Method and apparatus for matching heterogeneous feature spaces

Patent metadata

Field	Value
Publication number	US-11507775-B2
Application number	US-201816210844-A
Country	US
Kind code	B2
Filing date	Dec 5, 2018
Priority date	Dec 5, 2018
Publication date	Nov 22, 2022
Grant date	Nov 22, 2022

Abstract

An approach is provided for fully-automated learning to match heterogeneous feature spaces for mapping. The approach involves determining a first feature space comprising first features and a second feature space comprising second features, and classified by a feature detector into a first attribution category and a second attribution category, respectively. The approach further involves calculating a first similarity score for the first feature space based on a first distance metric applied to the first features, and a second similarity score for the second feature space based on a second distance metric applied to the second features. The approach also involves determining a transformation space comprising a first weight to be applied to the first similarity score and a second weight to be applied to the second similarity score based on matching the first attribution category and the second attribution category.

First claim

What is claimed is:

1. A computer-implemented method for generating a transformation space to compare feature spaces for digital map making comprising:
determining a first feature space represented by a first vector or matrix comprising one or more first mappable geographic features of a geographic area detected from raw sensor data and classified by a feature detector into a first attribution category, wherein the one or more first mappable geographic features of the geographic area are stationary;
determining a second feature space represented by a second vector or matrix comprising one or more second mappable geographic features of the geographic area detected from the raw sensor data and classified by the feature detector into a second attribution category, wherein the one or more second mappable geographic features of the geographic area are stationary;
calculating a first similarity score for the first feature space based on a first distance metric applied to the one or more first mappable geographic features, and a second similarity score for the second feature space based on a second distance metric applied to the one or more second mappable geographic features; and
determining the transformation space comprising a first weight to be applied to the first similarity score and a second weight to be applied to the second similarity score based on matching one or more combinations of the first attribution category and the second attribution category.

2. The method of claim 1 , further comprising:
processing the one or more first mappable geographic features, the one or more second mappable geographic features, or a combination thereof using the transformation space to create the digital map.

3. The method of claim 1 , wherein the first weight and the second weight perform a direct linear transformation (DLT) of the first feature space and the second feature space.

4. The method of claim 1 , further comprising:
tuning the first weight, the second weight, or a combination thereof based, at least in part, on the first similarity score, the second similarity score, or a combination thereof.

5. The method of claim 4 , wherein tuning the first weight and the second weight further comprising:
maximizing a total score based, at least in part, on the matching of the one or more combinations of the first attribution category and the second attribution category.

6. The method of claim 4 , wherein tuning the first weight and the second weight further comprising:
minimizing the total score based, at least in part, on determining the one or more combination of the first attribution category and the second attribution category is incorrect.

7. The method of claim 6 , further comprising:
automatically align the first feature space and the second feature space based, at least in part, on the total score.

8. The method of claim 7 , further comprising:
storing the first weight, the second weight, indices of the aligned feature spaces, or a combination thereof for processing additional features detected from the raw sensor data.

9. The method of claim 1 , wherein the first attribution category and the second attribution category comprises one or more lane lines, one or more road signs, terrain features, drivable surfaces, or a combination thereof.

10. The method of claim 1 , wherein determining the transformation space further comprising:
searching for a plurality of subset of the first attribution category and the second attribution category in the first feature space and the second feature space, respectively, to maximize the total score, and wherein the search comprises a discrete grid search.

11. An apparatus for generating a transformation space to compare feature spaces for digital map making, comprising:
at least one processor; and
at least one memory including computer program code for one or more programs,
the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to perform at least the following,
determine a first feature space represented by a first vector or matrix comprising one or more first mappable geographic features of a geographic area detected from raw sensor data and classified by a feature detector into a first attribution category, wherein the one or more first mappable geographic features of the geographic area are stationary;
determine a second feature space represented by a second vector or matrix comprising one or more second mappable geographic features of the geographic area detected from the raw sensor data and classified by the feature detector into a second attribution category, wherein the one or more first mappable geographic features of the geographic area are stationary;
calculate a first similarity score for the first feature space based on a first distance metric applied to the one or more first mappable geographic features, and a second similarity score for the second feature space based on a second distance metric applied to the one or more second mappable geographic features; and
determine the transformation space comprising a first weight to be applied to the first similarity score and a second weight to be applied to the second similarity score based on matching one or more combinations of the first attribution category and the second attribution category.

12. The apparatus of claim 11 , further comprising:
process the one or more first mappable geographic features, the one or more second mappable geographic features, or a combination thereof using the transformation space to create the digital map.

13. The apparatus of claim 11 , wherein the first weight and the second weight perform a direct linear transformation (DLT) of the first feature space and the second feature space.

14. The apparatus of claim 11 , further comprising: tune the first weight, the second weight, or a combination thereof based, at least in part, on the first similarity score, the second similarity score, or a combination thereof.

15. The apparatus of claim 14 , wherein tuning the first weight and the second weight further comprising:
maximize a total score based, at least in part, on the matching of the one or more combinations of the first attribution category and the second attribution category.

16. The apparatus of claim 14 , wherein tuning the first weight and the second weight further comprising:
minimize the total score based, at least in part, on determining the one or more combination of the first attribution category and the second attribution category is incorrect.

17. The apparatus of claim 16 , further comprising:
align automatically the first feature space and the second feature space based, at least in part, on the total score.

18. A non-transitory computer-readable storage medium for generating a transformation space to compare feature spaces for digital map making, carrying one or more sequences of one or more instructions which, when executed by one or more processors, cause an apparatus to at least perform the following steps:
determining a first feature space represented by a first vector or matrix comprising one or more first mappable geographic features of a geographic area detected from raw sensor data and classified by a feature detector into a first attribution category, wherein the one or more first mappable geographic features of the geographic area are stationary;
determining a second feature space represented by a second vector or matrix comprising one or more second mappable geographic features of the geographic area detected from the ra

Assignees

• Here Global Bv

Inventors

• Viswanathan Anirudh

Classifications

• G06F18/2413

  based on distances to training or reference patterns · CPC title

• G06F18/22

  Matching criteria, e.g. proximity measures · CPC title

• G06V20/56 Primary

  exterior to a vehicle by using sensors mounted on the vehicle · CPC title

• G06V20/182

  Network patterns, e.g. roads or rivers · CPC title

• G06V20/582

  of traffic signs · CPC title