Identifying and tracking convective weather cells

Patent metadata

Field	Value
Publication number	US-9613269-B2
Application number	US-201514631585-A
Country	US
Kind code	B2
Filing date	Feb 25, 2015
Priority date	Mar 31, 2014
Publication date	Apr 4, 2017
Grant date	Apr 4, 2017

Abstract

A method for tracking weather cells from a moving platform includes receiving, from a detection and ranging system, reflectivity data sampled for a volume of space and generating a feature map based on the reflectivity data, wherein the feature map is a representation of the volume of space that indicates locations with significant weather and generating a first segmented feature map based on the feature map that identifies the location and spatial extent of individual weather cells. The method further includes translating the first segmented feature map and a second segmented feature map, generated from data collected at a different point in time and/or space, to a common frame of reference and comparing the first segmented feature map to the second segmented feature map. The method further includes creating one or more track hypotheses based on the comparison of the first segmented feature map and the second segmented feature map.

First claim

What is claimed is:

1. A method comprising:
receiving, by a processor and from a detection and ranging system, reflectivity data sampled for a volume of space;
generating, by the processor, a feature map based on the reflectivity data, wherein the feature map is a representation of the volume of space that indicates the spatial extent of objects to be tracked;
generating, by the processor, a first segmented feature map based on the feature map;
translating, by the processor, the first segmented feature map and a second segmented feature map to a common frame of reference;
identifying, by the processor, one or more weather cells in the first segmented feature map, wherein each weather cell of the one or more weather cells corresponds to an area sampled with the detection and ranging system that has a VIR higher than a classification threshold;
creating, by the processor, a cell mask for a weather cell of the one or more weather cells in the first segmented feature map;
overlaying, by the processor, the cell mask from the first segmented feature map on one or more cell masks from the second segmented feature map; and
generating, by the processor, a track hypothesis based on a comparison of the cell mask from the first segmented feature map to the one or more cell masks from the second segmented feature map.

2. The method of claim 1 , further comprising:
creating, by the processor, a track history based on comparing the one or more weather cells in the first segmented feature map to one or more weather cells in the second segmented feature map, wherein the track hypothesis is further based on the track history; and
storing the track history for each of the one or more cell mask as one of a single cell track or a multi-cell track.

3. The method of claim 2 , further comprising updating the track history based on a most likely track hypothesis.

4. The method of claim 3 , wherein updating the track history based on a most likely track hypothesis comprises:
determining an amount of overlap for each identified weather cell between the first segmented feature map and the second segmented feature map; and
selecting the most likely track based on the weather cells with the highest amount of overlap.

5. The method of claim 1 , wherein generating the feature map comprises:
building a vertically integrated reflectivity feature map based on the reflectivity data;
generating a classified feature map based on applying a classification threshold to the vertically integrated reflectivity feature map; and
conditioning the classified feature map using one or more of a smoothing filter and a shape generalization filter.

6. The method of claim 5 , wherein conditioning the classified feature map comprises sparsely applying a morphological filter to the classified feature map.

7. The method of claim 1 , wherein generating the first segmented feature map comprises applying a classification threshold to the reflectivity data to identify weather cells.

8. The method of claim 7 , wherein creating the track hypothesis based on the comparison of the first segmented feature map and the second segmented feature map comprises creating a track hypothesis pool including one or more tracks for each identified weather cell, the method further comprising trimming expired tracks from the track hypothesis pool.

9. The method of claim 1 , wherein the reflectivity data is gathered at a first time instance, the second segmented feature map is based on reflectivity data gathered at a second time instance, and the second time instance precedes the first time instance.

10. The method of claim 1 , wherein the first segmented feature map includes a map of weighted centroid locations for each weather cell of the one or more weather cells, and wherein generating the track comparison is further based on a comparison of the weighted centroid locations in the first segmented feature map to weighted centroid locations in the second segmented feature map.

11. A system comprising:
a detection and ranging system configured to receive reflectivity data; and
one or more processors communicatively coupled to the detection and ranging system, wherein the one or more processors are configured to:
receive, from the detection and ranging system, the reflectivity data sampled for a volume of space;
build a vertically integrated reflectivity feature map based on the reflectivity data, wherein the vertically integrated reflectivity feature map includes one or more weather cells and a representation of the volume of space that indicates spatial extent of weather to be tracked;
generate a first classified feature map based on applying a classification threshold to the vertically integrated reflectivity feature map;
condition the first classified feature map by sparsely applying a morphological filter to the classified feature map;
translate the first classified feature map and a second classified feature map to a common frame of reference;
compare the first classified feature map to the second classified feature map; and
create a track hypothesis based on a comparison of the one or more weather cells in the first classified feature map to one or more weather cells in the second classified feature map.

12. The system of claim 11 , wherein the one or more processors are further configured to:
identify one or more weather cells in the first classified feature map, wherein each weather cell corresponds to an area sampled with the detection and ranging system that has a reflectivity higher than a classification threshold; and
create a track history based on comparing the one or more weather cells in the first classified feature map to one or more weather cells in the second classified feature map, wherein the track hypothesis is further based on the track history.

13. The system of claim 12 , wherein the one or more processors are further configured to create a cell mask for each of the one or more weather cells in the first classified feature map, and wherein to create the track history, the one or more processors are further configured to:
overlay the cell masks from the first classified feature map on cell masks from the second classified feature map;
generate a track hypothesis based on a comparison of the cell masks from the first classified feature map to cell masks from the second classified feature map; and
store the track history for each of the one or more cell mask as one of a single cell track or a multi-cell track.

14. The system of claim 13 , wherein the one or more processors are further configured to:
update the track history based on a most likely track hypothesis, wherein to update the track history based on a most likely track hypothesis the one or more processors are further configured to:
determine an amount of overlap for each identified weather cell between the first classified feature map and the second classified feature map; and
select a most likely track based on the weather cells with the highest amount of overlap.

15. The system of claim 11 , wherein the first classified feature map includes a map of weighted centroid locations for each weather cell of the one or more weather cells, and wherein the one or more processors are configured to create the track comparison based on a comparison of the weighted centroid locations in the first classified feature map to one or more weighted centroid locations in the second classified feature map.

16. The system of claim 11 , wherein to process the feature map into a first segmented feature map, the one or more processors are further configured to apply a class

Assignees

• Honeywell Int Inc

Inventors

• Kilty Brennan
• Christianson Paul E
• Jensen Robert J

Classifications

• G06T7/10

  Segmentation; Edge detection (motion-based segmentation G06T7/215) · CPC title

• G01S13/955 Primary

  mounted on satellite · CPC title

• G01S13/953

  mounted on aircraft · CPC title

• G06T3/00

  Geometric image transformations in the plane of the image · CPC title

• G01S13/951

  ground based · CPC title