3. Therefore, for the image corresponding to shooting position Pr9, the relative shooting position Pr9 is calculated as described above, and the shooting position correction unit 34 calculates the absolute shooting position P9 where the image was taken based on the installation position of the detected location-identifiable object M3. The shooting position correction unit 34 then determines the absolute shooting position P9 as the final calculated shooting position P9.

[0038] The shooting position correction unit 34 calculates the distance between the relative shooting position Pr9 and the absolute shooting position P9 and compares it with a threshold. In this case, the relative shooting position Pr9 and the absolute shooting position P9 are diverging, and the calculated distance is greater than the threshold. In such cases, it is thought that there is a large error in the relative shooting positions Pr5 to Pr8 estimated by the shooting position estimation unit 32, which are located on the path R between the shooting position P4 of the image in which the location-identifiable object M was detected immediately before, and the relative shooting position Pr9. Therefore, the shooting position correction unit 34 corrects these relative shooting positions Pr5 to Pr8. For example, the shooting position correction unit 34 corrects the relative shooting positions Pr5 to Pr8 by rotating and stretching the path R from shooting position P4 to relative shooting position Pr9 around shooting position P4 so that the relative shooting position Pr9 overlaps with the absolute position P9, and calculates the final calculation result of shooting positions P5 to P8. In this way, the shooting position correction unit 34 corrects the shooting position calculated as a relative position. This process is performed for each layer until the shooting position has been determined for all images taken in that layer.

[0039] Figure 5 is an explanatory diagram showing an example of display on a display device in a building condition recording system. The information display unit 36 ​​outputs the results of the above processing to the display device 4. The display device 4 may be a monitor device provided in the construction progress management system 40, or it may be a smartphone, tablet terminal, etc. When a worker specifies a building floor via an input device (not shown), such as a mouse or keyboard, the information display unit 36 ​​displays, for example, the drawing data of the construction site G on the specified floor on the display device 4. The information display unit 36 ​​also displays, superimposed on the drawing data, all the shooting positions P calculated for that floor, and lines indicating the movement path R that connect these shooting positions P in order. In this state, when, for example, an arbitrary shooting position P is selected by the worker, the information display unit 36 ​​displays the image associated with that shooting position P, which is stored in the database 35, on the display device 4.

[0040] The construction progress management system 40 uses the building condition recording system 1 to manage the progress of wall and ceiling construction within the construction site. The construction progress management system 40 estimates the progress of construction within the construction site G based on the movement information of site workers Q obtained by the photography recording device 2 and the captured images. As shown in Figure 1, the construction progress management system 40 functionally includes a construction detection unit 41, a construction range estimation unit 42, and a construction database 43. In this embodiment, the construction progress management system 40 is constructed such that the construction detection unit 41, the construction range estimation unit 42, and the construction database 43 are provided within the system body 3 of the building condition recording system 1.

[0041] The construction database 43 stores data related to wall and ceiling construction work performed within construction site G. This data includes, for example, drawing data for each floor within construction site G. Specifically, the drawing data for each floor within construction site G includes CAD data and BIM data for the walls and ceilings of each floor. More precisely, the data related to wall construction includes information on wall lines indicating the position of the wall surface, the types of materials used to make up each wall (structural materials, underlayment, insulation, finishing materials, etc.), area, and thickness. Similarly, the data related to ceiling construction includes, for example, information on the type of ceiling panel and the installation locations of equipment such as ducts and lighting.

[0042] The construction detection unit 41 detects construction work, including wall construction and ceiling construction, from the captured images at each level within the construction site G. The construction detection unit 41 determines and detects whether wall construction or ceiling construction is captured in each image captured by the image recording device 2. If wall construction or ceiling construction is captured in the image, the construction detection unit 41 estimates the progress of the construction. The construction detection unit 41 estimates which of several types of construction is being carried out as the progress. To this end, the construction detection unit 41 has trained models 411 and 412 that have been deep-trained using construction images of a construction site G and the types of construction corresponding to those images as training data (supervising data). The trained models 411 and 412 are deep-trained to estimate the type of construction of wall construction and ceiling construction when wall construction and ceiling construction are captured in the captured images. The trained models 411 and 412 are composed of, for example, convolutional neural networks. The trained models 411 and 412 are programmed modules that are part of artificial intelligence software and have been trained with appropriate learning parameters. The construction detection unit 41 executes these trained models 411 and 412 as programs on, for example, a CPU or GPU, so that when a captured image is input, it estimates the type of construction of the wall construction and ceiling construction captured in the captured image. In this embodiment, the trained models 411 and 412 each include a trained model 411 for wall construction and a trained model 412 for ceiling construction.

[0043] The pre-trained model 411 for wall construction uses deep learning to detect wall construction work being performed in areas where wall construction is being captured in images, and to estimate the type of construction work being performed. Specifically, the pre-trained model 411 for wall construction is trained to appropriately detect which of several types of construction work is being performed in the wall construction area captured in the images, including wall framing work using LGS (Light gauge steel), insulation work, spray painting, board work, puttying work, and finishing work. The construction detection unit 41 inputs the captured image into a pre-trained model 411 for wall construction and estimates the type of construction work for the wall construction in the captured image. The construction detection unit 41 creates a wall construction type estimation record, which is a record of estimated wall construction, based on the output of the trained model 411 for wall construction. For the wall construction type estimation record, for captured images in which some kind of wall construction is detected by the trained model 411 for wall construction, the shooting location associated with the captured image in the database 35 and the identification information of the type of construction of the detected wall construction are associated with each other and stored in the construction database 43.

[0044] The pre-trained model 412 for ceiling work is trained using deep learning to detect ceiling work being performed in areas of the ceiling that are captured in the image, and to estimate the type of work being performed. Specifically, the pre-trained model 412 for ceiling work is trained to appropriately detect which of several types of work is being performed in the areas of the ceiling captured in the image, including ceiling substructure work, board work, equipment installation, fireproofing work, putty work, finishing work, and unfinished work. The construction detection unit 41 inputs the captured image into a pre-trained model 412 for ceiling construction and estimates the type of construction work for the ceiling in the captured image. The construction detection unit 41 creates a ceiling construction type estimation record, which is a record of estimated ceiling construction, based on the output of the trained model 412 for ceiling construction. For the ceiling construction type estimation record, the shooting location associated with the shooting image in the database 35 and the identification information of the type of construction of the detected ceiling construction are associated with each other and stored in the construction database 43 for each captured image in which ceiling construction has been detected by the trained model 412 for ceiling construction.

[0045] The construction area estimation unit 42 estimates the location and extent of the construction work based on the shooting position associated with the captured images of wall and ceiling construction. The construction area estimation unit 42 estimates the location and extent of the wall and ceiling construction work within the construction site G based on the shooting position of the captured images in which the areas of wall and ceiling construction are being carried out. Figure 6 shows the estimated location and scope of wall construction by the construction scope estimation unit. In Figure 6, the location and scope Wm of wall construction are displayed on the building's construction drawing Z, for example, by color coding. Furthermore, the type of wall construction is displayed, for example, by color coding, within the location and scope Wm of wall construction. The construction area estimation unit 42 stores the estimated information indicating the location and extent of the wall construction in the construction database 43, associating it with the captured images. Figure 7 shows the estimated location and scope of ceiling work by the construction scope estimation unit. In Figure 7, the location Cm where ceiling work is being carried out is indicated on the building's construction drawing Z by, for example, color coding. Furthermore, the type of ceiling work is indicated by, for example, color coding within the location Cm where ceiling work is being carried out. The construction area estimation unit 42 stores the estimated information indicating the location where ceiling work is being carried out in the construction database 43, associating it with the captured images.

[0046] The information display unit 36 ​​refers to the construction database 43 and, for each floor, overlays the progress status (type of work) of wall and ceiling work onto the location and range of wall and ceiling work estimated by the construction range estimation unit 42 on the building construction drawing Z corresponding to that floor. This creates display data as shown in Figures 6 and 7, and outputs it to the display device 4 as the estimated result of the construction progress information. For example, when a worker specifies a floor of the building, the information display unit 36 ​​can be configured to acquire information indicating the location and range of work associated with that floor, and based on this, create display data by overlaying the construction drawing of the specified floor with the estimated result of the construction progress information within the construction site G. The display device 4 displays the building's construction drawings and the progress of the construction overlaid on each other, based on the display data transmitted from the information display unit 36.

[0047] Figure 8 is a flowchart showing the flow of the building condition recording method performed using the building condition recording system 1 of this embodiment. In order to implement the building condition recording method in the building condition recording system 1 of this embodiment, information regarding the locatable object M is registered in the database 35 in advance (step S11).

[0048] In step S12, the field worker Q carries the photography and recording device 2 and walks around the construction site G, taking photographs of the construction site G. The photography means 21 continuously takes photographs of the construction site G at predetermined time intervals as the field worker Q walks around the building construction site G, generating multiple images. After positioning is started, the movement information acquisition means 22 detects changes in acceleration and angular velocity that occur as the field worker Q moves at predetermined short intervals, and acquires acceleration data and angular velocity data. The captured images obtained by the imaging means 21 are stored in the storage unit 23. Similarly, the movement information obtained by the movement information acquisition means 22 is also stored in the storage unit 23, associated with the captured images. The transmission unit 24 transmits the associated captured images and movement information stored in the storage unit 23 to the system main unit 3.

[0049] The shooting position estimation unit 32 estimates the shooting position where each image was taken within the construction site G, based on the movement information associated with that image (step S13). The detection unit 33 detects a location-identifiable object M installed inside the building from the captured image (step S14). The shooting position correction unit 34 corrects the shooting position based on the installation position of the detected locatable object M (step S15).

[0050] Figure 9 is a flowchart showing the flow of the construction progress management method in the construction progress management system 40. In the construction progress management system 40, the construction detection unit 41 detects construction work, including wall construction and ceiling construction, from captured images at each level within the construction site G (process S21). Next, the construction area estimation unit 42 estimates the location and extent of the construction work based on the shooting location associated with the captured image. Subsequently, when a worker accesses the construction progress management system 40 and requests the display of information showing the progress of construction work within the construction site G, the information display unit 36 ​​creates display data for information showing the progress of construction work based on various data necessary for displaying the information (process S22). The information display unit 36 ​​creates display data as shown in Figures 6 and 7 by overlaying the progress status (type of work) of wall work and ceiling work onto the location and range of wall work and ceiling work estimated by the construction scope estimation unit 42 on the building construction drawing Z corresponding to the floor for each floor, and outputs the estimated progress information of the work to the display device 4 (step S23).

[0051] The building condition recording system 1 described above is a building condition recording system 1 that records the state inside a building (construction site G), and comprises: a shooting recording device 2 configured to be held or attached to a site worker Q or a mobile body configured to move within the building, and which takes pictures of the inside of the building and acquires captured images; a movement information acquisition means 22 provided inside the shooting recording device 2 for acquiring movement information of the site worker Q or the mobile body associated with the captured images; a shooting position estimation unit 32 that estimates the shooting position (relative shooting position) of each captured image based on the movement information associated with the captured image; a detection unit 33 that detects a location-identifiable object M installed inside the building that can be identified from the captured image; and a shooting position correction unit 34 that corrects the shooting position based on the installation position of the detected location-identifiable object M. With the above configuration, when a field worker Q or a mobile device configured to move within the building patrols the building, the building is photographed at any time by a photography and recording device 2, which is configured to be held or attached to the field worker Q or the mobile device, and the captured images are acquired. At the same time, movement information of the field worker Q or the mobile device is acquired at any time by a movement information acquisition means 22 provided inside the photography and recording device. This movement information is associated with the captured images, and the shooting position estimation unit 32 estimates the shooting position where each captured image was taken based on the movement information associated with that image. Therefore, by referring to these shooting positions and captured images, it is possible to understand what kind of images were taken at each shooting position within the building during the patrol. In this way, the state of the building can be recorded. Here, the detection unit 33 detects a location-identifiable object M from the image captured by the image recording device 2. Since the location-identifiable object M is installed within a building and its installation location can be identified, if a location-identifiable object M is detected from the captured image, the shooting location where the image was taken can be accurately estimated from the installation location of the detected location-identifiable object M. Therefore, even if the accuracy of the shooting location estimated by the shooting location estimation unit 32 based on the movement information is not high, the accuracy of the shooting location can be improved by correcting the shooting location based on the installation location of the location-identifiable object M detected as described above. In this way, it is possible to further improve the accuracy of the shooting location where the image was taken. Furthermore, since the movement information acquisition means 22 is located inside the recording device 2, for example, when a site worker Q patrols the construction site, the site worker Q does not need to carry the movement information acquisition means 22 and the recording device 2 separately and operate them individually. In this way, the configuration of the building condition recording system 1 is simple, making it easy to operate. Furthermore, the building condition recording system 1 automatically recognizes special, immobile objects and shapes from captured images as locatable objects M, thereby enabling the identification of location information by utilizing the existing environment and objects. In this way, when a field worker Q or a mobile device configured to move within the building patrols the building and takes pictures of the building with the shooting and recording device 2, and records the captured images along with their location, it becomes possible to improve the accuracy of the location while implementing a simpler configuration, making it easier to introduce and operate.

[0052] In particular, in the configuration of this embodiment, the accuracy of the shooting location from which the captured image was taken is improved by detecting the location-identifiable object M that constitutes the building. Therefore, in order to improve the accuracy of the shooting location, it is basically not necessary to separately install items for identifying the shooting location, such as identification tags such as 2D barcodes or AR markers, within the building. As a result, the configuration of the building condition recording system 1 is simple and can be easily implemented.

[0053] Furthermore, at least one locatable object M is provided on each floor of the building. With the configuration described above, the exact location where the captured image was taken can be accurately determined at each level.

[0054] Furthermore, the movement information acquisition means 22 consists of an acceleration sensor and an angular velocity sensor, and the movement information consists of acceleration data and angular velocity data, and the movement information is embedded within the captured image. With the configuration described above, movement information can be accurately associated with the images captured by the image recording device 2.

[0055] (First embodiment, first modified example) Next, a first modified example of the first embodiment described above will be explained using Figure 10. Figure 10 is a block diagram of a building condition recording system and a construction progress management system using the building condition recording system, relating to this first modified example. In the first embodiment described above, when the shooting means 21, for example, shoots a video, the movement information acquired by the movement information acquisition means 22 was embedded in the video. Alternatively, if the movement information acquired by the movement information acquisition means 22 is stored in the storage unit 23 separately from the captured image, it is necessary to associate the movement information with the captured image. This modified version includes an information integration unit 37 that associates the movement information with the captured image.

[0056] The captured images taken by the image recording device 2 are associated with the time of capture, which is the time the image was taken. Similarly, the movement information acquired by the movement information acquisition means 22, namely acceleration data and angular velocity data, is associated with the time of acquisition, which is the time the data was acquired. The information integration unit 37 associates the movement information and the captured images based on time. More specifically, the information integration unit 37, for each captured image, compares the capture time associated with the captured image with the acquisition time associated with each piece of movement information, and extracts the acquisition time closest to the capture time among the movement information, thereby extracting the movement information to be associated with the captured image, and associates the extracted movement information with the captured image by time. The information integration unit 37 stores the associated captured images and movement information in the database 35. The shooting position estimation unit 32, the detection unit 33, and the shooting position correction unit 34 use the associated captured images and movement information stored in the database 35 in this manner to perform processing in the same manner as in the first embodiment, as performed by the shooting position estimation unit 32, the detection unit 33, the shooting position correction unit 34, and the information display unit 36.

[0057] Thus, the building condition recording system 1 of this modified example includes an information integration unit 37 that associates movement information and captured images over time when the movement information and captured images are in a separate state. With the above configuration, even when movement information and captured images are separate, the movement information can be accurately associated with the captured images from the image recording device.

[0058] (Second modified example of the first embodiment) Next, a second modified example of the first embodiment will be described using Figure 11. Figure 11 is a block diagram of a building condition recording system and a materials and equipment management system using the building condition recording system, relating to this second modified example. In this modified example, a materials management system 50 that detects and manages the location of various equipment and materials (hereinafter referred to as site materials and equipment) used in carrying out work within the construction site G is implemented using the building condition recording system 1 of the first embodiment described above. The equipment management system 50 comprises the building condition recording system 1 described above and a site equipment location identification unit 51. More specifically, the equipment management system 50 is configured such that the site equipment location identification unit 51 is functionally implemented inside the system body 3 of the building condition recording system 1.

[0059] In the equipment management system 50, the detection unit 33 detects not only location-identifiable objects M but also on-site equipment from the captured image. The detection unit 33 can be implemented to detect on-site equipment using a trained model that has been deep-trained using training input images related to on-site equipment and identification data of on-site equipment captured in those training input images as training data. In this case, the trained model is preferably constructed using, for example, a convolutional neural network. The trained model is deep-trained to identify on-site equipment by outputting identification data corresponding to the on-site equipment if any on-site equipment is captured in the image when an image is input. Thus, the trained model is a program module that is part of artificial intelligence software and has learned appropriate training parameters. The detection unit 33 identifies on-site equipment when a captured image is input by executing this trained model as a program on, for example, a CPU or GPU. In this modified example, the identification data is, for example, the name data of the on-site equipment. The identification data may also be an identification number associated with each piece of on-site equipment.

[0060] Figure 12 shows an example of an image captured when the detection unit detected on-site equipment and materials. The detection unit 33 inputs the captured image N into a trained model configured as described above, for example, and outputs identification data for the on-site equipment 200, thereby estimating the on-site equipment 200 captured in the captured image N. In Figure 12, the detected on-site equipment 200 in the captured image N is shown enclosed in a frame Nx. If the detection unit 33 detects the on-site equipment 200 from the captured image N, it associates the identification data of the detected on-site equipment 200 with the captured image N in which the on-site equipment 200 was detected and records it in the database 35.

[0061] When using the equipment management system 50 described above to find, for example, a specific piece of on-site equipment 200, the worker inputs information about the on-site equipment 200 being sought into the equipment management system 50 via an input device such as a keyboard (not shown). When information about the on-site equipment 200 being sought is entered, the on-site equipment location identification unit 51 refers to the database 35, searches for the captured image N associated with the identification data corresponding to the entered on-site equipment 200, and obtains the shooting location associated with the captured image N. In this way, the site equipment location identification unit 51 estimates the location of the site equipment 200 to be searched within the construction site G. The information display unit 36 ​​highlights and displays, for example, the estimated location of the site equipment 200 on the drawing of the construction site G.

[0062] (Second Embodiment) In the first embodiment described above, the shooting position was corrected based on the installation position of the locatable object M. However, in this second embodiment, identification tags, such as those implemented by a two-dimensional barcode or AR marker, are placed at different locations on the building, and the shooting position is corrected using the identification tags instead of the locatable object M. In other words, the identification tags are installed on the building primarily for the purpose of performing each process in the building state recording system 1 of this embodiment, in particular the process related to correcting the shooting position in the shooting position correction unit 34. If the building state recording system is not applied to the building, the identification tags are not installed on the building. In this case, database 35 stores data related to the identification tag instead of the locatable object M. Database 35 stores the identification information, such as the identification number associated with the identification tag, and the correspondence between the horizontal and vertical position (floor) of the building where the identification tag is installed, and the orientation (direction or angle) in which the identification tag is displayed. The detection unit 33 detects an identification tag in place of a location-identifiable object M from multiple images captured by the image recording device 2 on each floor of the building. When the detection unit 33 detects an identification tag from the captured image, it refers to the database 35 and obtains identification information recorded in association with the identification tag based on the 2D barcode, AR marker, etc., written on the identification tag. Based on the obtained identification information, the detection unit 33 obtains the horizontal position (absolute position) and height position (floor) within the building where the identification tag is installed, as well as the orientation (direction or angle) in which the identification tag is displayed, which are recorded in the database 35 in association with that identification information. As a result, the detection unit 33 identifies the floor on which the captured image in which the identification tag was detected was taken, and the location of the detected identification tag on that floor within the building. In this way, the detection unit 33 detects the identification tag from the captured image, making it possible to estimate the shooting location of the image in which the identification tag was captured, similar to the case of a location-identifiable object M.

[0063] The building condition recording system described above is a building condition recording system that records the condition inside a building (construction site G), and comprises: a shooting recording device 2 configured to be held or attached to a site worker Q or a mobile body configured to move within the building, and which takes pictures of the inside of the building and acquires captured images; a movement information acquisition means 22 provided inside the shooting recording device 2 and which acquires movement information of the site worker Q or the mobile body associated with the captured images; a shooting position estimation unit 32 which estimates the shooting position (relative shooting position) of each captured image based on the movement information associated with the captured image; a detection unit 33 which detects an identification tag that can identify the installation position installed inside the building from the captured image; and a shooting position correction unit 34 which corrects the shooting position based on the installation position of the detected identification tag. With the above configuration, when a field worker Q or a mobile device configured to move within the building patrols the building, the building is photographed at any time by a photography and recording device 2, which is configured to be held or attached to the field worker Q or the mobile device, and the captured images are acquired. At the same time, movement information of the field worker Q or the mobile device is acquired at any time by a movement information acquisition means 22 provided inside the photography and recording device. This movement information is associated with the captured images, and the shooting position estimation unit 32 estimates the shooting position where each captured image was taken based on the movement information associated with that image. Therefore, by referring to these shooting positions and captured images, it is possible to understand what kind of images were taken at each shooting position within the building during the patrol. In this way, the state of the building can be recorded. Here, the detection unit 33 detects the identification tag from the image captured by the image recording device 2. Since the installation location of the identification tag within the building can be identified, if an identification tag is detected from the captured image, the shooting location where the image was taken can be accurately estimated from the installation location of the detected identification tag. Therefore, even if the accuracy of the shooting location estimated by the shooting location estimation unit 32 based on the movement information is not high, the accuracy of the shooting location can be improved by correcting the shooting location based on the installation location of the detected identification tag as described above. In this way, it is possible to further improve the accuracy of the shooting location where the image was taken. Furthermore, since the movement information acquisition means 22 is located inside the recording device 2, for example, when a site worker Q patrols the construction site, the site worker Q does not need to carry the movement information acquisition means 22 and the recording device 2 separately and operate them individually. In this way, the building condition recording system has a simple configuration and can be easily operated. In this way, when a field worker Q or a mobile device configured to move within the building patrols the building and takes pictures of the building with the shooting and recording device 2, and records the captured images along with their location, it becomes possible to improve the accuracy of the location while implementing a simpler configuration, making it easier to introduce and operate.

[0064] It should be noted that the building condition recording system of the present invention is not limited to the embodiments and modifications described above with reference to the drawings, and various other modifications are conceivable within its technical scope. For example, in the first embodiment, the shooting position was corrected based only on the installation position of the locatable object M, and in the second embodiment, it was corrected based only on the installation position of the identification tag. However, these can be combined, and the correction may be based on the installation positions of both the locatable object M and the identification tag. In this case, the building condition recording system is a building condition recording system that records the state inside a building (construction site G), and comprises: a shooting recording device 2 configured to be held or attached to a site worker Q or a mobile body configured to move within the building, and which takes pictures of the inside of the building and acquires captured images; a movement information acquisition means 22 provided inside the shooting recording device 2 for acquiring movement information of the site worker Q or the mobile body associated with the captured images; a shooting position estimation unit 32 that estimates the shooting position (relative shooting position) of each captured image based on the movement information associated with the captured image; a detection unit 33 that detects a location-identifiable object M and an identification tag installed inside the building from the captured image, and a shooting position correction unit 34 that corrects the shooting position based on the installation positions of the detected location-identifiable object M and the identification tag. In this case as well, similar to the first and second embodiments described above, when a field worker Q or a mobile body configured to move within the building patrols the building and takes photographs of the building with the photographic recording device 2, and records the captured images along with their locations, it becomes possible to improve the accuracy of the location while implementing a simpler configuration, thereby making it easier to introduce and operate.

[0065] In particular, when using objects such as electrical distribution boards or elevators as location-identifiable objects M, since these are installed on each floor of a building, if, for example, the location information of an electrical distribution board is located in a place where it is difficult to capture in the image, or if the elevator does not have a floor indicator, even if it is possible to detect the electrical distribution board or elevator from the image and determine the horizontal position where the image was taken, it may be difficult to determine the floor on which the image was taken. For this reason, in such cases, it is desirable to use identification tags in addition to location-identifiable objects M, as described above, and to identify the floor using the identification tags.

[0066] Furthermore, in the second embodiment described above, as explained as the first modification of the first embodiment, if the movement information acquired by the movement information acquisition means 22 is stored in the storage unit 23 separately from the captured image, the system may also be configured to include an information integration unit 37 that associates the movement information with the captured image. Alternatively, in the second embodiment described above, a building condition recording system may be used to implement the equipment management system, as explained as a second modification of the first embodiment.

[0067] Furthermore, while the construction progress management system of the first embodiment described above manages the progress of wall and ceiling construction within the construction site using a building condition recording system, it is also possible to manage the progress of any type of construction work within the construction site, not limited to wall and ceiling construction.

[0068] In addition to the above, it is possible to select or replace the configurations listed in each embodiment and each variation described above, or to change them to other configurations as appropriate. [Explanation of symbols]

[0069] 1. Building condition recording system 34. Shooting position correction unit 2. Shooting and recording device 37. Information integration unit 21. Filming methods 40. Construction progress management system 22 Means for acquiring movement information 50 Equipment management system 32 Shooting position estimation unit M Location identifiable object 33 Detection Unit Q Field Worker

Claims

1. A building condition recording system that records the conditions inside a building, A recording device configured to be held or attached to a site worker or a mobile body configured to be movable within the building, which photographs the inside of the building and acquires captured images, A means for acquiring movement information is provided inside the aforementioned recording device and is associated with the captured image, to acquire movement information of the field worker or the moving object. A shooting position estimation unit estimates the shooting position where each of the aforementioned captured images was taken, based on the movement information associated with the captured image. A detection unit that detects a location-identifiable object or identification tag installed within the building from the aforementioned captured image, A shooting position correction unit corrects the shooting position based on the installation position of the detected object that can be identified or the identification tag, A building condition recording system characterized by having the following features.

2. At least one of the aforementioned locatable objects is provided on each floor of the building. The building condition recording system according to feature 1.

3. The movement information acquisition means is an acceleration sensor and an angular velocity sensor, and the movement information is acceleration data and angular velocity data. The aforementioned movement information is either embedded within the captured image, or The system includes an information integration unit that, when the movement information and the captured image are separated, associates the movement information and the captured image over time. A building condition recording system according to feature 1 or 2.

Images