BE1026027B1 - METHOD FOR AUTOMATICALLY EVALUATING ROOFS OF BUILDINGS IN A REGION THROUGH IMAGE PROCESSING OF THE SHADE OF THE BUILDING. - Google Patents

Abstract

A computer-implemented method (100) is described for automatically evaluating roofs (114) of buildings (112) in a region (110) by image processing of the building's shadow.

Description

METHOD FOR AUTOMATICALLY EVALUATING ROOFS OF BUILDINGS IN A REGION THROUGH IMAGE PROCESSING OF THE SHADE OF THE BUILDING.

The invention relates to a computer-implemented method, as well as an associated system, for automatically evaluating roofs of buildings in a region by means of image processing. Such a method is particularly advantageous for automatically locating optimum locations for the installation of, for example, solar panels in a certain region.

In order to efficiently install solar panels in a certain region, in particular in an urban region, with dense buildings, in order to optimize the local production of renewable energy, it is necessary to use a large number of roofs from buildings the most suitable roofs. Often geographical information is available about the region, from which, for example, the built-up area for each building can be derived. However, often specific information is missing that allows to determine the specific surface and configuration of the roof, which is important, for example, for the evaluation of the roof in function of the installation of solar panels. Systems based on the automatic processing of height measurements and / or image processing of satellite images do not provide sufficiently accurate results to reliably evaluate a large number of roofs in a region. Moreover, such methods have limitations in automatically detecting obstacles such as roof windows, etc. Often time-consuming site visits are necessary to evaluate buildings with a suitable roof area in detail to determine whether the roof is of a suitable type or whether the roof is a suitable one orientation, whether there are no obstacles on or around the roof that would complicate the installation of solar panels and / or cast a shadow on installed solar panels, etc.

So there is a need for a method and device that allows to roof automatically and on a large scale and with greater accuracy

BE2018 / 5753 buildings in a region, in particular with a view to selecting suitable roofs for an installation of solar panels. There is also a need for a method and a system to detect faults in solar panels in a region in a reliable and efficient way.

To this end, according to a first aspect of the invention, a computer-implemented method is provided for automatically evaluating building roofs in a region by image processing, the method comprising the following steps:

- receiving geographical metadata from the region, the geographical metadata containing geographical information of the built-up area of the buildings of the region;

- receiving satellite images from the region;

- identifying images of the roofs of the buildings by selecting the images from the satellite images corresponding to the geographical information of the built-up area of the buildings from the geographical metadata;

- automatically identifying one or more segments in the images of the roofs; and

- identifying an image of the building's shadow by selecting an image with a lower brightness that connects to one side of the perimeter of the built surface of the building.

This makes it possible to automatically, on a large scale, reliably and efficiently evaluate roofs of buildings in a region, in particular for the selection of the most suitable roofs for the installation of solar panels. This is particularly suitable for maximizing the efficiency of installations of solar panels, in particular in an urban region, with a large number and / or a high density of buildings. Moreover, the use of the image of the building's shadow makes it possible to detect details about the segments of the roof with great accuracy by means of automatic image processing. Thus it can be deduced from the shadow or the type of it

BE2018 / 5753 roof is a sloping roof or a flat roof. Moreover, it is possible to deduce from the shade which slope a roof segment contains. Furthermore, the height of the obstacles can, for example, be derived from the shadow of obstacles, whereby a correct calculation of the shadow zones of these obstacles can be performed.

According to an embodiment, a method is provided, the method comprising the following step:

- automatically identifying a type of roof and / or the slope of segments of the roof based on the outline of the building's shadow.

According to an embodiment, a method is provided, the method comprising the following step:

- the automatic detection of one or more obstacles in the images of the roofs, based on:

- deviating color or color components;

- different brightness;

- adjacent shadow within the perimeter of the built-up area

- shape the obstacle of the shadow of the building.

According to an embodiment, a method is provided, the method comprising the following step:

- automatically identifying the usable area, orientation, slope and / or obstacles for placing one or more solar panels in the one or more segments in the images of the roofs.

According to an embodiment, a method is provided, the method comprising the following step:

- automatically identifying one or more segments in the images of the roofs, wherein each segment contains a surface with a certain evenness.

BE2018 / 5753

According to an embodiment, a method is provided, the method comprising the following step:

- automatically identifying one or more segments in the images of the roofs, wherein each segment contains a surface of which:

- more than 80% of the values of a parameter deviate less than 20% from the average value of that parameter of this segment; and

- more than 80% of the values of a parameter deviate more than 20% from the average value of that parameter of the adjacent segments.

According to an embodiment, a method is provided wherein the parameter contains one or more of the following components:

- the color or color components;

- the brightness

According to an embodiment, a method is provided, the method comprising the step of:

- automatically identifying a type of roof where:

the type of roof is a flat roof when only one segment is identified;

- the type of roof is a sloping roof when two or more adjacent segments are identified.

According to an embodiment, a method is provided, the method comprising the following step:

- automatically identifying one or more segments in the images of the roofs by automatically detecting internal, linear edges that form the perimeter of the one or more segments.

According to an embodiment, a method is provided, the method comprising the step of:

- automatically identifying a type of roof where:

BE2018 / 5753

the type of roof is a flat roof when only one segment is identified;

the type of the roof is a sloping roof when an internal edge is identified between two adjacent segments that extends parallel to and centrally between two opposite sides of the perimeter of the built-up area.

According to an embodiment, a method is provided wherein the automatic identification of one or more segments is performed such that the segments contain one or more of the following forms:

- a rectangle;

- a triangle;

- a trapezium.

According to an embodiment, a method is provided wherein the automatic identification of one or more segments is performed such that the segments contain at least one of the sides of the perimeter of the built-up area.

According to a second aspect of the invention there is provided a system configured to automatically evaluate roofs of buildings in a region by means of image processing according to the method of the first aspect of the invention, comprising a computer system configured to:

- receive geographical metadata from the region, the geographical metadata containing geographic information of the built-up area of the buildings of the region;

- receiving satellite images of the region;

- automatically identify images of the roofs of the buildings by selecting the images from the satellite images corresponding to the geographic information of the built-up area of the buildings from the geographic metadata;

- automatically identify one or more segments in the roof images; and

BE2018 / 5753

- automatically identify an image of the building's shadow by selecting an image with a lower brightness that connects to one side of the perimeter of the built area of the building.

According to a third aspect of the invention there is provided a computer program containing instructions which, when executed by a computer, cause the computer to perform the method according to the first aspect of the invention.

According to a following aspect, a computer-implemented method is provided for automatically monitoring solar panels on roofs of buildings in a region, the method comprising the following steps:

- receiving geographical metadata from the buildings in the region with solar panels;

- receiving power generation from the solar panels of the buildings of the region via monitoring modules;

- determining a control group comprising several monitoring modules for a monitoring module based on the geographical metadata of the corresponding buildings of the solar panels;

- determining a deviation of the power generation received via a monitoring module from the power generation of the monitoring modules of its control group; and

generating an error message when the deviation exceeds a certain value.

This allows reliable and efficient detection of errors in solar panels in a region.

According to an embodiment, a method is provided in which the control group of a monitoring module is determined such that the control group contains monitoring modules which:

- belong to a predefined defined geographical region;

BE2018 / 5753

- whose distance to the monitoring module is less than a predetermined maximum distance.

According to an embodiment, a method is provided, the method further comprising the following steps:

- receiving geographical metadata from the region, the geographical metadata containing geographical information of the built-up area of the buildings of the region;

- receiving satellite images from the region;

- identifying images of the roofs of the buildings by selecting the images from the satellite images corresponding to the geographic information of the built-up area of the buildings from the geographic metadata; and

- automatically identifying one or more segments in the images of the roofs;

- automatically identifying the presence of solar panels on one or more segments in the roof images;

- receiving the power consumption of monitoring modules of buildings from roofs where solar panels were identified.

According to an embodiment, a method is provided, the method further comprising the following step:

- the automatic detection of the orientation, slope and / or used surface area of the solar panels in the one or more segments in the images of the roofs.

According to an embodiment, a method is provided, wherein furthermore the control group of a monitoring module is determined such that it contains monitoring modules whose orientation, slope and / or surface area used deviates only within certain predetermined limits from the orientation, slope and / or area used of the monitoring module that is compared with the control group.

BE2018 / 5753

By way of example, some embodiments will be described in more detail with reference to the following figures, wherein:

Figure 1 shows schematically an embodiment of geographical metadata of a region with several buildings;

Figure 2 schematically shows an embodiment of a satellite image of the region of Figure 1;

Figure 3 schematically shows an embodiment of a computer-implemented method for automatically evaluating by means of image processing the roofs of the buildings of the region of Figures 1 and 2;

Figure 4 schematically shows an embodiment of a system for performing the computer-implemented method of Figure 3;

Figure 5 shows schematically a detail of Figure 3:

Figure 6 represents an embodiment of a monitoring system for a group of different installations of solar panels in a region;

Figure 7 schematically represents an embodiment of a method for monitoring a group of different installations of solar panels in a region; and

Figure 8 schematically shows an embodiment of a suitable computer system.

Figure 1 schematically shows an embodiment of geographic metadata 200 of a region 110 that includes a plurality of buildings 112. As can be seen, the embodiment of the geographic metadata includes, for example, geographic metadata such as, for example, the perimeter, location, area, etc. of each building 112 in the region 110. Typically, such geographic metadata includes metadata 200 for each building 112 in the region 100, geographic metadata 200 describing the location and perimeter of the plot 210 on which a respective building is located. As shown, the embodiment of the geographic metadata 200 also includes, for each plot 210 of a building 112, the location and perimeter of the built-up area 212 occupied by the building 112 on the plot 210. Such geographical metadata 200, for example, originates from a suitable geographic information system,

BE2018 / 5753 usually abbreviated to GIS. A GIS is offered, for example, in the form of a suitable database that contains geographical or spatial data or information about geographical objects, also known as geo-information. Thus, an embodiment of the geographic metadata 200 may, for example, be offered by governments or another suitable agency, in the form of a suitable GIS containing geo-information about the buildings of a region. Furthermore, as shown, a compass rose 120 is shown schematically to clarify the orientation of the region 110 and the associated geographic metadata 200. It is clear that alternative embodiments are possible, in which geographical metadata 200 of the region 110 generally contains geographical information of the built-up area 212 of the buildings 112 of the region 110.

Figure 2 schematically shows an embodiment of a satellite image 300 of the region 110 shown in Figure 1. As will be described in more detail, such a satellite image 300 includes a suitable digital image or photo in which the roofs 114 of the buildings 112 of the region 110 are visible. Such a satellite image 300 is made available according to a suitable embodiment, for example via a suitable database containing such satellite images for a specific region 110. These satellite images preferably contain 300 images of the region 110 with a sufficiently high resolution that allows structures such as a roof 114 of a building 112 to be identified by image processing. Such satellite images are offered, for example, via publicly available online services offered by companies and governments, such as Apple Maps, Google Maps, Bing Maps, etc.

Figure 3 schematically shows an embodiment of a computer-implemented method for automatically evaluating by means of image processing the roofs 114 of the buildings 112 of the region 110 of Figures 1 and 2. As visible, the embodiment of the computer-implemented method 100 includes a step 122 receiving geographical metadata 200 from the region 110. As indicated above, the

BE2018 / 5753 Geographical metadata 200 Geographical information of the built-up area 212 of the buildings 112 of the region 110, for example similar as shown in Figure 1. It is clear that such geographic information contains, for example, suitable coordinates that include the location, the perimeter, etc. of the built-up area 212 of the buildings 112. In step 124, satellite images 300 are received from this region 110, for example similar as shown in Figure 2.

Then, according to this embodiment of the method 100, in step 126, images 310 of the roofs 114 of the buildings 112 from the satellite images 300 are identified. This identification in step 126 is realized by selecting images 310 from the satellite images 300 of the region 110 corresponding to the built-up area 212 of the buildings of the region 110 from the geographical metadata 200. This means, for example, that for each building 112 shown in the region 110 in Figure 1, the location and perimeter of the built-up area 212 from the geographic metadata 200, is used to identify the portion of the satellite image 300 at a corresponding location with a corresponding perimeter, which is further referred to as a image 310 of the roof 114 of the corresponding building 112 will be considered. When reference is made in the following description to the built-up area 212, it is clear that this is a reference to suitable data for determining the location and the contours of the area 212 occupied by a corresponding building 112. Based on this data can be selected at a corresponding location in the satellite image 300, an image 310 with a corresponding contour or perimeter as determined by the built area 212 of the building from the geographic metadata 200. Although in the illustrated embodiment of Figures 1 and 2, If the built-up area 212 contains a rectangular outline, it is clear that alternatives are possible in which the built-up area 212 contains any suitable polygon or other suitable outline as a perimeter. It is clear that location data associated with the satellite image 300 is appropriately used for this purpose. It is clear that in this way the images 310 of the roofs 114 of the

BE2018 / 5753 buildings 112 are automatically identified by selecting the images 310 from the satellite images 300 corresponding to the geographic information of the built-up area 212 of the buildings 112 from the geographic metadata 200. This means that when the satellite images 300 of the region 110 contains multiple buildings 112, this will result in multiple corresponding images 310 of the roofs 114 of the buildings 112 of this region 110. This method of automatically identifying the images 310 of the roofs 114 in the satellite images 300 is efficient and reliable, since the geographic information of the built-up area 212 from the geographic metadata 212 allows a reliable and efficient automatic selection of these images from the satellite images 300. As described above, based on the geographic information, an image of the corresponding location should be selected from the satellite image 300. It is clear to a person skilled in the art that satellite images 300 also contain the necessary geographical information for this to be possible, such as, for example, the corresponding geographical coordinates of the satellite image and / or parts of the satellite image, the orientation of the satellite image, the scale, etc. . It is clear that when reference is made to the satellite images 300, this is a reference to one or more satellite images 310 which only, together, in overlap, etc. contain an aerial image of the region 110.

Although Figures 1 and 2 schematically show a representation where this data is presented in map form, it is clear that in the automatic processing of this data by means of the computer-implemented method 100, the metadata 200 and the satellite images 300 are processed in the form of suitable data or data, such as made available for example via a suitable database, web service, data files, etc. for automatic processing by a suitable system 10 as schematically shown in Figure 4. According to the exemplary embodiment shown, the system 10 comprises, for example, a suitable server 400 which automatically via a suitable network 600, such as, for example, the internet geographic metadata 200 and corresponding satellite images 300 of a region 110 from corresponding

BE2018 / 5753 can request databases. A user 502 can, for example, enter a desired region 110 via a suitable user interface of a computer 500 connected to the network 600 and send it to the server 400. The server 400 then executes the method 100 and, for example, presents the output to a user 502 suitably via the user interface of the computer 500 to the user 502.

For example, to automatically determine the usable area of the roofs 114 of the buildings 112 of the region 110, it is advantageous to automatically determine the type of roof. This means that it is automatically determined whether the roof is a sloping roof with, for example, two or more parts with different corresponding angles with respect to the ground surface, and / or whether the roof is a flat or flat roof that largely consists of a horizontal surface. According to the method shown in the embodiment of Figure 3, in step 128, segments 320 of the roofs 114 are identified based on the identified images 310 of the roofs 114. That is, in each image 310 of a roof 114, one or more segments 320 of the roof are automatically identified. Figure 5 shows schematically, in more detail, the images 310 of two roofs 114 from the satellite image 300 of Figure 2. The images 310 of these two roofs 114 are, as described above, selected from the satellite image 300 by means of the corresponding built-up area 212 from the geographic metadata 200. The images 310 are schematically indicated by the broken line indicated by references 312 and 314. It is clear that the perimeter of the images 310 of the roofs 114 does not have to exactly correspond to the perimeter of the built-up area and that the images 310 of the roofs 114 as they are automatically processed may contain a suitable margin or section of the satellite image 300 that includes, for example, a suitable margin or edge around the built-up area 212. It is clear that even if the image 310 as it is being processed automatically contains additional elements around the built-up area 212, the built-up area 212 still allows for the part and circumference thereof in this image 310 corresponding to the roof 114 to decide.

BE2018 / 5753

As indicated in the embodiment of Figure 3, the identification of the segments 320 in the image 310 of a roof 114 allows, for example, in step 130 to determine the type of the roof, in particular whether it is a sloping or flat roof. Furthermore, as shown in step 132, it also allows, for example, to automatically determine the slope and / or orientation of the segments 320 of the roof 114. With regard to the automatic evaluation of, for example, the electricity generated in an installation of solar panels, this is an important parameter, in particular with a sloping roof. Furthermore, as shown in step 134, this also makes it possible, for example, to detect obstacles on the roof such as fireplaces, venters, etc. that limit the usable surface area for installing solar panels and in some cases, through their shading effect, the efficiency of the solar panels. can influence solar panels.

As shown schematically in Figure 5, the automatic identification of one or more segments 320 in the images 310 of the roofs 114 can occur based on the smoothness of the surface of each segment 320. This means that, as schematically shown in Figure 5, a surface of the part of the image 312 of the roof 114 corresponding to the built-up area 212 of a first building 112, which for the most part has a single surface that exhibits a certain level of uniformity, can automatically be regarded as a roof 114 with only one segment 320. Om by image processing, automatically determining the evenness of a certain surface to segment the roof image 310 into one or more segments 320, for example, use is made of parameters containing one or more of the following components: the color, one or more color components, brightness etc. For example, each surface with a certain uniform brightness forms a segment 320, or a certain uniform color or one or more color components, etc. It is thereby clear that, as for example in the image 314 of a building 112 with a gable roof 114, the ridge of which runs along the east-west axis, as shown for example in Figure 5, two rectangular segments 320 can be automatically detected which extend on either side, symmetrically with respect to the cam, and together form the rectangular circumference of the

BE2018 / 5753 roof 114 corresponding to the built-up area e 212 of the building 112. This is the result of the fact that the light from the sun on the parts of the roof with a different slope is different and that the roof covering of a roof typically has an even color. In the exemplary embodiment shown, the brightness of the part of the roof directed to the south will be uniform and slightly higher than the part of the roof directed to the north. It is clear that instead of or in addition to the brightness, another parameter, such as for example the color, one or more color components such as for example RGB, YUV, etc. or any other suitable color system, and / or parameters derived therefrom can be used to identify the various segments 320 by image processing.

The automatic identification of one or more segments 320 in the images 310 of the roofs 114, for example, can be done using a suitable image processing algorithm. This image processing algorithm allows, for example, to identify a segment 320 as a surface whose: - more than 80% of the values of a parameter deviate less than 20% from the average value of that parameter of this segment 320; and - more than 80% of the values of a parameter deviate more than 20% from the average value of that parameter of the adjacent segments 320.

It is clear here that the image processing can be greatly simplified, since the images 310 of a roof 114, in particular for buildings with a rectangular perimeter, one or more segments 320 corresponding to a certain basic shape depending on the type of roof used. . For example, a flat roof will contain one segment 320 corresponding to the built-up area 212, a sloping roof designed as a gable roof containing two symmetrical rectangular segments 320 on either side of a central ridge, a sloping roof designed as a hipped two symmetrical roof trapezia on either side of the ridge and two contiguous, symmetrical triangles on the sides of the roof in line with the ridge, a sloping mansard roof will have four contiguous

BE2018 / 5753 contains rectangular segments 320, symmetrically arranged with respect to the central ridge, a pitched tent roof or pyramid roof will contain four contiguous triangles, etc. This allows the image processing algorithm to quickly and efficiently determine which configuration of such segments 320 the image 310 shows the best match. It is clear that still further alternatives are possible, but that the automatic identification of the one or more segments 320 is preferably performed such that the segments 320 comprise one or more of the following forms: a rectangle; a triangle; and / or a trapezium. It is clear that, as for example shown in Figures 3 and 5, the automatic identification by means of an image processing algorithm of the one or more segments 320 is carried out in such a way that the segments 320 at least one of the sides of the perimeter of the built-up area 212 contain. In other words, for most roof types, each segment 320 connects with at least one side to one side of the perimeter of the built-up area. For the image 312 in Figure 5, the single segment 320 connects completely to the sides of the rectangular perimeter of the built-up area 212. As visible in Figure 5, a first segment 322 connects to the south side of the rectangular perimeter of the built-up area 212 and a second rectangular segment 324 on the north side of the rectangular perimeter of the built-up area 212, for the image 314 of a building with a sloping gable roof. It is clear that one or more segments 320 that are identified, connect to each other and together correspond to the built-up area 212.

It is clear that the automatic identification of the segments 320 in the image 310 of the roof 114 by image processing allows to automatically identify the type of the roof as shown, for example, in step 130 of Figure 3. when only one segment becomes 320 identified in the image 310 of the roof 114, as for example in the image 312 in Figures 3 and 5, the type of the roof 114 can be identified as a falcon or flat roof. When two or more adjacent segments are identified, the type of the roof 114 can become an inclined roof

BE2018 / 5753, as for example in the image 314 as shown in Figures 3 and 5.

According to an alternative embodiment, the identification of segments 320 in the images 310 of the roofs 114 can take place instead of or in addition to the identification of smooth surfaces as described above by means of image processing with the aid of detecting internal, linear edges 330. This means the detection of edges that have a straight shape and are located within the circumference of the part of the image 310 of the roof 114 corresponding to the built-up surface 212. Thus, the automatic identification of one or more segments 320 in the images of the roofs 114 can be realized by automatically detecting internal, linear edges 330 which form the perimeter of the one or more segments 320. For example, as shown in the Figure 314 of Figure 5 allows this, for example, to detect by means of automatic image processing the internal edge 330 formed by the central ridge of the roof as an internal straight edge extending parallel and / or centrally between two opposite sides of circumference of the built-up area 212. In this case, where it concerns a gable roof, it is also clear that this edge 330 also extends between the two other opposite sides of the rectangular circumference of the illustrated embodiment of the part of the image 310 of the roof 114 corresponding to the built-up area 212.

Similar as described above, this allows the type of roof 114 to be automatically identified as a flat or flat roof 114 when only one segment 320 is identified, the perimeter of which corresponds to the perimeter of the built-up area. In other words, when, as in the image 312 in Figure 5, no linear internal edge 330 is detected, i.e. a linear, internal edge 330 sufficiently corresponding to an internal edge that could be a side of a segment of a roof represent. That is, for example, extending parallel and / or centrally between two opposite sides of the circumference of the built-up area 212, or for example starting from a corner point of

BE2018 / 5753 the circumference of the built-up area in order to form a trapezium or triangle together with another internal edge. As will be further described, internal edges, even planes corresponding to segments 320 of the image 310 of the roof 114 can thus be distinguished from obstacles 340, which typically contain only a smaller even surface or shorter internal edges than the segments 320 corresponding to different contiguous roof sections that have a different slope. It is furthermore clear that by edge detection image processing, the type of the roof 114 can be automatically identified as a sloping roof when an internal edge 330 is identified between two adjacent segments 320 parallel to and centrally between two opposite sides. of the circumference of the built-up area 212, as shown, for example, in the image 314 in Figure 5.

Instead of and / or in addition to embodiments where the segments 320 are detected by image processing in the image 310 on the basis of smooth surfaces and / or edge detection, it is also possible, as shown in Figure 5, to enter or supplement the roof image 310 automatically identifies an image 350 of the shadow of the building 112 by selecting a lower brightness image that connects to one side of the perimeter of the built area 212 of the building 112. As can be seen, the circumference of the shadow to automatically deduce the type of the roof 114 or the slope of the segments 320 from the circumference of the shadow of the building. For example, the outline of the shadow 350 of the building 112 with a flat roof of the image 312 includes a corresponding rectangular outline in Figure 5. The outline of the shadow 350 that connects to an east side of the built-up area 212 of the building 112 of the image 314 comprises, as indicated in Figure 5, two sloping sides 354, 356 on the part of the circumference of the shade 350 corresponding to the roof of the building, on the basis of which the type of roof can be identified as a sloping roof. According to an advantageous embodiment, the automatically identified shadow, and more particularly the shape of its circumference, further allows to determine the slope of the segments 320 of the roof. On base

BE2018 / 5753 of the relative position of the shadow with respect to the building and the geographic data of the built-up area 212 and / or the time and location the satellite image 300 can be adjusted by photogrammetry of the height, the orientation, the shape of the building and therefore also the slope of the roof on the shadow side of the built-up area can be determined

As already mentioned above, and as visible in the embodiment of Figure 5, a number of obstacles can be present on the roof 114 which prevent the placement of solar panels. According to the exemplary embodiment shown, a number of obstacles 340 in the form of, for example, light domes, skylights, technical installations, etc., are present on the flat roof on the roof 114. In the exemplary embodiment of image 314 in Fig. 5, for example, obstacle 340 in the form of a fireplace. It is clear that such obstacles have a relatively small area and / or circumference with respect to the built-up area 212, for example less than 10%, or less than 5%, and can therefore be easily distinguished from segments 320 corresponding to roof parts with a different slope. Furthermore, such obstacles 340, as also schematically shown in Figure 5, can be automatically detected and identified in the images 310 of the roofs 114, based on a deviating color or color components and / or a deviating brightness with respect to the segment 320 where the part of it. For example, a fireplace will typically have a lower brightness and a different color than the roof covering, for example gray, compared to a red color of a roof. Skylights, rooflights, etc. will also have a different brightness with respect to the roof covering, for example a higher brightness, or a light color that differs from the dark color of a flat roof roof covering. An obstacle, such as a mantelpiece, antenna, etc., which contains a certain height can also be detected on the basis of an adjacent shadow within the circumference of the built-up surface 212. Finally, the detection can also be done on the basis of a shape which appears in the shadow of the building. For example, as shown in the exemplary embodiment of Figure 5, the shadow of the

BE2018 / 5753 obstacle 340 clearly detectable as a form 356 in the shade 350 of the building. Similar to that described above, it may be possible to derive, based on the shadow of the obstacle 340, for example the height of the obstacle 340, making it possible to automatically identify which zones of the roof 114 are covered by the shadow of the obstacle 340 and when. Such information makes it possible, for example, to determine the optimum zones for placing solar panels on the roof, as well as to more reliably predict or evaluate the course of the power production of solar panels that are installed at this position on the roof.

It is clear that further variant embodiments are possible, the method comprising, for example, further steps in which, for example, it automatically includes the usable area, orientation, slope and / or obstacles for placing one or more solar panels in the one or more segments 320 in the images. 310 of the roofs 114 are identified. If, for example, additional data about, for example, the climate, the weather forecast, the technical parameters of the solar panels, the electricity prices, the electricity consumption of the houses, etc., for example from a database 610 are available for the system 600, then this way for the various homes 112 in the region automatically identify the most efficient roofs 114 for installing solar panels.

An alternative embodiment is, for example, when, as shown in the exemplary embodiment of Figure 6, the geographic information can also be used advantageously in detecting errors in installations of solar panels already in operation on buildings 112 in a region 110. The embodiment of the system from Figure 6 is similar to the embodiment shown in Figure 4, and similar elements are designated with similar references and perform a similar function. As can be seen, Figure 6 shows sensors or several modules 620 for monitoring the power production of installations of solar panels on buildings 112 in the region. Every building with an installation of

BE2018 / 5753 solar panels in a certain region 110, for example a city, province, region, etc. or other suitable geographical group that is subject to similar weather conditions that determine the power production of solar panels. Although four such monitoring modules 620 are shown in the exemplary embodiment, it is clear that any other suitable number of monitoring modules 620 are possible in a given region. Preferably, the number of monitoring modules 620 is sufficiently large, for example ten or more, one hundred or more, etc. Such a monitoring module 620 which is configured to monitor the power production of a solar panel installation of a home 112 can be designed, for example, as a separate module that monitors the power generation of the solar panels and makes it available via a suitable network 600, for example for further processing and storage in database 610. According to an advantageous embodiment, however, the monitoring module 620 may also consist of, or be integrated into, a monitoring module or meter for the electricity consumption of the house, which offers the advantage that not only the electricity production of the solar panels, but also the electricity consumption of the house itself can be monitored and made available for further processing and storage in a database such as for example database 610. Such a monitor g module 620 may, for example, contain a suitable communication module that can exchange the position of the electricity meter and / or the processed power of the solar panels via a suitable protocol. It is clear that still other variant embodiments are possible for the monitoring modules 620, as long as in general the monitoring modules 620 make the power consumption of the solar panels of a house 112 and / or the house 112 and / or the evolution thereof available at different times. via a suitable data network 600 for further processing by means of a suitable computer system 400, 500 and / or storage in a suitable database.

Figure 7 then shows an embodiment of a method for detecting errors that uses a system 10 similar to that described above with respect to Figure 6. Similarly described above, the illustrated embodiment of the method may

BE2018 / 5753 use of geographical metadata 200 and / or satellite images 300 that allows to determine the location of the monitoring modules 620 and their associated home with solar panels, as shown in step 702. Next, the method embodiment in step 704 also uses of the data from the power generation of the solar panels made available by the monitoring modules 620. Thus, in step 706 for the shown embodiment of the method a nearby control group can be determined for each monitoring module 620 or group of monitoring modules. According to a simple embodiment, the control group 630 for a particular monitoring module 620 can be determined as a function of a predetermined geographically defined region in the vicinity of this monitoring module 630. According to a simple embodiment, the nearby geographic control group 630 for a monitoring module 620 can be consist of a group of monitoring modules 620 that contain geographical metadata from which it can be deduced that they belong to the same city, they contain the same zip code, they belong to the same region or region, they in themselves are less than a predetermined maximum distance from the monitoring module 620, etc. As shown in Figure 6, a monitoring group 630 can be determined for the region 110 for the monitoring module 620 of a particular home, which also contains two other monitoring modules 620 of other homes, and where, such as yet another monitoring module 624 is not shown makes more of, for example, because this monitoring module 624 is related to solar panels on a home 112 of another city, or that is further than a predetermined maximum distance from the home 112 with solar panels from the monitoring module 622. As indicated in step 708 of the embodiment of the method, then, for the monitoring module 622, the deviation of the measured current production from the solar panels is compared with the values of the measured current production from the monitoring modules 620 of the control group 630. This may be for example, by comparing the evolution of the flow production of the monitoring module 622 with the evolution of the average flow production of the monitoring modules 620 of the geographic control group 630 of which this monitoring module 620 is a part. It is however

BE2018 / 5753 clearly that alternative embodiments are possible to detect a deviation. For example, it is possible to monitor if the total, average, weighted, historical, predicted, etc. current production of the geographic control group 630 exceeds a certain minimum limit value and / or also the current production of the monitoring module 620 exceeds a certain minimum limit value or not. It is clear that this allows to detect errors or defects in an efficient and reliable manner as indicated in step 710, when the deviation detected between a monitoring module 620 and its geographic control group 630 exceeds a predetermined threshold value. This makes it possible, for example, to reliably and quickly detect defective solar panels, inverters, etc., so that the power generation of the solar panels can be guaranteed.

It is clear that alternative embodiments are possible and that the nearby geographic control group 630 is preferably chosen to be large enough to contain sufficient monitoring modules 620 to be able to detect a reliable deviation of individual monitoring modules 620 from the control group 630. However, preferably the geographic control group 630 is also kept limited enough so that the monitoring modules 620 are linked to solar panels of homes 112 in a region that is exposed to similar climatic conditions, etc.

To be able to determine the deviations of a monitoring module 620 relative to the control group even better, for example, similar as described above, based on the geographical metadata 200 and / or the satellite images 300, the location, orientation, slope, etc. of the solar panels on the roof 114 of the corresponding home 112 are determined. This then allows for this to be taken into account when comparing the measured current production of a monitoring module 620 with respect to the current production of the control group 630. According to another variant embodiment, this can also allow the geographical control group 630 to be monitored module 620 to be further refined into monitoring modules

BE2018 / 5753

620 of solar panels from nearby houses with a similar orientation, slope, etc.

Figure 8 shows a suitable computer system 400, 500 for hosting the system 10 or any of its components as described with reference to the above-mentioned embodiments. Computer system 400 can generally be configured as a suitable general-purpose computer and a bus 410, a processor 402, a local memory 404, one or more optional input interfaces 414, one or more optional output interfaces 416, a communication interface 412, a storage element interface 406 and one or more storage elements 408. Bus 410 can comprise one or more conductors that allow communication between the components of the computer system. Processor 402 can include any type of conventional processor or microprocessor that interprets and executes programming instructions. Local memory 404 may include a Random Access Memory (RAM) or any other type of dynamic storage device that stores information and instructions to be executed by processor 402 and / or a Read Only Memory (ROM) or any other type of static storage device that has static stores information and instructions for use by processor 402. Input interface 414 may include one or more conventional mechanisms that allow an operator to enter information into the computer system 400, such as a keyboard 420, a mouse 430, a pen, voice recognition and / or or biometric mechanisms, etc. Output interface 416 may include one or more conventional mechanisms that output information to the operator, such as a display 440, a printer, a speaker, etc. Communication interface 412 may include one or more transceiver-like mechanisms such as two 1 Gb Ethernet interfaces that computer system 400 via a suitable data communication network enable k 600 to communicate with other devices and / or systems, for example mechanisms for communicating with one or more other computer systems 500. The communication interface 412 of computer system 400 may be connected to such other computer system 500 by means of a LAN (local area network) or WAN (wide area network), such as the internet.

BE2018 / 5753

Storage element interface 446 may include a storage interface such as a Serial Advanced Technology Attachment (SATA) interface or a Small Computer System Interface (SCSI) to connect bus 410 to one or more storage elements 408, such as one or more local disks, e.g. 1TB SATA disk drives, and control the reading and writing of data to and / or of these storage elements 408. Although the storage elements 408 are described above as a local disk, any other computer-readable media can generally also be used, such as a removable magnetic disk, optical storage media such as a CD or DVDROM, SSDs, flash memory cards, etc.

The automatic system 10 according to the above-mentioned embodiments can be implemented by means of a suitable computer program that is executed on a computer system 400 that is locally available to a user 502, such as a personal computer, laptop, etc. or on a remote accessible computer system such as one or more servers available to one or more users. Alternatively, the system 10 may also include web service or cloud-based servers. It is clear that the components of the system 10 and their associated computer-implemented method can be implemented as programming instructions that are stored in the local memory 404 of the computer system 400 to be executed by its processor 402. Alternatively, these components can be stored on the storage element 408 or accessible from another computer system 500 via the communication network 600. In general, the system 10 and the associated computer-implemented method 100 are provided in this way as a computer program comprising software code adapted to this implement a computer-implemented method if performed by a computer system. Alternatively, the system 10 and the associated computer-implemented method 100 may also be provided as a computer-readable storage medium comprising a computer

BE2018 / 5753 executable instructions that, if implemented by a computer system, implement the method implemented by a computer.

Although the present invention has been illustrated with reference to specific embodiments, it will be apparent to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be practiced with various modifications and modifications without thereby to leave the scope of the invention. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being described by the appended claims and not by the foregoing description, and all modifications falling within the scope of the claims are therefore here included.

In addition, the reader of this patent application will understand that the words comprising or comprising, "containing" or "containing" do not exclude other elements or steps, that the word excludes a plural, and that a single element, such as a computer system, a processor or another integrated unit can perform the functions of different tools mentioned in the claims. Any references in the claims should not be construed as limiting the claims in question. The terms first, second, third, a, b

c and the like, when used in the description or in the claims, are used to distinguish between similar elements or steps and do not necessarily describe a sequential or chronological order. Similarly, the terms top, bottom, over, under and the like are used for the purposes of the description and do not necessarily refer to relative positions. It is to be understood that those terms are interchangeable under proper conditions and that embodiments of the invention are capable of functioning in accordance with the present invention in sequences or orientations other than described or illustrated above.

Claims (12)

CONCLUSIONS A computer-implemented method (100) for automatically evaluating roofs (114) of buildings (112) in a region (110) by image processing, the method comprising the following steps: - receiving geographical metadata (200) from the region (110), wherein the geographical metadata (200) contains geographic information of the built-up area (212) of the buildings (112) of the region (110); receiving satellite images (300) from the region (110); -identifying images (310) of the roofs (114) of the buildings (112) by selecting the images (310) from the satellite images (300) corresponding to the geographical information of the built-up area (212) of the buildings (112) from the geographical metadata (200); - automatically identifying one or more segments (320) in the images (310) of the roofs (114); and - identifying an image (350) of the building's shadow (112) by selecting a lower-brightness image that connects to one side of the perimeter of the built-up area (212) of the building (112). A method according to claim 1, wherein the method comprises the following step: - automatically identifying a type of roof (114) and / or the slope of segments (320) of the roof based on the circumference of the shadow of the building (112). A method according to one or more of the preceding claims, wherein the method comprises the following step: - automatically detecting one or more obstacles in the images (310) of the roofs (114), based on: - deviating color or color components; - different brightness; - adjacent shadow within the perimeter of the built-up area (212) BE2018 / 5753 - shape (356) of the obstacle of the shadow of the building. A method according to one or more of the preceding claims, wherein the method comprises the following step: - automatically identifying the usable area, orientation, slope and / or obstacles for placing one or more solar panels in the one or more segments (320) in the images (310) of the roofs (114). A method according to one or more of the preceding claims, wherein the method comprises the following step: - automatically identifying one or more segments (320) in the images (310) of the roofs (114), wherein each segment (320) contains a surface with a certain evenness. A method according to one or more of the preceding claims, wherein the method comprises the following step: - automatically identifying one or more segments (320) in the images (310) of the roofs (114), wherein each segment (320) comprises a surface of which: - more than 80% of the values of a parameter deviate less than 20% from the average value of that parameter of this segment (320); and - more than 80% of the values of a parameter deviate more than 20% from the average value of that parameter of the adjacent segments (320). A method according to claim 6, wherein the parameter contains one or more of the following components: - the color or color components; - the brightness A method according to one or more of the preceding claims, wherein the method comprises the following step: BE2018 / 5753 - automatically identifying one or more segments (320) in the images of the roofs (114) by automatically detecting internal, linear edges (330) that form the perimeter of the one or more segments (320). A method according to one or more of the preceding claims, wherein the automatic identification of one or more segments (320) is carried out such that the segments (320) comprise one or more of the following forms: a rectangle; -a triangle; -a trapezium. A method according to one or more of the preceding claims, wherein the automatic identification of one or more segments (320) is performed such that the segments (320) contain at least one of the sides of the perimeter of the built-up area (212) . A system (10) configured to automatically evaluate roofs (114) of buildings (112) in a region (110) by image processing according to the method of one or more of the preceding claims, comprising a computer system (400, 500) configured to: - receive geographical metadata (200) from the region (110), the geographical metadata (200) containing geographical information of the built-up area (212) of the buildings (112) of the region (110); - receiving satellite images (300) of the region (110); - automatically identify images (310) from the roofs (114) of the buildings (112) by selecting the images (310) from the satellite images (300) corresponding to the geographic information of the built area (212) of the buildings (112) from the geographical metadata (200); and - automatically identify one or more segments (320) in the images (310) of the roofs (114); and BE2018 / 5753 - automatically identify an image (350) of the building's shadow (112) by selecting a lower-brightness image that connects to one side of the perimeter of the built-up area (212) of the building (112). A computer program containing instructions that, when executed by a computer, cause the computer to perform the method according to one or more of claims 1 to 10.