Techniques for analyzing and mapping geographic areas using UAS
The present disclosure provides systems and methods for extracting information from image data captured using an unmanned aircraft system (UAS). The disclosed systems and methods may capture image data using an imaging device mounted to the UAS, which may include a plurality of images of an environment. Each image may be associated with a particular area of the environment and a timestamp. The imagery may be used to construct one or more multi-temporal orthomosaics. The disclosed systems and methods may generate information representative of a dynamic evolution of an event impacting the environment over time based on the multi-temporal orthomosaic(s). The systems and methods may also calculate a rate of spread (ROS) of the event based on the information representative dynamic evolution of an event impacting the environment over time. In an additional aspect, techniques for synchronizing image data captured by the UAS to calibrated image data are disclosed.
1 . A method comprising:
collecting, by an unmanned aircraft system, a plurality of repeat-pass time-sequential images of an environment, wherein each image of the plurality of repeat-pass time-sequential images is associated with a particular area of the environment and a timestamp;
constructing, by one or more processors, one or more multi-temporal orthomosaics based on the plurality of repeat-pass time-sequential images;
generating, by the one or more processors, information representative of a dynamic evolution of a fire impacting the environment over time based on the one or more multi-temporal orthomosaics, wherein the generating the information representative of the dynamic evolution of the fire includes:
generating a grid of pixels for each of the one or more multi-temporal orthomosaics;
classifying, for each pixel of the grid of pixels, whether the pixel represents a fire front based on the pixel meeting an intensity variance threshold; and
generating, by performing a delineation on those pixels of the grid of pixels classified to represent the fire front, a fire front curve; and
calculating, by the one or more processors, a rate of spread (ROS) of the fire based on the information representative of the dynamic evolution of the fire impacting the environment over time.
2 . The method of claim 1 , the method further comprising:
extracting a fire front from each of the one or more multi-temporal orthomosaics to generate a plurality of fire fronts;
generating the fire front curve based on the plurality of fire fronts; and
determining the ROS based at least in part on the fire front curve.
3 . The method of claim 1 , wherein the generating the fire front curve comprises:
extracting the pixels classified as representing the fire front to produce an initial fire front; and
performing the delineation on the initial fire front to obtain the fire front curve.
4 . The method of claim 1 , wherein the intensity variance threshold is determined based on a pixel intensity metric obtained from the grid of pixels.
5 . The method of claim 1 , wherein intensity variance threshold is determined based on a range of pixel intensity values of the grid of pixels.
6 . The method of claim 1 , wherein the fire front curve is determined based on a plurality of fire fronts extracted from the one or more multi-temporal orthomosaics, and wherein the ROS is determined based on the fire front curve.
7 . The method of claim 1 , further comprising generating a fire evolution map based on the fire front curve.
8 . The method of claim 1 , further comprising registering an area of the one or more multi-temporal orthomosaics with respect to a calibrated image.
9 . The method of claim 8 , further comprising enhancing the calibrated image based on image content included in at least one of the one or more multi-temporal orthomosaics.
10 . The method of claim 1 , wherein the intensity variance threshold is determined based on a pixel intensity metric obtained from the grid of pixels and based on a range of pixel intensity values of the grid of pixels.
11 . A non-transitory computer-readable storage medium storing instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising:
receiving a plurality of repeat-pass time-sequential images of an environment collected by an unmanned aircraft system, wherein each image of the plurality of repeat-pass time-sequential images is associated with a particular area of the environment and a timestamp;
constructing one or more multi-temporal orthomosaics based on the plurality of repeat-pass time-sequential images;
generating information representative of a dynamic evolution of a fire impacting the environment over time based on the one or more multi-temporal orthomosaics, wherein the generating the information representative of the dynamic evolution of the fire includes:
generating a grid of pixels for each of the one or more multi-temporal orthomosaics;
classifying, for each pixel of the grid of pixels, whether the pixel represents a fire front based on the pixel meeting an intensity variance threshold; and
generating, by performing a delineation on those pixels of the grid of pixels classified to represent the fire front, a fire front curve; and
calculating, by the one or more processors, a rate of spread (ROS) of the fire based on the information representative of the dynamic evolution of the fire impacting the environment over time.
12 . The non-transitory computer-readable storage medium of claim 11 , the operations further comprising:
extracting a fire front from each of the one or more multi-temporal orthomosaics to generate a plurality of fire fronts;
generating the fire front curve based on the plurality of fire fronts; and
determining the ROS based at least in part on the fire front curve.
13 . The non-transitory computer-readable storage medium of claim 12 , wherein the generating the fire front curve comprises:
extracting the pixels classified as representing the fire front to produce an initial fire front; and
performing the delineation on the initial fire front to obtain the fire front curve.
14 . The non-transitory computer-readable storage medium of claim 11 , wherein the intensity variance threshold is determined based on a pixel intensity metric obtained from the grid of pixels and based on a range of pixel intensity values of the grid of pixels.
15 . The non-transitory computer-readable storage medium of claim 11 , wherein the fire front curve is determined based on a plurality of fire fronts extracted from the one or more multi-temporal orthomosaics, and wherein the ROS is determined based on the fire front curve.
16 . The non-transitory computer-readable storage medium of claim 11 , the operations further comprising generating a fire evolution map based on the fire front curve.
17 . The non-transitory computer-readable storage medium of claim 11 , the operations further comprising registering an area of the one or more multi-temporal orthomosaics with respect to a calibrated image.
18 . The non-transitory computer-readable storage medium of claim 17 , the operations further comprising enhancing the calibrated image based on image content included in at least one of the one or more multi-temporal orthomosaics.