Abstract: Automating conversion of drawings to indoor maps and plans. One example is a computer-implemented method of machine learning, the method comprising: receiving, by a device, a first set of updates to a first indoor map, the first indoor map previously created by a production machine-learning system having a production map accuracy; training, by a device, a supporting machine-learning system using the first set of updates to the first indoor map; and then applying, by a device, test data to the supporting machine-learning system, the applying results in first-evaluation indoor map with a first-evaluation map accuracy; and when the first-evaluation map accuracy is within a predetermined window above the production map accuracy refraining, by a device, from updating the production machine-learning system based on the first set of updates.

Abstract: Automating conversion of drawings to indoor maps and plans. One example is a computer-implemented method of determining a geo-location, the method comprising: determining a floor-level outline of a floor depicted in a CAD drawing; receiving an approximate geo-location of a building to which the CAD drawing applies; obtaining an overhead image of a target area encompassing the approximate geo-location, the overhead image comprising a plurality of buildings within the target area; identifying a plurality of building footprints within the target area; calculating, by a device, a plurality of distance functions that relate the floor-level outline to the each of the plurality of building footprints, the calculating creates a plurality of similarity scores; selecting a building footprint from plurality of building footprints, the selecting based on the plurality of similarity scores; and calculating a final geo-location of the building corresponding to the building footprint.

Abstract: Automating conversion of drawings to indoor maps and plans. One example is a computer-implemented method of creating an indoor map from a CAD drawing, the method comprising: preprocessing an original CAD drawing to create a modified CAD drawing, a text database containing text from the original CAD drawing, a CAD vector-image of the modified CAD drawing, and a CAD raster-image of the modified CAD drawing; creating a floor-level outline; sensing furniture depicted on the floor, the sensing creates set of furniture entities; identifying a room depicted in the CAD drawing; and creating the indoor map for a floor using the floor-level outline, the room-level outline and the room identity.

Abstract: Automating conversion of drawings to indoor maps and plans. One example is a computer-implemented comprising: preprocessing an original CAD drawing to create a modified CAD drawing, a text database containing text from the original CAD drawing, a CAD vector-image of the modified CAD drawing, and a CAD raster-image of the modified CAD drawing; determining a floor depicted in the CAD drawing, the determining results in a floor-level bounding line; sensing furniture depicted on the floor by applying the floor-level bounding line, the CAD vector-image, and the text database to machine-learning algorithms, the sensing results in a plurality of furniture entities and associated location information; identifying each room depicted in the CAD drawing, the identifying results in a plurality of room outlines; and creating an indoor map for the floor by combining the plurality of furniture entities and associated location information with the plurality of room outlines.

Abstract: Automating conversion of drawings to indoor maps and plans. One example is a computer-implemented comprising: preprocessing an original CAD drawing to create a modified CAD drawing, a text database containing text from the original CAD drawing, a CAD vector-image of the modified CAD drawing, and a CAD raster-image of the modified CAD drawing; determining a floor depicted in the CAD drawing, the determining results in a floor-level bounding line; sensing furniture depicted on the floor by applying the floor-level bounding line, the CAD vector-image, and the text database to machine-learning algorithms, the sensing results in a plurality of furniture entities and associated location information; identifying each room depicted in the CAD drawing, the identifying results in a plurality of room outlines; and creating an indoor map for the floor by combining the plurality of furniture entities and associated location information with the plurality of room outlines.

Abstract: Automating conversion of drawings to indoor maps and plans. One example is a computer-implemented method comprising: preprocessing a CAD drawing to create a text database containing text from the CAD drawing and associations of the text with locations within the CAD drawing; determining a floor depicted in the CAD drawing, the determining results in a floor-level outline; identifying a plurality of room-level outlines within the floor-level outline, the plurality of room-level outlines corresponds to a respective plurality of rooms; selecting a name of a first room from the plurality of rooms, the selecting based on text within the text database; and creating an indoor map including the name of the first room, the name of the first room associated with a location of the first room within the floor-level outline.

Abstract: Automating conversion of drawings to indoor maps and plans. One example is a computer-implemented method of machine learning, the method comprising: receiving, by a device, a first set of updates to a first indoor map, the first indoor map previously created by a production machine-learning system having a production map accuracy; training, by a device, a supporting machine-learning system using the first set of updates to the first indoor map; and then applying, by a device, test data to the supporting machine-learning system, the applying results in first-evaluation indoor map with a first-evaluation map accuracy; and when the first-evaluation map accuracy is within a predetermined window above the production map accuracy refraining, by a device, from updating the production machine-learning system based on the first set of updates.

Abstract: Automating conversion of drawings to indoor maps and plans. One example is a computer-implemented method of determining a geo-location, the method comprising: determining a floor-level outline of a floor depicted in a CAD drawing; receiving an approximate geo-location of a building to which the CAD drawing applies; obtaining an overhead image of a target area encompassing the approximate geo-location, the overhead image comprising a plurality of buildings within the target area; identifying a plurality of building footprints within the target area; calculating, by a device, a plurality of distance functions that relate the floor-level outline to the each of the plurality of building footprints, the calculating creates a plurality of similarity scores; selecting a building footprint from plurality of building footprints, the selecting based on the plurality of similarity scores; and calculating a final geo-location of the building corresponding to the building footprint.

Abstract: Automating conversion of drawings to indoor maps and plans. One example is a computer-implemented comprising: preprocessing an original CAD drawing to create a modified CAD drawing, a text database containing text from the original CAD drawing, a CAD vector-image of the modified CAD drawing, and a CAD raster-image of the modified CAD drawing; determining a floor depicted in the CAD drawing, the determining results in a floor-level bounding line; sensing furniture depicted on the floor by applying the floor-level bounding line, the CAD vector-image, and the text database to machine-learning algorithms, the sensing results in a plurality of furniture entities and associated location information; identifying each room depicted in the CAD drawing, the identifying results in a plurality of room outlines; and creating an indoor map for the floor by combining the plurality of furniture entities and associated location information with the plurality of room outlines.

Abstract: Automating conversion of drawings to indoor maps and plans. One example is a computer-implemented method of creating an indoor map from a CAD drawing, the method comprising: preprocessing an original CAD drawing to create a modified CAD drawing, a text database containing text from the original CAD drawing, a CAD vector-image of the modified CAD drawing, and a CAD raster-image of the modified CAD drawing; creating a floor-level outline; sensing furniture depicted on the floor, the sensing creates set of furniture entities; identifying a room depicted in the CAD drawing; and creating the indoor map for a floor using the floor-level outline, the room-level outline and the room identity.

Abstract: Automating conversion of drawings to indoor maps and plans. One example is a computer-implemented method of creating an indoor map from a CAD drawing, the method comprising: preprocessing an original CAD drawing to create a modified CAD drawing, a text database containing text from the original CAD drawing, a CAD vector-image of the modified CAD drawing, and a CAD raster-image of the modified CAD drawing; creating a floor-level outline; sensing furniture depicted on the floor, the sensing creates set of furniture entities; identifying a room depicted in the CAD drawing; and creating the indoor map for a floor using the floor-level outline, the room-level outline and the room identity.

Abstract: Automating conversion of drawings to indoor maps and plans. One example is a computer-implemented method comprising: preprocessing an original CAD drawing and thereby creating a modified CAD drawing, a text database containing text from the modified CAD drawing, a CAD vector-image of the modified CAD drawing, and a CAD raster-image of the modified CAD drawing; creating a floor-level bounding line that encircles a floor depicted in the modified CAD drawing, the creating by applying the CAD raster-image, the CAD vector-image, and the text database to a floor-level machine-learning algorithm; applying an active contour model to an initial floor-level segmentation created from the floor-level bounding line to create an intermediate floor outline that delineates the floor; removing drawing-entities from the modified CAD drawing that are a predetermine distance away from the intermediate floor outline to create a final floor outline; and creating an indoor map using the final floor outline.

Abstract: Automating conversion of drawings to indoor maps and plans. One example is a computer-implemented method comprising: preprocessing an original CAD drawing to create a modified CAD drawing, a text database, a CAD vector-image, and a CAD raster-image; determining a floor depicted in the CAD drawing by applying the CAD raster-image, the CAD vector-image, and the text database to a floor-level machine-learning algorithm resulting in a floor-level outline; sensing furniture depicted on the floor by applying the floor-level outline, the CAD vector-image, and the text database to a furniture-level machine-learning algorithm resulting in a set of furniture entities; identifying each room depicted in the CAD drawing by applying the floor-level outline, the furniture entities, CAD vector-image, and the text database to room-level machine-learning algorithm resulting in a plurality of room-level outlines; and creating an indoor map by combining the furniture entities and the plurality of room-level outlines.