# US 2020/0348132 A1 · System, computer program product and method for using a convolution neural network to auto-determine a floor height and floor height elevation of a building

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2020/0348132 A1
- Document ID: US20200348132A1
- Status: Published application
- Kind: A1
- Application number: 16/864,342
- Filed: 2020-05-01
- Published: 2020-11-05
- Earliest claimed: 2019-05-02
- Assignees: CoreLogic Solutions, LLC
- Applicants: Corelogic Solutions, LLC
- Inventors: Wei DU; David Frederick SMITH; Taylor Bryant Brown
- Web page: https://patentgrep.com/patent/US20200348132A1

## CPC

- G06N 3/08: Learning methods
- G01C 5/00: Measuring height; Measuring distances transverse to line of sight; Levelling between separated points; Surveyors' levels (G01C3/20, G01C3/30 take precedence)
- G01C 15/002: {Active optical surveying means (optical plumbing G01C15/105)}
- G06F 18/2413: based on distances to training or reference patterns
- G06N 3/04: Architecture, e.g. interconnection topology
- G06N 3/045: Combinations of networks
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/09: Supervised learning
- G06N 3/096: Transfer learning
- G06V 10/764: using classification, e.g. of video objects
- G06V 10/82: using neural networks
- G06V 20/10: Terrestrial scenes (scenes under surveillance with static cameras G06V20/52; scenes perceived from the exterior of a vehicle G06V20/56; scenes perceived from the interior of a vehicle G06V20/59)
- G06V 20/176: {Urban or other man-made structures}
- G06V 20/20: in augmented reality scenes
- Y02A 10/40: Controlling or monitoring, e.g. of flood or hurricane; Forecasting, e.g. risk assessment or mapping
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G01C: MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY (measuring liquid level G01F; radio navigation, determining distance or velocity by use of propagation effects, e.g. Doppler effects, propagation time, of radio waves, analogous arrangements using other waves G01S)
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- Y02A: TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE

## Abstract

A system, apparatus, computer program product, and method automatically, accurately, and reliably detect floor elevations of improvements (e.g., houses) on parcels of land without human inspection of the property. Fusing aspects of image processing, artificial intelligence, and information on adjacent grade points along the footprint of a specific property, the present computer-based system identifies a first floor height (FFH), and first floor elevation (FFE) of the improvement with respect to the terrain or surface of the parcel of land on which the improvement is located. In turn, by knowing the FFH of the improvement on the parcel, it is possible to improve on conventional methods for estimating flood risk to a property without requiring a personal inspection of the parcel by a human.

## Claims

### Claim 1

A method of detecting a first floor height (FFH) of a first floor of a building relative to a terrain or surface of a parcel of land on which the building is located, the method comprising:
obtaining information on a building footprint of the building on the parcel of land;
applying an image of the building to a CNN-based AI engine that has been trained to identify a first floor of a building from the image;
analyzing the image with the CNN-based AI engine and determining the FFH of the building;
extracting digital elevation map information of the terrain and/or surface from a dataset for the parcel of land;
converting the FFH of the building to a first floor elevation (FFE) from the FFH and the digital elevation map information; and
identifying a location at an adjacent grade point along the building footprint so as to determine an elevation of the location and the FFE of the building at the location.

### Claim 2 (depends on 1)

The method of claim 1, further comprising:
populating a data structure with fields associated with the building, the fields including an address of the parcel of land,
the FFH,
the first floor elevation,
a lowest floor elevation,
the elevation at the adjacent grade point
a latitude and a longitude of the adjacent grade point, and
at least one property characteristic.

### Claim 3 (depends on 1)

The method of claim 1, further comprising:
training the CNN-based AI engine by inputting training images and ground truth images to the CNN-based AI engine, and backpropagating losses so as to establish data extraction parameters for a data extraction network portion of the CNN-based AI engine.

### Claim 4 (depends on 3)

The method of claim 3, wherein the training comprises
training the CNN-based AI engine to detect building types including applying training images of different types of building to the CNN-based AI engine, the different types of buildings including basement-below grade, basement-daylight, and basement-walkout.

### Claim 5 (depends on 3)

The method of claim 3, wherein the training comprises
training the CNN-based AI engine to detect foundation types by applying training images of buildings with different types of foundations.

### Claim 6 (depends on 3)

The method of claim 3, wherein the training comprises
training the CNN-based AI engine to detect a whether a building has a floor below a main floor.

### Claim 7 (depends on 6)

The method of claim 6, wherein the method further comprises
adjusting the FFH to a lower height in response to the CNN-based AI engine having detected a presence of the floor below the main floor.

### Claim 8 (depends on 6)

The method of claim 6, wherein the training of the CNN-based AI engine to detect whether the building has the floor below the main floor includes detecting from the training images at least one of a basement window, a crawlspace, and a back entrance at a level that is below the main floor.

### Claim 9 (depends on 3)

The method of claim 3, wherein the training comprises
training the CNN-based AI engine to detect a main entrance of the building.

### Claim 10 (depends on 9)

The method of claim 9, wherein the training of the CNN-based AI engine to detect the main entrance includes training the CNN-based AI engine with images includes buildings having front steps above ground that lead to the main entrance.

### Claim 11 (depends on 9)

The method of claim 9, wherein the training of the CNN-based AI engine to detect the main entrance includes training the CNN-based AI engine with overhead images of buildings having different walkways leading to the main entrance.

### Claim 12 (depends on 9)

The method of claim 9, wherein the training of the CNN-based AI engine to detect the main entrance includes training the CNN-based AI engine to detect the main entrance from street-view images of buildings having at least one of
entrance lights adjacent to the main entrance,
a doorbell or a door knocker,
a package adjacent to the main entrance;
a door handle, and
a lock.

### Claim 13 (depends on 9)

The method of claim 9, wherein the training of the CNN-based AI engine to detect the main entrance includes training the CNN-based AI engine with images of building layouts.

### Claim 14 (depends on 1)

The method of claim 1, wherein
the analyzing includes determining that a main entrance is not detectable from the image within a predetermined detection threshold, and in response to determining that the main entrance is not detectable, detecting a street segment that aligns with a front portion of the building footprint, and determining the main entrance is in a center of the front portion of the building footprint that is aligned with the street segment.

### Claim 15 (depends on 1)

The method of claim 1, further comprising:
obtaining characteristics of the building recording in a database;
forming an initial assessment of the FFH from the characteristics of the building; and
performing the analyzing the image with the CNN-based AI engine to adjust the initial assessment of the FFH from an analytical model.

### Claim 16 (depends on 1)

The method of claim 1, wherein the training comprises
training the CNN-based AI engine to estimate a height of the main entrance of the building, and the image being a street view image of the building, wherein the image includes at least one of steps and a ramp leading to the main entrance.

### Claim 17 (depends on 1)

The method of claim 1, further comprising:
retrieving base flood elevation (BFE) information for the structure; and
determining a floor elevation of the building that is at least higher than the BFE by a predetermined amount so as to determine whether the building would experience flood damage during an occurrence of a flood that corresponds in size with a base flood.

### Claim 18 (depends on 6)

The method of claim 6, wherein the training of the CNN-based AI engine to detect whether the building has a floor below a main floor comprising:
training the CNN-based AI engine to
detect an interior feature of the building that is associated with the main floor,
determine a distance of the interior feature to a building footprint geometric point, associate the building geometric footprint point with a georeferenced point, and
estimate the FFE using the georeferenced point and the distance of the interior feature to the building footprint geometric point.

### Claim 19

A system for detecting a first floor height of a first floor building relative to a terrain or surface of a parcel of land on which the building is located, the system comprising:
circuitry configured to
obtain information on a building footprint of the building on the parcel of land;
apply an image of the building to a CNN-based AI engine that has been trained to identify the first floor of a building from the image;
analyze the image with the CNN-based AI engine and determine the first floor height of the building;
extract digital elevation map information of the terrain and/or surface from a dataset for the parcel of land;
convert the first floor height of the building to a first floor elevation from the first floor height and the digital elevation map information; and
identify a part of the building footprint at an adjacent grade point along the building footprint so as to detect elevation of the parcel of land and the first floor elevation of the building at the adjacent grade point along the building footprint.

### Claim 20

A non-transitory computer readable storage device that contains computer readable instructions that when executed by a processor cause the processor to perform a method of detecting a first floor height of a first floor of a building relative to a terrain or surface of a parcel of land on which the building is located, the method comprising:
obtaining information on a building footprint of the building on the parcel of land;
applying an image of the building to a CNN-based AI engine that has been trained to identify the first floor of a building from the image;
analyzing the image with the CNN-based AI engine and determining the first floor height of the building;
extracting digital elevation map information of the terrain and/or surface from a dataset for the parcel of land;
converting the first floor height of the building to a first floor elevation from the first floor height and the digital elevation map information; and
identifying a part of the building footprint at an adjacent grade point along the building footprint so as to detect an elevation of the parcel of land and the first floor elevation of the building at an adjacent grade point.
