JP6511410B2 - Inspection apparatus for solar panel, inspection method and inspection program - Google Patents

Description

  The present invention relates to a solar panel inspection device, inspection method and inspection program.

  In conventional solar panel inspection, the inspector travels through a large solar power station to view the solar panel or use a desired type of image, such as a visible or infrared image of the acquired solar panel. Since the solar panels are visually inspected one by one to inspect the presence or absence of hot spots which are local high temperature points, a lot of time is required for the inspection.

  In recent years, from the viewpoint of reducing the movement load of inspectors, for example, a remote-controlled mobile body such as a radio control helicopter equipped with an imaging device for imaging a solar panel instead of the inspectors There has been proposed a solar panel inspection technique for moving and acquiring an image (inspection image) to be used for inspection of a solar panel such as an infrared image.

JP, 2015-146371, A

  In the case of inspection of a solar panel, since we want to obtain an overhead view image of the solar panel taken from above, the type of mobile object is, for example, radio control aircraft, Unmanned Aircraft System (UAS), small manned aircraft etc. An airborne flight vehicle with a desired image quality and range is preferred.

  However, even if the above moving object is made to acquire a desired type of image, the number of images to be acquired is enormous and varies depending on the size of the solar power plant, but at least the number of images of solar panels requiring confirmation Hundreds, in many cases thousands. In the conventional solar panel inspection apparatus, it is the inspector who confirms the acquired image, and the inspection burden on the inspector is still enormous.

  Moreover, in a solar power plant, since solar panels are arranged in order, it is difficult for the inspector to specify the abnormal position from the acquired image, and the burden is not small.

  On the other hand, focusing on the accuracy of the inspection, the method of inspecting a hot spot by visual inspection using an image (hereinafter referred to as “thermal image”) showing a temperature distribution such as an infrared image for inspection is visual inspection. Since it depends on the judgment of the inspector to confirm, there is a problem that the judgment result is varied.

  Moreover, when using a thermal image for an inspection, the risk of a misjudgment rises with temperature environment etc. at the time of image acquisition. For example, if the temperatures other than the solar panel and the solar panel are close, it is difficult to determine the contour of the solar panel, and the possibility of mistaking the range of the solar panel is increased. In addition, under a temperature environment exceeding a certain temperature, there is a high possibility that a high temperature place other than the solar panel may be erroneously determined as a hot spot.

  Furthermore, when using a thermal image for inspection, there is also an abnormality that does not accompany the heat generation phenomenon such as a crack, so in the confirmation by the heat image, there is a possibility that an abnormal part that does not accompany the heat generation phenomenon is overlooked. That is, there is a possibility of an erroneous determination in which an abnormal part which originally does not exert the ability at the time of soundness is judged as a normal part which exhibits the ability at the time of soundness.

  The present invention has been made in consideration of the above-described circumstances, and it is an object of the present invention to provide a solar panel inspection apparatus, inspection method and inspection program capable of automatically detecting an abnormal part of a solar panel with higher accuracy than before. I assume.

In order to solve the problems described above, a solar panel inspection apparatus according to an embodiment of the present invention detects a contour line of a solar panel based on a temperature distribution appearing in a thermal image taken from above the solar panel to be inspected. A first inspection range detection unit for detecting the range of each of the solar panels to be inspected; and a second detection of each range of the solar panels to be inspected from the acquired visible image by acquiring a visible image taken from above Inspection range detecting means for detecting an outline of each solar panel from the visible image and projecting the detected outline on the thermal image to detect an area of each solar panel to be inspected If, Oyo brightness L of each partial region plurality sets a partial region having a predetermined area on the thermal image of the solar panel, respectively, wherein the inspecting range detecting means detects A standard deviation value σ of the luminance L is calculated, and a threshold value t defined by the following equation [Equation 1] based on the average value μ of the luminance L and the standard deviation value σ in a partial region where the standard deviation value σ is the smallest In the case where it is determined that there is a portion showing temperature abnormality in the whole of the thermal image by comparing with the threshold value t as a threshold value t for determining the presence or absence of a temperature abnormality point or the visible image is inclined with respect to the outline. If it is determined that there is a line segment vector in another direction, the solar panel is determined to be abnormal, and the inspection range detection unit is inspected with the abnormality detection unit having the abnormality detection unit that detects the abnormal solar panel The position information acquisition means for acquiring the position information of the photographing position of the image used when detecting the range of each solar panel, and the map information including the position information in which the position of the solar panel to be inspected is written When the range of each of the solar panels to be inspected detected by the inspection range detection means, the position information of the photographing position of the image acquired by the position information acquisition means, and the abnormality detection means detects an abnormal solar panel Is an inspection result including inspection result information indicating at least the abnormal solar panel among the solar panels to be inspected on the map indicated by the map information, based on the abnormal solar panel and the read out map information And display processing means for generating an image.
t = μ + nσ (n is any positive number)

A solar panel inspection method according to an embodiment of the present invention detects a contour of a solar panel based on a temperature distribution appearing in a thermal image taken from above of the solar panel to be inspected, in order to solve the above-mentioned problems. A first inspection range detection unit for detecting the range of each of the solar panels to be inspected; and a second detection of each range of the solar panels to be inspected from the acquired visible image by acquiring a visible image taken from above Inspection range detecting means for detecting an outline of each solar panel from the visible image and projecting the detected outline on the thermal image to detect an area of each solar panel to be inspected If, Oyo brightness L of each partial region plurality sets a partial region having a predetermined area on the thermal image of the solar panel, respectively, wherein the inspecting range detecting means detects A standard deviation value σ of the luminance L is calculated, and a threshold value t defined by the following equation [Equation 1] based on the average value μ of the luminance L and the standard deviation value σ in a partial region where the standard deviation value σ is the smallest In the case where it is determined that there is a portion showing temperature abnormality in the whole of the thermal image by comparing with the threshold value t as a threshold value t for determining the presence or absence of a temperature abnormality point or the visible image is inclined with respect to the outline. If it is determined that there is a line segment vector in another direction, the solar panel is determined to be abnormal, and the inspection range detection unit is inspected with the abnormality detection unit having the abnormality detection unit that detects the abnormal solar panel Position information acquisition means for acquiring information of the imaging position of an image used when detecting the range of each solar panel, the range of each solar panel to be inspected detected by the inspection range detection means, and the position information Information on the photographing position of the image acquired by the acquiring means, and the abnormal one of the solar panels to be inspected based on the abnormal solar panel when the abnormal detection means detects the abnormal solar panel A solar panel inspection procedure using a solar panel inspection apparatus comprising: a display processing means for generating an inspection result image including inspection result information in which a solar panel is distinguished from other solar panels, the solar panel inspection procedure A first inspection range detection step in which the inspection range detection means detects the range of each of the solar panels to be inspected based on the temperature distribution appearing in the thermal image, and the second inspection range detection unit A second inspection area detection step of detecting the area of the solar panel to be inspected from the acquired visible image; If the detecting step and the abnormality detecting means determine that there is a portion indicating temperature abnormality from the range of each of the solar panels detected in the inspection range detecting step, the solar panel is determined to be abnormal, and the abnormality is abnormal. An abnormality detection step including an abnormality detection step of detecting a solar panel, and a position where the position information acquisition means acquires information of a photographing position of an image used when detecting the range of each solar panel in the inspection range detection step An information acquisition step, the display processing means detects the range of each of the solar panels detected in the inspection range detection step, information of the photographing position of the image acquired in the position information acquisition step, and the abnormality detection step Out of the solar panels to be inspected based on the detected abnormal solar panel if an abnormal solar panel is detected in And a display processing step of generating an inspection result image including inspection result information in which the abnormal solar panel is shown separately from other solar panels.
t = μ + nσ (n is any positive number)

A solar panel inspection program according to an embodiment of the present invention detects a contour of a solar panel based on a temperature distribution appearing in a thermal image taken from above of the solar panel to be inspected, in order to solve the problems described above. A first inspection range detection step of detecting the range of the solar panel to be inspected; acquiring a visible image taken from above, and detecting an outline of each of the solar panels to be inspected from the acquired visible image; A second inspection range detection step of detecting a range of each of the solar panels to be inspected by projecting the detected outline onto the thermal image; and an inspection range detection step, and the detection detected in the inspection range detection step a plurality set a partial area to calculate the standard deviation value σ of the luminance L and the luminance L of each partial region having a predetermined area in the thermal image of the solar panel, The threshold value t defined by the following equation [Equation 1] based on the average value μ of the luminance L and the standard deviation value σ in a partial region where the standard deviation value σ is the smallest is a threshold value t for determining the temperature abnormal point presence or absence It is determined that there is a portion showing temperature abnormality in the whole of the thermal image by comparison with the threshold value t or that there is a line segment vector in a direction inclined with respect to the outline in the visible image In this case, an abnormality detection step including an abnormality detection step of determining the solar panel as abnormal and detecting an abnormal solar panel, and an imaging position of an image used when detecting the range of the solar panel in the inspection range detection step Position information acquiring step of acquiring information of the information, the range of the solar panel detected in the inspection range detecting step, information of the photographing position of the image acquired in the position information acquiring step, and the information If an abnormal solar panel is detected in the normal detection process, the abnormal solar panel is determined based on the detected abnormal solar panel, and the solar panels and zones not determined as abnormal in the abnormal detection process. A computer is caused to execute a solar panel inspection procedure including a display processing step of generating an inspection result image indicating the solar panel to be inspected.
t = μ + nσ (n is any positive number)

  According to the present invention, it is possible to automatically detect an abnormal part of a solar panel with higher accuracy than before.

The schematic which showed the hardware structural example of the test | inspection apparatus of the solar panel which concerns on embodiment of this invention. Explanatory drawing which shows the example of acquisition of the image used when the test | inspection apparatus of the solar panel which concerns on embodiment of this invention determines the presence or absence of abnormality about each of a solar panel. BRIEF DESCRIPTION OF THE DRAWINGS The functional block diagram which shows the functional structural example of the panel test | inspection apparatus which concerns on the 1st Embodiment of this invention. The processing flow figure showing the flow of processing of the panel inspection procedure (the 1st-3rd panel inspection procedure) which the panel inspection device concerning an embodiment of the present invention performs. The processing flow figure showing the flow of processing of the 1st inspection range detection process in the panel inspection procedure which the panel inspection device concerning an embodiment of the present invention performs. The processing flow figure showing the flow of processing of the hot spot detection process in the panel inspection procedure which the panel inspection device concerning a 1st embodiment of the present invention performs. The processing flow figure showing the flow of processing of the unusual part detection process in the panel inspection procedure which the panel inspection device concerning a 1st embodiment of the present invention performs. The processing flow figure which shows the flow of processing of the 1st inspection result picture generation process in the panel inspection procedure which the panel inspection device concerning a 1st embodiment of the present invention performs. The functional block diagram which shows the functional structural example of the panel test | inspection apparatus which concerns on the 2nd Embodiment of this invention. The processing flow figure which shows the flow of a process of the 2nd test result image generation process in the panel test procedure which the panel test | inspection apparatus which concerns on the 2nd Embodiment of this invention performs. A functional block diagram showing an example of functional composition of a panel inspection device concerning a 3rd embodiment of the present invention. The processing flow figure which shows the flow of processing of the 2nd inspection range detection process in the panel inspection procedure which the panel inspection device concerning a 3rd embodiment of the present invention performs. The processing flow figure which shows the flow of a process of the 3rd test result image generation process in the panel test procedure which the panel test | inspection apparatus which concerns on the 3rd Embodiment of this invention performs.

  Hereinafter, an inspection device, an inspection method, and an inspection program of a solar panel according to an embodiment of the present invention will be described with reference to the drawings.

  FIG. 1 is a schematic view showing an example of a hardware configuration of a solar panel inspection apparatus (hereinafter, referred to as “panel inspection apparatus”) according to an embodiment of the present invention.

  The panel inspection apparatus according to the embodiment of the present invention is, for example, a computer 1 which is hardware having an arithmetic processing function as a solar panel inspection program (hereinafter referred to as “panel inspection PG”) according to the embodiment of the present invention. ) Is realized by causing the computer 1 to function as a solar panel inspection apparatus provided with means for inspecting the presence or absence of an abnormality of a solar panel (object to be inspected) installed in a solar power plant etc. Be done. That is, the cooperation of the computer 1 which is hardware and the panel inspection PG which is software realizes a function of inspecting the presence or absence of abnormality of the solar panel in the computer 1.

  Here, the panel inspection PG 10 (10A to 10C) is an example of a panel inspection program according to the embodiment of the present invention, and is a program that causes the computer 1 as hardware to function as the panel inspection device 50 (50A to 50C). . From the viewpoint of allowing the computer 1 to execute a panel inspection procedure, which is an example of a solar panel inspection method (hereinafter referred to as “panel inspection method”) according to the embodiment of the present invention, the panel inspection PG 10 is a computer 1 is a program that executes the panel inspection procedure.

  The computer 1 includes, for example, a CPU (central processing unit) 2 which is an example of a processor, a main storage device 3 such as a random access memory (RAM), and a read only memory (ROM) and a hard disk drive (HDD). An auxiliary storage device 4, an input device 5 such as a keyboard and a mouse, an output device 6 such as a display and a printer, and a communication device 7 for communicating with an external device.

  For example, the computer 1 can be connected to an external device on the network such as the calculation server 8 via the communication device 7, and can transmit and receive data to and from the external device. For example, the computer 1 can cause the calculation server 8 to execute at least a part of the arithmetic processing or receive an execution result. The computer 1 can also make network connection with external devices other than the calculation server 8.

  In the auxiliary storage device 4 accessible by the CPU 2 such as a ROM, a program or data used for inspection, for example, a panel inspection PG 10, a map which is an electronic map showing a place where a solar panel to be inspected is installed. Data 11, ortho-image 13 including an area where all the solar panels to be inspected are photographed, panel layout data 15 etc. which is an electronic drawing showing the arrangement of all solar panels to be inspected It is stored in the device 4.

  The computer 1 loads the panel inspection PG 10 stored in the auxiliary storage device 4 and the data necessary for the execution of the panel inspection PG 10 into the main storage device 3 such as RAM, and executes processing according to this program. The main storage device 3 such as a RAM provides a work area for temporarily storing programs executed by the CPU 2 and data.

  Then, the panel inspection apparatus which concerns on embodiment of this invention demonstrates the image used when determining the presence or absence of abnormality with respect to each of a solar panel (test object).

  FIG. 2 shows an image (thermal image 22) used when the panel inspection apparatus 50 for inspecting the solar panels 21 installed in the solar power generation plant 20 determines the presence or absence of an abnormality in each of the solar panels 21. It is an explanatory view showing an example of acquisition of a visible image 23 and a high resolution visible image 24).

  In the solar power plant 20, in addition to a plurality of solar panels 21 for solar power generation, a building 25 and the like in the office are installed. In the panel inspection apparatus 50, for example, while moving the UAS 30 by following the moving path such as the flight path R, the images 22, 23, 24 of all the solar panels 21 to be inspected without leakage are communicated by wireless Alternatively, the images 22, 23, 24 are acquired via a storage medium such as a magnetic disk, an optical disk, or a semiconductor memory storing the images 22, 23, 24.

  The images 22, 23, 24 are all images obtained by photographing the solar panel 21 constituting at least a part of the inspection object from above, and for example, an unmanned aerial vehicle equipped with an imaging device capable of imaging a desired image quality and range. It is a so-called bird's-eye view image photographed using a mobile body such as a system (hereinafter simply referred to as "UAS") 30 or the like.

  The thermal image 22 is an image showing a temperature distribution, and is acquired, for example, using a thermography 26 as an imaging device.

  The visible image 23 is an image representing the reflection of a light wave (visible light) in a wavelength band (about 380 nm to 780 nm) visible to human beings, and is acquired using, for example, a visible camera 27 as a photographing device.

  It is preferable that the thermal image 22 and the visible image 23 have the same range captured in one sheet, but they do not have to be the same.

  For example, when obtaining a thermal image 22 and a visible image 23 using different imaging devices, such as using a thermography 26 for obtaining the thermal image 22 and using a visible camera 27 for obtaining the visible image 23 It is possible that the ranges captured in the thermal image 22 and the visible image 23 differ due to the difference between the angle of view and the imaging timing of both imaging devices.

  If the angle of view between the thermography 26 and the visible camera 27 is different, the same range as the imaging range of the thermography 26 may be set in the visible camera 27. By this setting, it is possible to set a range corresponding to the whole range of the thermal image 22 with respect to the visible image 23, similarly to the case where the photographed range is matched to one of the thermal image 22 and the visible image 23. The panel inspection procedure described later can be performed.

  A camera (hereinafter, simply referred to as an “integrated camera”) 28 integrally provided with the thermography 26 and the visible camera 27 may be used as an imaging device for acquiring the thermal image 22 and the visible image 23.

  Moreover, when using the integrated camera 28, the resolution of the visible image 23 to be obtained may not be sufficient as compared with the case of using the general visible camera 27 (without the acquisition function of the thermal image 22). Therefore, a visible camera (hereinafter referred to as “high resolution visible camera”) 29 having a resolution higher than the resolution of the visible image 23 obtained by the integrated camera 28 may be mounted separately from the integrated camera 28.

  In the case where the panel inspection apparatus 50 acquires the visible images 23 of two different resolutions by using the UAS 30 equipped with the integrated camera 28 and the high resolution visible camera 29, for example, from the viewpoint of distinguishing between both visible images 23 The relatively high resolution visible image 23 is referred to as a high resolution visible image 24, and the relatively low one is simply referred to as a visible image 23. In particular, when it is not necessary to distinguish between the two visible images 23, they are simply referred to as visible images 23.

  The images 22, 23, 24 used in the panel inspection apparatus 50 are usually not a single still image, but a plurality of still images obtained by capturing mutually different areas (including some overlapping areas). It covers the area including all of the solar panels 21 to be inspected.

  Note that the images 22, 23, 24 are not necessarily limited to still images. If all of the solar panels 21 to be inspected as a whole are included, it may be a moving image. The advantage of using a moving image is that the moving speed of the UAS 30 can be increased by performing moving image shooting with a frame rate higher than that of a still image. The advantage of being able to increase the flight speed of the UAS 30 means that a wider range can be taken in one aerial shot. Therefore, when photographing the same area, it is possible to further reduce the aerial photographing time.

  In the following description, when it is necessary to distinguish between the thermal image 22, the visible image 23 and the high resolution visible image 24 between a moving image and a still image, the thermal moving image 22, the visible moving image 23 and the high The still images are referred to as a thermal still image 22, a visible still image 23, and a high resolution visible still image 24, respectively.

  In addition, it is preferable that the thermal moving image 22 and the visible moving image 23 (including the high resolution visible moving image 24) synchronize moving images of the thermography 26 and the visible camera 27 or the integrated camera 28 and the high resolution visible camera 29. When the two are synchronized and shooting moving images, the thermal moving image 22 and the visible moving image 23 can be input to the panel inspection apparatus 50 as they are without the pre-processing step.

On the other hand, even when the timing of starting the moving image shooting of the thermography 26 and the visible camera 27 or the integrated camera 28 and the high resolution visible camera 29 is deviated (the moving image is taken asynchronously), the panel inspection device 50 Although it is possible to inspect the solar panel 21 using video, this case, in consideration of the shift of the imaging start timing of the thermal video 22 and the visible video 23 to enter into the panel inspection apparatus 50, starts the input moving A process of adjusting the position of the frame and a process of cutting out a moving image frame at the same time as the thermal still image 22 and the visible still image 23 are required.

  In the panel inspection apparatus 50, the inspection range of the solar panel 21 and the individual solar panels 21 are identified and identified using the thermal image 22 obtained by photographing at least the area including the whole of the solar panel 21 to be inspected from above. The presence or absence of abnormality of the solar panel 21 present in the inspection range is determined, and the solar panel 21 determined to be abnormal is displayed in a distinguishable manner as the solar panel 21 determined to be non-abnormal (healthy).

  Subsequently, a panel inspection apparatus, a panel inspection method, and a panel inspection PG according to each embodiment will be described. The panel inspection devices 50 (50A to 50C) described later are devices realized by causing the computer 1 (FIG. 1) to execute panel inspection PGs 10 (10A to 10C) (FIG. 1).

First Embodiment
FIG. 3 is a functional block diagram showing a functional configuration of a panel inspection apparatus 50A which is an example of the panel inspection apparatus according to the first embodiment of the present invention.

  The panel inspection apparatus 50A (FIG. 3) includes, for example, an inspection range detection including an input unit 51, an output unit 52, a communication unit 53, a first inspection range detection unit 541, and a second inspection range detection unit 542. Means 54A, an abnormality detection means 55 comprising a hot spot detection unit 551 and an unusual place detection unit 552, a position information acquisition means 56, a display processing means 57A comprising a first inspection result image generation unit 571; It comprises the means 58, the control means 59, and the still picture extraction means 61.

  The input unit 51 is realized, for example, by an input device connected to a computer via an interface or by an input unit such as a keyboard or a mouse provided in the computer itself. The input unit 51 receives an input of information and gives the received information to the control unit 59. From the input means 51, there are, for example, a request for execution of a panel inspection procedure to be described later, a request for selection of an image to be read or panel layout data 15, a request for setting of various conditions, and the like.

  The output unit 52 is realized by, for example, a display unit connected to the computer via an interface or a display unit such as a display provided in the computer itself, a printing unit such as a printer connected to the computer via the interface, or the like. When receiving the display request, the output unit 52 displays the contents corresponding to the display request on the screen. Further, when receiving the printing request, the output means 52 prints out the contents according to the printing request.

  The communication unit 53 includes external devices such as the computer server 8 illustrated in FIG. 1 and the unmanned aerial system (UAS) 30 illustrated in FIG. 2 (including equipment such as a mounted control device and an imaging device) and data. It has a function to transmit and receive. The communication means 53 transmits the data received from the control means 59 to the external device, and gives the data sent from the external device to the control means 59.

  The inspection area detection means 54A has a function of detecting the area of each solar panel 21 for each of the solar panels 21 to be inspected from the provided inspection image (thermal image 22 or thermal image 22 and visible image 23). Have. The inspection range detection means 54A includes, for example, a first inspection range detection unit 541 for detecting the range of each solar panel 21 to be inspected based on the temperature distribution appearing in the thermal image 22, and the visible image 23 (high resolution visible image 24). And a second inspection range detection unit 542 for detecting the range of the individual solar panels 21 for each of the solar panels 21 to be inspected.

  In addition to the first inspection range detection unit 541, the inspection range detection unit 54 </ b> A further includes a second inspection range detection unit 542 because only the temperature distribution appearing in the thermal image 22 is exactly the solar panel 21. In consideration of the possibility that the range of

  For example, depending on the surrounding environment in which the solar panel 21 is installed, the temperature environment at the time of shooting, etc., there may be no difference in temperature between the solar panel 21 and the surroundings of the solar panel 21. The contour of the solar panel 21 does not appear clearly only by the temperature distribution. Therefore, by configuring the inspection range detection means 54A so that the contour of the solar panel 21 appearing in the visible image 23 can be identified, the panel inspection device 50A reduces the possibility of erroneous detection of the range of the solar panel 21.

  The abnormality detection means 55 determines whether or not there is an abnormality state (normal state) in a state in which the ability in the normal state can not be exhibited within each range of the solar panel 21 detected by the inspection range detection means 54A. Identification information (hereinafter referred to as "image identification information") of an image used for the determination and the result of determining whether each solar panel 21 is normal or abnormal, and identification information (hereinafter referred to as "image identification information") And "panel identification information."

  The abnormality detection unit 55 detects, for example, a hot spot detection unit 551 that detects whether or not there is a temperature abnormality spot (hot spot) in the solar panel 21, and some external force is applied to the solar panel 21 to damage such as cracks. A non-ordinary part detection unit that detects a part where the solar panel 21 is in a state different from the normal state (hereinafter referred to as “unusual”), such as when there is an obstacle And 552.

  The abnormality detection means 55 illustrated in FIG. 3 is a hot spot detection unit, for example, regarding whether or not there is a location (hot spot) in the solar panel 21 where a temperature abnormality occurs in the solar panel 21 regarding the presence or absence of an abnormality of the solar panel 21. The abnormal point detection unit 552 detects whether or not an abnormal point is detected in the solar panel 21 by detecting 551.

  The hot spot detection unit 551 detects, as a hot spot, a spot having a brightness exceeding a threshold value to be set based on the temperature distribution in the solar panel 21, that is, the brightness distribution in the solar panel 21 in the thermal image 22. Do.

  The non-ordinary point detection unit 552 detects line segment vectors from a region in the solar panel 21 (including the outer edge forming the contour) shown in the visible image 23 to form the contour of the solar panel 21 in the longitudinal direction and the lateral direction. It is determined whether or not an unusual part exists by determining whether a line segment vector in a direction inclined (oblique) with respect to the line segment exists.

  Since the unusual location includes a location in an unusual and abnormal state such as a crack or other damage, the anomaly detecting means 55 including the unusual location detecting unit 552 causes a temperature anomaly. It is possible to detect an abnormality other than the location (hot spot). In addition, if the current condition of the sound (currently operating) condition is left unhealthy, there is an event that can be the cause of hot spot occurrence (hereinafter referred to as "abnormal standby state") Since the location is also included, the anomaly detection unit 55 including the non-ordinary location detection unit 552 can detect an anomaly reserve state of the solar panel 21. By making it possible to eliminate the abnormal reserve state in advance, the risk of the future occurrence of an abnormality in the solar panel 21 can be further reduced.

  The abnormality detection means 55 has two operations: a first event that there is a location (hot spot) causing a temperature anomaly in the solar panel 21 and a second event that there is an unusual location in the solar panel 21. Among the events, when at least the first event occurs, it is determined as "abnormal", and when no event occurs, it is determined as "no abnormality" (normal).

It should be noted that, in the case where only the second event has occurred, more accurately, "(There is suspicion of abnormal) non-normal" becomes to a determination that, as viewed in the safer side from the viewpoint of promoting the inspection by the people "abnormal The abnormality detection unit 55 may be configured to determine that there is a “presence”.

  The position information acquisition means 56 has a function of acquiring position information indicating the shooting position of the inspection image (for example, the thermal image 22 etc.) used when the inspection range detection means 54A detects each range of the solar panel 21. .

  The position information acquisition unit 56 acquires the position information added thereto, for example, when the position information indicating the latitude, longitude, and altitude measured by the GPS, such as a geo tag, is added to the image, the thermal image 22 And the like (hereinafter, referred to as “shooting position information”) indicating the shooting position of the inspection image.

  Further, when the imaging position information of the inspection image is not added to the inspection image such as the thermal image 22, the thermal image is acquired by acquiring the information including the imaging position information from elements other than the image such as the UAS 30. The imaging position information of the inspection image such as 22 is obtained.

  Usually, the image includes information of shooting time (hereinafter referred to as “shooting time information”). Therefore, if there is shooting position information of the image associated with the time information, the time when the image was shot It is possible to acquire shooting position information in

  For example, since the flight log of UAS 30 used for shooting includes the flight time of UAS 30 and position information indicating the flight position at that time, using this flight log, an image such as a thermal image 22 etc. The flight position of the UAS 30 at the shooting time, that is, the shooting position can be specified, and shooting position information of an image such as the thermal image 22 can be acquired.

  The display processing means 57A is, for example, an image showing the range of the solar panel detected by the inspection range detection means 54A, an image showing an abnormal solar panel detected when the abnormal solar panel detected by the abnormality detection means 55 is detected, etc. The present invention has a function of generating display information for causing display means such as a display to display information including an inspection result image indicating the inspection result of the above. That is, in the panel inspection apparatus 50A, the display processing unit 57A includes a first inspection result image generation unit 571 as an inspection result image generation unit that generates an inspection result image.

  The first inspection result image generation unit 571 performs at least on an image (hereinafter, referred to as a “background image”) in which the position of the solar panel 21 to be inspected, such as a map image used to display the inspection result. It has a function of generating an inspection result image which is superimposed and displayed by matching (positioning) the thermal image 22 captured by the solar panel 21 detected and determined as abnormal by the abnormality detection means 55 and corresponding to the position of the map image. The map image is included in the map data 11 as map information including position information of the place where the solar panel 21 is installed, and is acquired by reading out the given map data 11.

  In addition, the first inspection result image generation unit 571 has a function of generating an inspection result image that displays the solar panel 21 determined to have no abnormality by the abnormality detection unit 55 in correspondence with the position of the map image. It is also good. That is, all the solar panels 21 whose abnormality detection means 55 inspects the presence or absence of abnormality may have a function of generating an inspection result image to be displayed corresponding to the position of the map image.

  When the display processing means 57A receives the information to be displayed such as the inspection result image, the display processing means 57A generates display information for displaying the received content, and gives the generated display information to the control means 59. How to display the information may be initially set, or may be set from the input unit 51 each time.

  In addition to the case where the first inspection result image generation unit 571 generates the inspection result image using the map image included in the map data 11 as the background image, the position of the region including the solar panel 21 to be inspected When the ortho image 13 having information is provided by a private sector or a government agency, the ortho image 13 can also be used as a background image.

  When using the ortho image 13 as a background image, alignment between the images is performed using the position information of the thermal image 22 acquired by the position information acquisition means 56 and the position information such as geotags possessed by the ortho image 13 The thermal image 22 including the information of the inspection result can be projected at 13.

  The storage unit 58 includes a storage area capable of reading (writing) and writing (writing) information, and has a function of holding information in the storage area. The storage unit 58 provides the inspection range detection unit 54A, the abnormality detection unit 55, the display processing unit 57A, and the control unit 59 with a storage area in which information can be read and written. Reading and writing are performed.

  The control unit 59 is a unit that controls the processing of the entire panel inspection apparatus 50A, and includes an input unit 51, an output unit 52, a communication unit 53, an inspection range detection unit 54A, an abnormality detection unit 55, a still image extraction unit 61, It has a function of mutually exchanging data with the display processing means 57A and the storage means 58 and controlling them.

  The still image extraction means 61 has a function of cutting out a frame of the moving image as a still image such as the heat still image 22 when the inspection image such as the given thermal image 22 is a moving image. The still image such as the thermal still image 22 obtained by the still image extraction means 61 is given to the inspection range detection means 54A.

  The panel inspection apparatus 50A configured in this manner automatically detects the range of each solar panel 21 using the thermal image 22 obtained by photographing the solar panel 21 to be inspected from above, and detects each solar panel 21 Since it is detected whether or not there is a place where an abnormality occurs in the inside, the burden on the inspector required for the inspection of the solar panel 21 can be reduced.

  Further, in the panel inspection apparatus 50A, the inspection area detection unit 54A includes the second inspection area detection unit 542 so that a crack or the like can not be sufficiently specified from the thermal image 22 in the range of each solar panel 21. The outline of each solar panel 21 can be detected using the visible image 23 used to detect the range of the solar panels 21 also when detecting the range of each solar panel 21.

  Therefore, in the panel inspection apparatus 50A including the second inspection range detection unit 542, the outline of each solar panel 21 can be detected more accurately than from the thermal image 22 alone, and the outline detected from the visible image 23 is By projecting onto the thermal image 22, the range of each solar panel 21 can be detected more accurately, and an abnormal place appearing on each solar panel 21 can be automatically detected more accurately than in the past.

  The panel inspection device 50A is not limited to the configuration illustrated in FIG. For example, when the thermal image 22 and the visible image 23 to be input are still images, etc., it is not necessary to cut out the still image from the moving image, even if the panel inspection apparatus 50A does not necessarily have the still image extraction unit 61. Good.

  Further, in the panel inspection apparatus 50A, the thermal image 22 having a high resolution can be obtained so that the individual solar panels 21 can be sufficiently identified, and there is no particular problem in detecting the range of the solar panels 21 from the thermal image 22. If it exists, the inspection range detection means 54A does not necessarily have to include the second inspection range detection unit 542.

  Furthermore, in the panel inspection apparatus 50A, while the hot spot of the solar panel 21 is detected, when the abnormal part is not detected, it is necessary that the abnormality detecting unit 55 necessarily includes the abnormal part detecting unit 552 There is no.

  Next, as a panel inspection method according to the first embodiment of the present invention, a first panel inspection procedure performed by the panel inspection apparatus 50A will be described.

  FIG. 4 is a processing flow chart (flow chart) showing the flow of processing of first to third panel inspection procedures (steps S1 to S5) performed by the panel inspection apparatus 50 (50A to 50C).

  The first to third panel inspection procedures illustrated in FIG. 4 are processing procedures performed by the panel inspection devices 50A to 50C, respectively, and among the first to third panel inspection procedures, the preprocessing step (steps) S1), the hotspot detection step (step S3) of the abnormality detection step and the non-ordinary part detection step (step S4) are common processing steps in any panel inspection procedure.

On the other hand, it is other processing steps, the inspection range first inspection range detecting step as a detecting step (step S2) and a second inspection area detection step (step S 7), as well as the inspection result image first as generating step inspection result image generating process (step S5), and a second test result image generation step (step S 6) and the third test result image generation step (step S8) is different in the first to third panel test procedure It is a processing step.

  Among the first to third panel inspection procedures illustrated in FIG. 4, the first panel inspection procedure includes a pre-processing step (step S1) performed as needed, and a first inspection range detection step (steps). S2), a hot spot detection step (step S3) and an abnormal part detection step (step S4) of the abnormality detection step, and a first inspection result image generation step (step S5).

In the first panel inspection procedure, when the processing step is started (START), a pre-processing step (step S1) is performed as necessary. Before and processing steps, for example, thermal video 22 (Fig. 2) to enter into the panel inspection apparatus 50A (FIG. 3) and the visible video 23 same imaging from when and videos imaging start timing is not synchronized (Figure 2) It is performed when cutting out a still image that covers the range. When the pretreatment process is completed, a first inspection range detection process (step S2) is subsequently performed.

  The preprocessing step (step S1) may be performed by the panel inspection device 50A, but may be performed by another device. In this case, the pre-processing step is skipped, and is performed from the subsequent inspection range detection step (step S2) using the thermal image 22 and the visible image 23 obtained by performing the pre-processing step in advance with another device. Also, the pre-processing step is skipped even when the pre-processing itself is unnecessary.

  In the first inspection range detection step (step S2), the inspection range detection means 54A (FIG. 3) specifies each range of the solar panel 21 (FIG. 2) to be inspected with respect to the thermal image 22. , And a process step of detecting the range of each of the solar panels 21. When the range of each solar panel 21 is detected, an abnormality detection step (steps S3 and S4) is subsequently performed.

  The abnormality detecting step (steps S3 and S4) is a processing step in which the abnormality detecting unit 55 (FIG. 3) detects an abnormality in each solar panel 21 detected in the first inspection range detecting step (step S2). And a hotspot detection step (step S3) for detecting a hotspot at least. The abnormality detection process illustrated in FIG. 4 includes a non-ordinary part detection process for detecting an abnormal part other than the hot spot (step S3) and an abnormal part other than the hot spot which may be a precursor of an abnormality or an occurrence of an abnormality. (Step S4) is further provided.

  In the first inspection result image generation step (step S5), the first inspection result image generation unit 571 (FIG. 3) of the display processing means 57A (FIG. 3) is abnormal on the map used to display the inspection results. It is a processing step which generates an inspection result picture which shows whether solar panel 21 exists, and is a processing step which constitutes a part of display processing step. When the inspection result image is generated and the first inspection result image generation step (step S5) is completed, the first panel inspection procedure ends all processing steps (END).

  Subsequently, in a first panel inspection procedure, a first inspection range detection step (step S2), a hot spot detection step (step S3), a non-ordinary part detection step (step S4) and a first inspection result image generation step (Step S5) will be described.

  FIG. 5 is a processing flow diagram showing the flow of processing of the first inspection range detection step (step S2: FIG. 4).

  The first inspection range detection step identifies each range of the solar panel 21 with respect to the inspection image such as the thermal image 22 used in the later abnormality detection step (FIG. 4), and the identified range is the inspection range It is a processing step to detect as.

  In the first inspection range detection step, each solar panel reflected in the thermal image 22 on the basis of the outline detected from the thermal image 22 or the visible image 23 (including the high resolution visible image 24. The same shall apply hereinafter). When detecting the outline of 21, the selection as to whether or not to use the visible image 23 is received, and the processing steps of step S 21 to step S 26 are performed according to the received selection content.

  The first inspection range detection step (steps S21 to S26) is, for example, a step of detecting an outline of each of the solar panels 21 from the thermal image 22 (step S21), and is directly detected from the thermal image 22. The range of each solar panel 21 is specified on the basis of the outline of each solar panel 21, and the step of setting the range of each solar panel 21 (step S23) and the outline for each of the visible images 23 to the solar panel 21 (Step S24), and projecting an outline detected from the visible image 23 onto the thermal image 22 (step S25), and each solar panel 21 detected from the visible image 23 and projected onto the thermal image 22. Identifying the range of each solar panel 21 based on the outline of the Step S26) and a.

  When the first inspection range detection step is started (ENTER), first, the first inspection range detection unit 541 detects the outline of the solar panel 21 to be inspected from the thermal image 22 obtained by photographing the solar panel 21. (Step S21). Here, when the visible image 23 is not used to detect the outline of each solar panel 21 (in the case of NO at step S22), the first inspection range detection unit 541 detects each solar panel directly detected from the thermal image 22. The range of each solar panel 21 is specified based on the outline of 21 and detected as the range of each solar panel 21 (step S23).

  When the first inspection range detection unit 541 detects each range of the solar panels 21, panel identification information is attached to each of the detected solar panels 21, and an image of a detection source image (here, the thermal image 22) The first inspection range detection process is completed in association with the identification information (RETURN).

  On the other hand, when the visible image 23 is used to detect the outline of each solar panel 21 (in the case of YES at step S22), the second inspection range detection unit 542 applies the visible image 23 to each of the solar panels 21. An outline is detected (step S24), and the detected outline is projected on the thermal image 22 (step S25).

  The projection of the outline of each solar panel 21 detected from the visible image 23 onto the thermal image 22 is, for example, position information of a plane ignoring the height of the visible image 23 and the thermal image 22 (hereinafter referred to as “planar position information” The visible image 23 is temporarily decided on the thermal image 22 with reference to the), and the feature quantity extracted from the thermal image 22 is searched in the visible image 23, and the feature quantity matches from the tentatively decided position. The visible image 23 is moved to the closest position to be. Then, the position of the visible image 23 after movement where the feature amounts match is determined as the position of the final visible image 23, and the position on the thermal image 22 overlapping the position of the outline detected from the visible image 23 is It is performed by using the outline of the panel 21.

  By performing matching processing of feature amounts in such an image, even if the thermography 26 for capturing the thermal image 22 and the visible camera 27 for capturing the visible image 23 are not coaxial, they are detected from the visible image 23 The outline of each solar panel 21 can be aligned with the thermal image 22 without misalignment.

  When the projection of the outline detected from the visible image 23 onto the thermal image 22 is completed, the second inspection range detection unit 542 is projected on the thermal image 22 using the thermal image 22 after the outline projection. The range of each solar panel 21 is specified based on the outline of each solar panel 21 and detected as the range of each solar panel 21 (step S26).

  When the second inspection range detection unit 542 detects each range of the solar panels 21, panel identification information is attached to each of the detected solar panels 21, and an image of the detection source image (here, the thermal image 22) The first inspection range detection process is completed in association with the identification information (RETURN).

  Note that the first inspection range detection step described above does not necessarily have to include all the processing steps of step S21 to step S26.

  For example, assuming that the visible image 23 is not used when detecting the outline of each solar panel 21, the range of each of the solar panels 21 to be inspected is detected based on the temperature distribution appearing in the thermal image 22. It may be a first inspection range detection step including the first inspection range detection step. That is, processing steps (steps S21 and S23) obtained by omitting steps S22 and S24 to S26 from the first inspection range detection step (steps S21 to S26) described above are used as the first inspection range detection step. It can also be an inspection range detection step.

  Also, conversely, assuming that the visible image 23 is used when detecting the outline of each solar panel 21, the outline of each solar panel 21 is detected from the visible image 23 and the detected outline is detected. It may be a first inspection range detection step including a second inspection range detection step of detecting the range of each of the solar panels 21 to be inspected as a reference. That is, an inspection range detection step including processing steps (steps S21 and S23) obtained by omitting steps S22 and S23 from the first inspection range detection step (steps S21 to S26) described above as a second inspection range detection step It can also be done.

  FIG. 6 is a process flow diagram showing a process flow of the hot spot detection process (step S3: FIG. 4).

  The hotspot detection process (steps S31 to S38) includes, for example, a brightness average calculation step (step S31), a brightness standard deviation calculation step (step S32), a threshold automatic setting step (step S33), and a brightness determination step (step S33). And step S34 to step S37).

  When the hot spot detection process is started (ENTER), first, the hot spot detection unit 551 (FIG. 3) calculates the average value μ of the luminance L in one solar panel 21 (FIG. 2) appearing in the thermal image 22. Then, the standard deviation value σ of the luminance L in the same solar panel 21 is calculated (step S32).

After calculating the average value μ and the standard deviation value σ, the hot spot detection unit 551 subsequently calculates a threshold t defined by the following equation (1), and the calculation result (calculated value) is used to determine the presence or absence of a hot spot. Set as a threshold t.
[Equation 1]
t = μ + nσ (1) (n is any positive number)

  In panel inspection device 50, calculation is performed in step S31 and step S32 without setting threshold value t itself by giving an arbitrary value to positive number n in the right-hand side term together with the above equation (1). The hot spot detection unit 551 can be configured to automatically set the threshold value t calculated further using the average value μ and the standard deviation value σ as the threshold for hot spot presence / absence determination.

  When the hot spot detection unit 551 sets the threshold value t (step S33), subsequently, the hot spot detection unit 551 compares whether the luminance L exceeds the threshold value t with respect to the threshold value t (is equal to or less than the threshold value t). (Step S34). This is because, in the thermal image 22, the higher the temperature is, the higher the luminance is. Therefore, the portion where the higher luminance L is observed with respect to the peripheral temperature (luminance) can be regarded as a hot spot.

  As a result of the hot spot detection unit 551 comparing the threshold value t with the brightness L, if the brightness L is equal to or less than the threshold value t (YES in step S34), the hot spot detection unit 551 determines that the spot giving the brightness L is a hot spot Is not determined (no abnormality) (step S35), and when the luminance L exceeds the threshold t (in the case of NO in step S34), it is determined that the spot to which the luminance L is given is a hot spot (abnormal) (Step S36).

  When it is determined whether the spot giving luminance L is a hot spot (step S35 or step S36), it is confirmed whether there is another solar panel 21 detected in the thermal image 22 (step S35). S37) When it remains (in the case of NO in step S37), the processing steps from step S31 to step S36 are performed for the remaining solar panel 21.

  On the other hand, in the case where the determination (step S35 or step S36) as to whether or not the spot giving luminance L is a hot spot for all the solar panels 21 detected in thermal image 22 has been completed (step S37) In the case of YES), the hot spot detection unit 551 further confirms whether or not another thermal image 22 including the solar panel 21 whose inspection has not been completed remains (step S38).

  When another thermal image 22 including the solar panel 21 to be inspected remains (in the case of NO at step S38), the hot spot detection unit 551 performs step S31 for each solar panel 21 appearing in the remaining thermal image 22. The processing step of step S37 is performed.

  On the other hand, when there is no other thermal image 22 showing the solar panel 21 to be inspected, that is, it is judged whether or not it is a hotspot for all the inspection objects (solar panel 21) (step S38) (YES), the hot spot detection unit 551 completes the hot spot detection process (RETURN).

  In the above-described hot spot detection step (steps S31 to S38), as shown in the above equation (1), the threshold value t for hot spot presence / absence determination is determined in consideration of the average value μ and the standard deviation value σ. The reason is that the temperature may not be high due to the hot spot due to the influence of the temperature, the amount of solar radiation, the amount of power generation, etc., but the temperature may be simply high as a whole. This is to detect hotspots more accurately by eliminating the influence of quantity.

  In the above-described hot spot detection step, in the luminance standard deviation calculation step (step S32), an area used to calculate the standard deviation value σ (hereinafter referred to as a “standard deviation value calculation area”) is obtained. The standard deviation value σ is determined based on the luminance obtained for the entire range in the solar panel 21, but the standard deviation calculation region is not necessarily the solar panel 21. It does not have to be the entire range. A smaller area may be set in one solar panel 21, and the standard deviation value σ may be determined based on the luminance in the area.

  The advantage of determining the standard deviation value σ based on the luminance in a partial region in the solar panel 21 is compared to the case of determining the standard deviation value σ based on the luminance in the entire range in one solar panel 21. The variation of the standard deviation value σ between the solar panels 21 can be suppressed small, and the detection accuracy of the hot spot can be further enhanced.

  More specifically, when there is a hot spot in the solar panel 21, a locally high-brightness part is included, so the standard deviation value σ becomes relatively large while the hot spot does not exist. The standard deviation value σ of the solar panel 21 is relatively small. Therefore, when the standard deviation value calculation area is the entire range of the solar panel 21, the threshold t of the solar panel 21 in which the hot spot is present becomes larger than the threshold t of the healthy solar panel 21.

  On the other hand, when the standard deviation value calculation area is a partial area in the solar panel 21, if a place where there is no hot spot is set as the standard deviation value calculation area, locally high brightness parts are included. Since the solar panels 21 do not have the hot spots, there is an advantage that the standard deviation value σ of each of the solar panels 21 can be made approximately the same (variation within a predetermined range) regardless of whether the solar panels 21 have hot spots or not.

  In the place where there is no hot spot, the temperature is almost the same as the temperature around, so the variation of the luminance L is small. Therefore, a partial region having a predetermined area is set in the solar panel 21, and the standard deviation value σ in the set partial region is calculated, and the luminance of the partial region giving the smallest standard deviation value σ is calculated. The calculated standard deviation value σ may be used to calculate the threshold value t. The partial area giving the smallest standard deviation value σ can be specified by, for example, calculating the standard deviation value σ each time while shifting a partial area in the area within the solar panel 21 in pixel units.

  FIG. 7 is a processing flow diagram showing a flow of processing of the non-ordinary part detection step (step S4: FIG. 4).

  The non-ordinary part detection step (steps S41 to S45) includes, for example, a line segment detection step (step S41) and a normal / non-ordinary part determination step (steps S42 to S44).

  When the non-ordinary part detection step is started (ENTER), first, the non-ordinary part detection unit 552 detects a line segment vector from the area within the range of the detected solar panel 21 of the visible image 23 (step S41) Then, it is determined whether or not there is a line segment vector in a (oblique) direction which is inclined with respect to the vertical and horizontal line segments forming the outline of the solar panel 21 (step S42).

  When the non-ordinary part detection unit 552 detects a line segment vector in the oblique direction in the solar panel 21 shown in the visible image 23 (in the case of YES in step S42), it is determined that the non-normal part is present in the solar panel 21 (Step S43). On the other hand, when the non-ordinary part detection unit 552 does not detect a line segment vector in the diagonal direction in the solar panel 21 (in the case of NO in step S42), it is determined that the non-normal part is not present in the solar panel 21 ( Step S44).

  When all the solar panels 21 detected in the visible image 23 have the solar panels 21 for which the determination (step S43 or step S44) of the non-ordinary part is not completed remains (NO in step S45) Line segment detection step (step S41) and a normal / non-normal point determination step (steps S42 to S44) are performed on the solar panel 21 for which the determination of the presence / absence of the non-normal point is not completed.

  On the other hand, if the determination of the presence / absence of the non-ordinary part has been completed for all the solar panels 21 detected in the visible image 23 (YES in step S45), the non-ordinary part detection step is completed. (RETURN).

  FIG. 8 is a processing flow diagram showing a flow of processing of a first inspection result image generation step (step S5: FIG. 4) in the first panel inspection procedure (FIG. 4).

  The first inspection result image generation step (steps S51 to S57) includes, for example, an inspection result acquisition step (step S51), an abnormal solar panel emphasizing step (step S53), and an image position information acquisition step (step S54). And a position matching step (step S56).

  When the first inspection result image generation process is started (ENTER), first, the first inspection result image generation unit 571 (FIG. 3) determines the presence / absence of abnormality for each of the solar panels 21 (FIG. 2) That is, the result of the abnormality detection step (step S3 and step S4: FIG. 4) is acquired (step S51).

  When the determination result of the abnormality presence or absence is abnormality existence (in the case of YES in step S52), the first inspection result image generation unit 571 highlights the solar panel 21 with the abnormality existence (step S53). For highlighting of the solar panel 21, various types such as changing the thickness of the outline of the solar panel 21, changing the color of the outline of the solar panel 21, or marking the range of the solar panel 21 A technique can be adopted. Also, the method of highlighting may be changed according to the content of the abnormality.

  When the solar panel 21 with abnormality is highlighted, subsequently, the first inspection result image generation unit 571 acquires position information of the inspection image such as the thermal image 22 giving the judgment result of the abnormality (step S54). Acquisition of the positional information on the inspection image such as the thermal image 22 is performed by acquiring the positional information acquired by the positional information acquisition means 56.

  When the first inspection result image generation unit 571 acquires the position information of the inspection image such as the thermal image 22, subsequently, the map data 11 or the like that is the basis of the background image such as the map image used to display the inspection result The position based on the position information of the inspection image is associated with the position on the map indicated by the map data 11, and the inspection image is positioned on the map (step S56).

  When the first inspection result image generation unit 571 positions the inspection image such as the thermal image 22 on the map indicated by the map data 11, the generation of the inspection result image is completed, and the generation of the inspection result image is completed (step S57), the first inspection result image generation unit 571 completes the first inspection result image generation process (RETURN).

  On the other hand, in the first inspection result image generation process, when the determination result of the abnormality presence / absence is no abnormality (in the case of NO at step S52), the first inspection result image generation unit 571 skips step S53 and proceeds to step S54. Perform the subsequent processing steps.

  According to the panel inspection apparatus 50A, the first panel inspection procedure, and the first panel inspection PG 10A, each of the inspection images such as the thermal image 22 taken from above the solar panel 21 to be inspected is automatically used for each Since the range of the solar panel 21 is detected and it is detected whether or not there is a place where an abnormality occurs in each of the detected solar panels 21, the burden on the inspector required for the inspection of the solar panel 21 can be reduced. it can.

  Further, in the panel inspection device 50A, the first panel inspection procedure and the first panel inspection PG 10A, the outline of each solar panel 21 is detected from the visible image 23 used when detecting a crack or the like, and the detected outline is By making it possible to project onto the thermal image 22 including the same range as the visible image 23, the range of each solar panel 21 can be detected even when the thermal image 22 alone can not detect the outline of each solar panel 21 sufficiently. It is possible to automatically detect an abnormal point appearing on each solar panel 21 more accurately than in the prior art.

  Furthermore, in the panel inspection apparatus 50A, the first panel inspection procedure, and the first panel inspection PG 10A, it is possible to detect whether there is an unusual place in the solar panel 21 or not, thereby making it possible to Such anomalies other than hot spots can be detected. In addition, since it is possible to detect an unusual part other than the part where damage such as a crack has occurred, it is also possible to detect an abnormal reserve state. Therefore, in addition to the currently occurring abnormality, it is possible to detect an event that may become an abnormality in the future, and prevent the occurrence of an abnormality in the future.

  According to the panel inspection apparatus 50A, the first panel inspection procedure and the first panel inspection PG 10A, the thermal image 22 etc. can be received from the UAS 30 and the panel inspection can be performed in real time by using the communication means 53. The inspection of the solar panel 21 can be started even if the UAS 30 is in the non-landing state (during flight continuation) after the start of photographing. In addition, if an abnormal part is detected on site, repair can be started earlier.

  According to panel inspection apparatus 50A, the first panel inspection procedure, and the first panel inspection PG 10A, threshold value t for hot spot presence / absence determination is determined in consideration of average value μ and standard deviation value σ, Hot spots can be detected more accurately by eliminating the effects of temperature, solar radiation, and power generation.

  In addition, as a standard deviation value σ to be used for the threshold value t, a smaller area is set in one solar panel 21 and a standard deviation value σ giving a minimum value among the standard deviation values σ obtained based on the luminance in the area Can be used to minimize the variation in the standard deviation value σ between the solar panels 21 and to further enhance the detection accuracy of the hot spot.

  According to the panel inspection apparatus 50A, the first panel inspection procedure, and the first panel inspection PG 10A, moving images such as the heat moving image 22 can be used as the inspection image, so that the moving image shooting has a frame rate higher than that of the still image. The solar panel 21 to be inspected can be photographed with the flying speed of the UAS 30 further increased. Therefore, it is possible to shoot a wider range or shoot a predetermined range in a shorter time by one aerial shooting for acquiring an inspection image.

Second Embodiment
FIG. 9 is a functional block diagram showing a functional configuration of a panel inspection apparatus 50B which is an example of a panel inspection apparatus according to a second embodiment of the present invention.

  The panel inspection apparatus 50B (FIG. 9) further includes an image combining unit 62 with respect to the panel inspection apparatus 50A (FIG. 3), and a display processing unit 57A including a first inspection result image generation unit 571. The second embodiment differs in that the display processing means 57B including the second inspection result image generation unit 572 is provided, but there is substantially no difference in other points. So, in this embodiment, about the component which is not substantially different from panel inspection device 50A, the same numerals are attached and the overlapping explanation is omitted.

  The panel inspection apparatus 50B, for example, includes an input unit 51, an output unit 52, a communication unit 53, an inspection range detection unit 54A, an abnormality detection unit 55, a position information acquisition unit 56, and a second inspection result image generation. A display processing unit 57B including a unit 572, a storage unit 58, a control unit 59, a still image extraction unit 61, and an image combining unit 62 are configured.

  The second inspection result image generation unit 572 in the display processing unit 57B inspects the combined thermal image (hereinafter, referred to as “combined thermal image”) 32 (FIG. 10). Since the synthetic thermal image 32 is a thermal image showing the temperature distribution of all the solar panels 21 to be inspected, the display processing means 57B uses the synthetic thermal image 32 to generate an inspection result image indicating the solar arrangement and the inspection result.

  The image combining unit 62 has an image combining function of the thermal image 22 and the visible image 23, and has a plurality of thermal images 22 or visible images 23 obtained by capturing different regions (a part of the overlapping region is included). By superimposing in overlapping regions, a thermal image (synthetic thermal image 32) or a visible image (hereinafter, “synthetic visible image”) 33 (FIG. 10) covering a wider area is obtained.

  The control means 59 in the panel inspection apparatus 50B is provided with a control function of the component added to the panel inspection apparatus 50A. That is, the control means 59 in the panel inspection apparatus 50B exchanges data with the display processing means 57B including the second inspection result image generation unit 572 and the image combining means 62, and further has a function of controlling these. There is.

  The panel inspection apparatus 50B acquires an inspection image such as an individual thermal image 22 or the like obtained by photographing each divided area with respect to divided areas into which a predetermined area including the entire range of the solar panel 21 to be inspected is divided into a plurality. Therefore, a composite image such as a composite thermal image 32 covering the entire range of the solar panel 21 to be inspected is obtained, and using this composite image, a first inspection range detection step (step S2: FIG. 4), a hot spot The detection step (step S3: FIG. 4), the non-ordinary part detection step (step S4: FIG. 4), and the second inspection result image generation step (step S6: FIG. 4) are performed.

  In panel inspection device 50B configured in this manner, inspection is performed using the synthetic thermal image 32 in which all the solar panels 21 to be inspected by the second inspection result image generation unit 572 appear as a panel layout diagram showing the arrangement of the solar panels 21. The resulting image can be generated. Therefore, even when there is no map data 11, it is possible to generate an inspection result image indicating a part where an abnormality is detected and display the inspection result on the display means.

  Panel inspection device 50B is not limited to the configuration illustrated in FIG. For example, the panel inspection apparatus 50B may not necessarily include the still image extraction unit 61 as in the panel inspection apparatus 50A. Further, when the image combining process is performed by the external device and the combined thermal image 32 (FIG. 10) or the combined visible image 33 (FIG. 10) is given, the image combining process in the panel inspection device 50B is unnecessary. If so, the panel inspection apparatus 50B does not necessarily have to include the image combining means 62.

  Furthermore, when it is not necessary to display the solar panel 21 in which the abnormality is detected on the map using the map data 11, the display processing means 57B need not necessarily include the first inspection result image generation unit 571. Also, it is not necessary to have the position information acquisition means 56.

  Next, as a panel inspection method according to the second embodiment of the present invention, a second panel inspection procedure performed by the panel inspection device 50B will be described.

  The second panel inspection procedure is different from the first panel inspection procedure in that the second inspection result image generation step (step S6) is included instead of the first inspection result image generation step (step S5). Although different, the other processing steps are not substantially different.

  FIG. 10 is a process flow diagram showing a process flow of a second inspection result image generation step (step S6: FIG. 4) in the second panel inspection procedure.

  The second inspection result image generation step (step S61, steps S51 to S53 and step S57) is similar to the first inspection result image generation step (FIG. 8: step S51 to step S57). The difference is that the step of acquiring the position information of the inspection image such as the thermal image 22 (steps S54 and S55) and the position associating step (step S56) are omitted while further including the step (step S61). There is no substantial difference in other points. Therefore, the same step numbers are assigned to processing steps that are not substantially different from the first inspection result image generation step, and redundant description will be omitted.

  The second inspection result image generating step includes, for example, a panel layout diagram image acquiring step (step S61) for acquiring the thermosynthetic image 32 which is a composite image of the thermal image 22; an inspection result acquiring step (step S51); And a solar panel emphasizing step (step S53).

  When the second inspection result image generation process is started (ENTER), first, the second inspection result image generation unit 572 (FIG. 9) acquires the thermally synthesized image 32 as a panel layout plan image (step S61). When the second inspection result image generation unit 572 acquires the thermosynthesis image 32, subsequently, if the second inspection result image generation unit 572 determines that there is an abnormal solar panel 21 in the thermosynthesis image 32 (step S52). In the case of YES), the solar panel 21 with abnormality is highlighted (step S53), and the generation of the inspection result image is completed (step S57).

  When the second inspection result image generation unit 572 completes the generation of the inspection result image (step S57), the second inspection result image generation unit 572 completes the second inspection result image generation process (RETURN).

  When the panel inspection apparatus 50B (FIG. 9) includes the first inspection result image generation unit 571 in the display processing means 57B, whether the first inspection result image generation step (FIG. 8) is selected or not It is possible to receive a selection as to whether to select the second examination result image generation process, and to execute one examination result image generation process according to the received selection content.

  According to the panel inspection device 50B, the second panel inspection procedure and the second panel inspection PG 10B, the same effects as the panel inspection device 50A, the first panel inspection procedure and the first panel inspection PG 10A can be obtained In addition to the above, the second inspection result image generation unit 572 generates an inspection result image by using the synthetic thermal image 32 as a panel layout diagram showing the layout of the solar panel 21 without using the map data 11 etc. Even if there is no map data 11 grade | etc., The test result of the solar panel 21 can be displayed so that the location with abnormality can be grasped | ascertained.

  Further, in the panel inspection apparatus 50B, when the display processing means 57B includes the first inspection result image generation unit 571, since the map data 11 etc. can be used, the map data 11 etc. show according to the selection of the display method. The inspection result can be displayed by selecting whether to display the synthetic thermal image 32 as an example of the thermal image 22 on the image or to display the synthetic thermal image 32 without the image indicated by the map data 11 or the like.

Third Embodiment
FIG. 11 is a functional block diagram showing a functional configuration of a panel inspection apparatus 50C which is an example of a panel inspection apparatus according to a third embodiment of the present invention.

  Panel inspection apparatus 50C (FIG. 11) is different from panel inspection apparatus 50A (FIG. 3) in that inspection area detection means 54C and display processing means 57C are provided instead of inspection area detection means 54A and display processing means 57A. Although different, they are not substantially different in other points. Then, about the component which is not substantially different from panel inspection device 50A, the same numerals are attached and the overlapping explanation is omitted.

  The panel inspection apparatus 50C includes, for example, an inspection range detection unit 54C including an input unit 51, an output unit 52, a communication unit 53, and a third inspection range detection unit 543, an abnormality detection unit 55, a position information acquisition unit 56, a display processing means 57C having a third examination result image generation unit 573, a storage means 58, a control means 59, and a still image extraction means 61.

  The inspection range detection unit 54C further includes, for example, a third inspection range detection unit 543 in addition to the first inspection range detection unit 541 and the second inspection range detection unit 542.

  The third inspection range detection unit 543 has a function of projecting an inspection image such as the thermal image 22 used in the abnormality detection step (FIG. 4) onto the panel layout, and each solar panel based on the outline of the panel layout. 21 has a function of detecting the range of 21 and after projecting an inspection image such as the thermal image 22 used in the abnormality detection step onto the panel layout diagram, the range of each solar panel 21 based on the outline of the panel layout diagram To detect

  The display processing unit 57C includes a third inspection result image generation unit 573 instead of the first inspection result image generation unit 571 with respect to the display processing unit 57A.

  The third inspection result image generation unit 573 reads out the panel layout diagram data 15 (FIG. 11) that is the basis of the panel layout diagram image (background image) used to display the inspection result, and acquires the panel layout diagram image And a function for specifying which of the solar panels 21 in the panel layout diagram image is an inspection image such as the thermal image 22 used in the abnormality detection step and each of the solar panels 21 in the inspection image And the function of reflecting the determination result of the abnormality presence or absence of the solar panel 21 on the panel layout drawing image.

  The third inspection result image generation unit 573 sequentially acquires inspection images such as the thermal image 22 used in the abnormality detection step, and sequentially determines the determination result of the presence or absence of the solar panel 21 captured in the acquired image. To generate an inspection result image.

  The control means 59 in the panel inspection apparatus 50C is added with a control function of the component added to the panel inspection apparatus 50A. That is, the control means 59 in the panel inspection apparatus 50C further has a function of mutually exchanging data with the inspection range detection means 54C and the display processing means 57C and controlling them.

  Panel inspection device 50C is not limited to the configuration illustrated in FIG. For example, like the panel inspection device 50A, the panel inspection device 50C may not necessarily include the still image extraction unit 61. In addition, if the display processing means 57 in the panel inspection apparatus 50C includes at least the third inspection result image generation unit 573, the first inspection result image generation unit 571 and the second inspection result image generation unit 572 are further added. The configuration may be provided.

  Next, as a panel inspection method according to the third embodiment of the present invention, a third panel inspection procedure performed by the panel inspection device 50C will be described. In the description of the third panel inspection procedure, processing steps that are not substantially different from the above-described first and second panel inspection procedures are given the same step numbers and redundant description will be omitted.

  The third panel inspection procedure (FIG. 4) is the first inspection range detection step (step S2) and the first inspection result image generation step (step S5) with respect to the first panel inspection procedure (FIG. 4). Instead of the second inspection range detection step (step S7) and the third inspection result image generation step (step S8), the other processing steps are substantially the same.

  FIG. 12 is a processing flow diagram showing a flow of processing of the second inspection range detection step (step S7: FIG. 4).

  The second inspection range detection step further includes an image projection step (step S71) and a panel range detection step (step S72) with respect to the first inspection range detection step (FIG. 5), while the panel range detection step Although the difference is that step S26) is omitted, the other points are not substantially different.

  When the second inspection range detection step is started (ENTER), the inspection range detection means 54C performs the processing step of step S21 to step S25 in the same manner as the first inspection range detection step, and the subsequent abnormality detection step An inspection image such as the thermal image 22 used for (FIG. 4) is prepared. Subsequently, the third inspection range detection unit 543 projects the prepared inspection image for projection onto the panel layout diagram (step S71).

  The projection of the inspection image to be used in the abnormality detection step (FIG. 4) on the panel layout view is carried out, for example, referring to the inspection image to be used in the abnormality detection step (FIG. 4) with reference to the plane position information of the inspection image. Temporarily determine on the panel layout diagram, search for the feature quantity extracted from the inspection image in the panel layout diagram, and finally determine the closest position where the feature quantity matches from the tentatively decided position. Determined as a position in the layout and projected.

When the projection onto the panel layout diagram is completed, the third inspection range detection unit 543 identifies the range of each solar panel 21 based on the outline of each solar panel 21 in the panel layout diagram, and It is detected as a range (step S72). When the third inspection range detection unit 543 detects each range of the solar panels 21, the panel identification of each solar panel 21 included in the panel layout view data 15 (FIG. 11) is detected for each of the detected solar panels 21. Information is added and associated with the image identification information of the inspection image to complete the second inspection range detection step (RETURN).

  FIG. 13 is a processing flow diagram showing a flow of processing of the third inspection result image generation step (step S8: FIG. 4).

  The third inspection result image generation step (steps S81 to S83, S52, S53, S84 and S57) is the same as the first inspection result image generation step (FIG. 8: steps S51 to S57). While further including a panel layout drawing image acquisition step (step S81), a panel identification step (step S82), and an abnormality presence / absence determination result acquisition step (step S83), position information of the inspection image such as the thermal image 22 The difference is that the obtaining step (step S54 and step S55) and the position matching step (step S56) are omitted, but the other points are not substantially different.

  When the third inspection result image generation process is started (ENTER), first, the third inspection result image generation unit 573 (FIG. 11) uses the panel layout image (background image) to be used to display the inspection results. The panel layout diagram data 15 (FIG. 11) to be obtained are read out, and a panel layout diagram image is acquired (step S81).

  When the third inspection result image generation unit 573 acquires a panel layout drawing image, subsequently, the third inspection result image generation unit 573 refers to the identification information of the inspection image and the panel identification information to detect an abnormality. It is specified which solar panel 21 in the panel layout drawing image is each of the solar panels 21 shown in the inspection image used in the process (step S3 and step S4: FIG. 4) (step S82).

  When the third inspection result image generation unit 573 identifies the solar panel 21 (FIG. 2) in the panel layout image, the determination result of the identified presence or absence of the identified solar panel 21 is acquired (step S83). If there is a panel 21 (YES in step S52), the abnormal solar panel 21 is highlighted (step S53).

  When the third inspection result image generation unit 573 reflects the judgment result of the abnormality presence / absence for all the solar panels 21 specified, the third inspection result image generation unit 573 reflects the judgment result of the abnormality presence / absence of the solar panel 21. If an uncompleted image remains (in the case of NO at step S84), the remaining images are read out, and all the judgment results of abnormality presence or absence are reflected on the solar panel 21 (step S82, step S83, step S52 and step S53) .

  On the other hand, when the third inspection result image generation unit 573 reflects all the inspection results from all the images giving the determination result of the presence or absence of abnormality of the solar panel 21 (in the case of YES in step S84), third inspection result image generation The unit 573 completes the generation of the inspection result image (step S57), and completes the third inspection result image generation process (RETURN).

  According to the panel inspection device 50C, the third panel inspection procedure and the third panel inspection PG 10C, the same effect as the panel inspection device 50A, the first panel inspection procedure and the first panel inspection PG 10A can be obtained In addition to the above, since the inspection result can be displayed on the panel layout drawing indicated by the panel layout drawing data 15, the inspection result can be provided in an easy-to-understand manner for the user. Further, by forming data of the inspection result image into a layer structure and layering each information to be displayed, desired information to be displayed can be selected and displayed on the panel layout diagram.

  For example, by setting the hotspot detection result based on the thermal image 22 and the non-ordinary detection result based on the visible image 23 as separate layers, the hotspot detection result and the non-ordinary detection result simultaneously on the panel layout diagram or It can be individually displayed on the panel layout drawing.

  As described above, according to the panel inspection apparatus 50 (50A to 50C), the first to third panel inspection procedures, and the panel inspection PG 10 (10A to 10C), the thermal image 22 taken from above of the solar panel 21 to be inspected. Because the range of each solar panel 21 is automatically detected using the inspection image such as, it is detected whether or not there is a location where an abnormality has occurred in each solar panel 21 detected, so the solar panel The burden on the inspector required for the 21 tests can be reduced.

  Further, in the panel inspection apparatus 50, the first to third panel inspection procedures and the panel inspection PG10, the outline of each solar panel 21 is detected from the visible image 23 used when detecting a crack or the like, and the detected outline is By making it possible to project onto the thermal image 22 including the same range as the visible image 23, the range of each solar panel 21 can be detected even when the thermal image 22 alone can not detect the outline of each solar panel 21 sufficiently. It is possible to automatically detect an abnormal point appearing on each solar panel 21 more accurately than in the prior art.

  Furthermore, in the panel inspection apparatus 50, the first to third panel inspection procedures, and the panel inspection PG 10, it is possible to detect whether or not there is an unusual place in the solar panel 21, such as damage such as a crack or the like. Abnormalities other than hot spots can be detected. In addition, since it is possible to detect an unusual part other than the part where damage such as a crack has occurred, it is also possible to detect an abnormal reserve state. Therefore, in addition to the currently occurring abnormality, it is possible to detect an event that may become an abnormality in the future, and prevent the occurrence of an abnormality in the future.

  According to the panel inspection apparatus 50, the first to third panel inspection procedures, and the panel inspection PG10, the thermal image 22 and the like can be received from the UAS 30 and the panel inspection can be performed in real time by using the communication means 53. The inspection of the solar panel 21 can be started even under the situation where the UAS 30 continues the flight after the start of photographing. In addition, if an abnormal part is detected on site, repair can be started earlier.

  According to panel inspection device 50, the first to third panel inspection procedures, and panel inspection PG10, threshold value t for hot spot presence / absence determination is determined in consideration of average value μ and standard deviation value σ, Hot spots can be detected more accurately by eliminating the effects of temperature, solar radiation, and power generation.

  In addition, as a standard deviation value σ to be used for the threshold value t, a smaller area is set in one solar panel 21 and a standard deviation value σ giving a minimum value among the standard deviation values σ obtained based on the luminance in the area Can be used to minimize the variation in the standard deviation value σ between the solar panels 21 and to further enhance the detection accuracy of the hot spot.

  Further, according to the panel inspection apparatus 50, the first to third panel inspection procedures, and the panel inspection PG10, the moving image such as the thermal moving image 22 can be used as the inspection image, whereby the frame rate is higher than that of the still image. A moving image can be taken, and the solar panel 21 to be inspected can be taken with the flying speed of the UAS 30 further increased. Therefore, it is possible to shoot a wider range or shoot a predetermined range in a shorter time by one aerial shooting for acquiring an inspection image.

  According to the panel inspection procedure performed by the panel inspection device 50 including the second inspection result image generation unit 572, the panel inspection procedure performed by the panel inspection device 50, and the panel inspection PG 10 according to the panel inspection PG 10, the inspection result An image can be generated to display the test results.

  The panel inspection apparatus 50 including the third inspection range detection unit 543 and the third inspection result image generation unit 573, the panel inspection procedure performed by the panel inspection device 50, and the panel inspection PG10 according to the panel inspection PG10 Since the inspection result can be displayed on the layout drawing, the inspection result can be provided more easily for the user. Further, by forming data of the inspection result image into a layer structure and layering each information to be displayed, desired information to be displayed can be selected and displayed on the panel layout diagram.

  The method described in each of the above-described embodiments may be applied to various devices as a program that can be executed by a computer, for example, written to a storage medium such as a magnetic disk, an optical disc, or a semiconductor memory, or transmitted via a communication medium It can also be applied to various devices. A computer for realizing the present apparatus reads a program recorded in a storage medium and executes the above-mentioned processing by controlling the operation by this program.

  Furthermore, the present invention is not limited to the above-described embodiment as it is, and can be implemented in various forms other than the above-described embodiment at the implementation stage, without departing from the scope of the invention. Various omissions, additions, replacements and changes can be made. These embodiments and modifications thereof are included in the scope and the gist of the invention, and are included in the invention described in the claims and the equivalent scope thereof.

1 computer 2 CPU
3 main storage device 4 auxiliary storage device 5 input device 6 output device 7 communication means 8 calculation server 10 (10A to 10C) panel inspection PG (program)
11 map data 13 ortho image 15 panel layout data 20 solar power station 21 solar panel 22 thermal image (thermal still image or thermal video)
23 Visible image (visible still image or visible video)
24 High Resolution Visible Image (High Resolution Visible Still Image or High Resolution Visible Video)
25 in-house building 26 thermography 27 visible camera 28 integrated camera 29 high resolution visible camera 30 unmanned aerial vehicle system (UAS)
50 (50A to 50C) Panel inspection device 51 Input unit 52 Output unit 53 Communication unit 54A, 54C Inspection range detection unit 541 First inspection range detection unit 542 Second inspection range detection unit 543 Third inspection range detection unit 55 Abnormality detection means 551 hot spot detection unit (first abnormality detection unit)
552 Unusual location detector (second anomaly detector)
56 Position information acquisition means 57A, 57B, 57C Display processing means 571 First examination result image generation unit 572 Second examination result image generation unit 573 Third examination result image generation unit 58 Storage means 59 Control means 61 Still image extraction Means 62 image synthesizing means R flight path

Claims (20)

A first inspection range detection unit configured to detect an area of each of the solar panels to be inspected by detecting an outline of the solar panels based on a temperature distribution appearing in a thermal image taken from above the solar panels to be inspected; A second inspection range detection unit for acquiring a visible image photographed from the image and detecting an area of each of the solar panels to be inspected from the acquired visible image; detecting an outline of each of the solar panels from the visible image; Inspection range detection means for projecting the detected outline on the thermal image to detect the range of each of the solar panels to be inspected;
Setting a plurality of partial regions having a predetermined area in the thermal image of each of the solar panels detected by the inspection range detection means, and calculating the luminance L and the standard deviation value σ of the luminance L for each partial region; A threshold value t defined by the following equation [Equation 1] based on the average value μ of the luminance L and the standard deviation value σ in the partial area where the standard deviation value σ is the smallest is a threshold value t for determining the temperature abnormal point presence or absence When it is determined that there is a portion showing temperature abnormality in the whole of the thermal image by comparison with the threshold value t, or it is determined that the visible image has a line segment vector in a direction inclined with respect to the outline. And an abnormality detection unit having an abnormality detection unit that determines that the solar panel is abnormal and detects an abnormal solar panel.
Position information acquisition means for acquiring position information of a shooting position of an image used when the inspection range detection means detects the range of each of the solar panels to be inspected;
The map information including the position information indicating the position of the solar panel to be inspected is read out, the range of each of the solar panels to be inspected detected by the inspection range detecting means, and the image acquired by the position information acquiring means The inspection indicated on the map indicated by the map information based on the position information of the photographing position and the abnormal solar panel when the abnormality detection means detects an abnormal solar panel and the read map information read out And a display processing means for generating an inspection result image including inspection result information indicating at least the abnormal solar panel among the solar panels to be inspected.
t = μ + nσ (n is any positive number) The layout of the solar panels including all of the solar panels to be inspected provided to the display processing means is an ortho image in which the solar panels including all of the solar panels to be inspected are photographed. The inspection apparatus of the solar panel as described in. The display processing means is provided with the range of each of the solar panels to be inspected detected by the inspection range detection means, and the abnormal solar panel when the abnormal detection means detects an abnormal solar panel. It is projected on the layout of the solar panel to be inspected obtained by reading out data of the solar panel layout including all of the solar panels to be inspected, and at least the inspection on the layout of the solar panel to be inspected A display according to claim 1 or 2, characterized in that it is configured to generate the inspection result image to which a display showing each solar panel within the range of and a display showing the abnormal solar panel is added. Solar panel inspection equipment. For each divided area obtained by dividing a predetermined area including the entire range of the solar panel to be inspected into a plurality of divided areas, when each divided image obtained by photographing each divided area is acquired, based on the acquired divided image, The inspection of a solar panel according to any one of claims 1 to 3, further comprising image combining means for obtaining a single composite image in which at least the entire range of the solar panel to be inspected is photographed. apparatus. 5. The solar panel according to claim 4, wherein the layout diagram of the solar panel including all of the solar panels to be inspected provided to the display processing means is one composite image obtained by the image combining means. Inspection device. The display processing means includes, in addition to the layout of the first solar panel based on the data of the layout of the solar panel given in advance, one composite image obtained by the image combining means as the second solar When obtained as a layout drawing of the panel,
Arrangement of the first solar panel given an object to which a display showing each solar panel within the scope of the inspection and a display showing the abnormal solar panel are given when generating the inspection result image A function of selecting one of the layout of the first solar panel and the layout of the second solar panel based on a command to select one of the layout and the second solar panel. The inspection apparatus of the solar panel of Claim 4 characterized by having. The display processing means matches the feature amount appearing in the layout drawing of the solar panel with the feature amount appearing in an image used when detecting the range of each of the solar panels to be examined. The range of each of the inspected solar panels detected by the range detecting means and the abnormal solar panel when the abnormality detecting means detects an abnormal solar panel are projected onto the layout diagram of the inspected solar panel The inspection apparatus for a solar panel according to any one of claims 3 to 6, which is configured as follows. Position information of the photographing position of the image used when the position information acquisition means detects the range of each of the solar panels to be inspected, and a predetermined position included in a solar panel layout including all of the solar panels to be inspected If the location information of is obtained from the location information determined by GPS,
When the display processing means detects the range of each of the solar panels to be inspected detected by the inspection range detection means and the abnormality detection means detects an abnormal solar panel based on the position information measured by the GPS. 7. An apparatus according to any one of claims 3 to 6, wherein the apparatus is configured to project the abnormal solar panel onto the layout of the solar panel to be inspected. The image processing apparatus further comprises still image extraction means for extracting a predetermined frame from a moving image obtained by imaging all of the inspected solar panels as an image used for detecting the range of the inspected solar panels and generating a still image. The inspection apparatus of the solar panel in any one of the Claims 1-8 characterized by the above-mentioned. A first inspection range detection unit configured to detect an area of each of the solar panels to be inspected by detecting an outline of the solar panels based on a temperature distribution appearing in a thermal image taken from above the solar panels to be inspected; A second inspection range detection unit for acquiring a visible image photographed from the image and detecting an area of each of the solar panels to be inspected from the acquired visible image; detecting an outline of each of the solar panels from the visible image; the inspection range detecting means for detecting a range of solar panels each being examined by projecting the detected contour on the thermal image, the inspection range detection means of the solar panel respectively detected the thermal image to a predetermined A plurality of partial areas having an area are set, and luminance L for each partial area and standard deviation value σ of luminance L are calculated, and the standard deviation value σ is the smallest in the partial area with the smallest value A threshold value t defined by the following equation [Equation 1] based on the average value μ of the luminance L and the standard deviation value σ is used as the threshold value t for determining the presence or absence of a temperature abnormal point. When it is determined that there is a portion showing temperature abnormality in the whole, or when it is determined that the visible image has a line segment vector in a direction inclined to the outline, the solar panel is determined to be abnormality, Position detecting means having an abnormality detecting unit for detecting a large number of solar panels, and position information acquiring means for acquiring information of a photographing position of an image used when the inspection range detecting means detects the range of each solar panel to be inspected And information on the range of each of the solar panels to be inspected detected by the inspection area detecting unit, the information on the photographing position of the image acquired by the position information acquiring unit, and the solar of the abnormality detecting unit being abnormal. When an array is detected, based on the abnormal solar panel, an inspection result image including inspection result information indicating the abnormal solar panel among the solar panels to be inspected distinguished from other solar panels is generated A solar panel inspection procedure using a solar panel inspection device comprising display processing means, the solar panel inspection procedure comprising
The first inspection range detection step in which the inspection range detection means detects the range of each of the solar panels to be inspected based on the temperature distribution appearing in the thermal image, and the second inspection range detection unit acquires A second inspection range detection step of detecting the range of the solar panel to be inspected from a visible image;
When the abnormality detection means determines that there is a portion showing temperature abnormality from the range of each of the solar panels detected in the inspection range detection step, the solar panel is determined as abnormal and an abnormal solar panel is detected An anomaly detection step comprising an anomaly detection step;
A position information acquisition step of acquiring information of a photographing position of an image used when detecting the range of each of the solar panels in the inspection range detection step;
The display processing means includes the range of each of the solar panels detected in the inspection range detecting step, the information of the photographing position of the image acquired in the position information acquiring step, and the abnormal solar panel in the abnormality detecting step Inspection result image including inspection result information showing the abnormal solar panel among the solar panels to be inspected distinguished from other solar panels based on the detected abnormal solar panel when A display processing step of generating a solar panel inspection method.
t = μ + nσ (n is any positive number) The layout of the solar panels including all of the solar panels to be inspected provided to the display processing means is an ortho image in which the solar panels including all of the solar panels to be inspected are photographed. The inspection method of the solar panel as described in. The display processing step provides the range of each of the solar panels detected in the inspection range detection step, and the abnormal solar panel detected when an abnormal solar panel is detected in the abnormality detection step. Projected on the layout of the solar panel to be inspected obtained by reading out data of the solar panel layout including all of the solar panels to be inspected, and at least the layout of the solar panel to be inspected; 12. The method according to claim 10, further comprising the step of generating the inspection result image to which a display indicating each solar panel within the inspection range and a display indicating the abnormal solar panel are added. Inspection method of solar panel. The solar panel inspection procedure is acquired when each divided image obtained by photographing each divided region is acquired with respect to divided regions into which a predetermined region including the entire range of the solar panel to be inspected is divided into a plurality An image combining means for obtaining a single composite image in which at least the entire range of the solar panel to be inspected is photographed based on the divided images, and a single composition in which the entire range of the solar panel to be inspected is photographed Further comprising an image combining step for obtaining an image;
The layout of the solar panels to be inspected, to which the display showing each solar panel within the scope of the inspection and the display showing the abnormal solar panel are added in the display processing step, is the image combining means The inspection method of a solar panel according to claim 12, characterized in that it is a single composite image obtained. In addition to the layout of the first solar panel based on the data of the layout of the solar panel given in advance by the display processing means, in addition to the layout of the first solar panel, When obtained as a layout drawing of the panel,
The solar panel inspection procedure is acquired when each divided image obtained by photographing each divided region is acquired with respect to divided regions into which a predetermined region including the entire range of the solar panel to be inspected is divided into a plurality The image combining means for obtaining a single composite image in which at least the entire area of the solar panel to be inspected is photographed based on the divided images is a sheet in which the entire area of the solar panel to be inspected is photographed Further comprising an image combining step of obtaining a combined image;
The display processing step includes: a layout of the first solar panel selected based on a command to select one of the layout of the first solar panel and the layout of the second solar panel; An image including at least a display indicating each solar panel within the inspection range and a display indicating the abnormal solar panel is generated as the inspection result image in any one of the layout diagrams of the solar panels of 2. The method of inspecting a solar panel according to claim 13, comprising the steps of: In the display processing step, the inspection is performed by matching the feature appearing in the layout drawing of the solar panel with the feature appearing in an image used when detecting the range of each of the solar panels to be inspected. Range of each of the solar panels detected in the range detecting step and an abnormal solar panel detected if an abnormal solar panel is detected in the abnormality detecting step to the layout diagram of the solar panel to be inspected The method of inspecting a solar panel according to any one of claims 12 to 14, wherein the projection is performed. Position information of the photographing position of the image used when the position information acquisition means detects the range of each of the solar panels to be inspected, and a predetermined position included in a solar panel layout including all of the solar panels to be inspected If the location information of is obtained from the location information determined by GPS,
In the display processing step, a range of each of the solar panels detected in the inspection range detection step and an abnormal solar panel in the abnormality detection step are detected based on the position information measured by the GPS. The method of inspecting a solar panel according to any one of claims 12 to 14, wherein projected abnormal solar panels are detected on the layout of the solar panel to be inspected. 17. The image according to any one of claims 10 to 16, wherein the image used when detecting the range of each of the solar panels in the inspection range detection step is a moving image obtained by imaging all of the solar panels to be inspected. The inspection method of the solar panel as described in a section. The solar panel inspection procedure is a still image which cuts out a predetermined frame from a moving image obtained by imaging each of the solar panels to be inspected as an image used when detecting the range of the solar panels to be inspected. The image extracting means further comprises the step of generating the still image,
The inspection method of a solar panel according to claim 17, wherein the inspection range detection step is performed using the still image. The thermal image and the visible image used when detecting the range of each of the solar panels in the inspection range detection step are both moving images captured with synchronized timing,
17. The method according to any one of claims 10 to 16, wherein the first inspection range detection step and the second inspection range detection step continuously detect the range of each solar panel to be inspected from the moving image. The inspection method of the solar panel of 1 item. A first inspection range detection step of detecting an outline of the solar panel based on a temperature distribution appearing in a thermal image taken from above the solar panel to be inspected and detecting the range of the solar panel to be inspected; Acquiring a visible image captured from the image, detecting an outline of each of the solar panels to be inspected from the acquired visible image, projecting the detected outline onto the thermal image to detect an area of each of the solar panels to be inspected An inspection range detection step including a second inspection range detection step;
Setting a plurality of partial areas having a predetermined area in the thermal image of the solar panel detected in the inspection range detection step, and calculating the luminance L and the standard deviation value σ of the luminance L for each partial area; A threshold value t defined by the following equation [Equation 1] based on the average value μ of the luminance L and the standard deviation value σ in the partial area where the standard deviation value σ is the smallest is a threshold value t for determining the temperature abnormal point presence or absence It is determined that there is a portion showing temperature abnormality in the whole of the thermal image by comparison with the threshold value t or that there is a line segment vector in a direction inclined with respect to the outline in the visible image In this case, an abnormality detection step including an abnormality detection step of judging the solar panel as abnormal and detecting an abnormal solar panel;
A position information acquisition step of acquiring information of an imaging position of an image used when detecting the range of the solar panel in the inspection range detection step;
When the range of the solar panel detected in the inspection range detection step, the information on the photographing position of the image acquired in the position information acquisition step, and the abnormal solar panel is detected in the abnormality detection step Based on the detected abnormal solar panel, the abnormal solar panel is distinguished from the solar panel not determined to be abnormal in the abnormal detection step to generate an inspection result image showing the solar panel to be inspected An inspection program of a solar panel, which causes a computer to execute a solar panel inspection procedure including a display processing step of
t = μ + nσ (n is any positive number)

Images