Method and device for determining location of object in lane map

1 . A method of determining a location of an object on a lane map, the method comprising:

acquiring image data captured by a camera mounted on a vehicle;

generating feature information regarding a plurality of features included in the image data by performing certain processing on the image data;

determining a plurality of features corresponding to a same object in a plurality of pieces of image data, on the basis of a degree of correlation determined on the basis of an arrangement relationship and a size relationship between the plurality of features included in the image data;

determining locations of objects corresponding to the plurality of features on the lane map, on the basis of location information of the vehicle on the lane map and feature location information on the plurality of pieces of image data; and

displaying a final location of an object determined as the same object on the lane map by aggregating the determined locations of the objects corresponding to the plurality of features,

the determining of the plurality of features corresponding to the same object further comprising:

searching for neighbor feature existing around each of the first feature of the first image data and the second feature of the second image data among the plurality of image data;

registering a graph including the first feature and the second feature as key features for each of the first neighbor feature and the second neighbor feature based on a result of searching for the first neighbor feature and the second neighbor feature; and

determining the first feature of the first image data and the second feature of the second image data indicates the same object based on the similarity between the registered key features.

2 . The method of claim 1 , wherein the object comprises at least one of a traffic light, a road sign, a road marking, and a bump.

3 . The method of claim 1 , wherein the determining of the plurality of features corresponding to the same object comprises determining a plurality of features corresponding to a same object in a plurality of pieces of image data by aggregating features within the plurality of pieces of image data.

4 . The method of claim 1 , wherein the determining of the plurality of features corresponding to the same object comprises determining a plurality of features indicating a same object in a plurality of pieces of image data by using a graph-matching method.

5 . The method of claim 4 , further comprising:

selecting a first feature of first image data and a second feature of second image data;

searching for other features around the first feature and the second feature;

when a second neighbor feature having a similar attribute to a first neighbor feature around the first feature is present around the second feature, registering a graph including, as key features of the first neighbor feature and the second neighbor feature, respectively, the first feature and the second feature;

repeating, on all features around the first feature, the searching for and the registering;

after attempting to construct a graph for all features around the first feature, calculating a similarity between the first feature and the second feature, on the basis of a number of features and a similarity between registered key features; and

determining that the second feature of the second image data having a highest similarity to the first feature of the first image data indicates a same object, by performing, on remaining features other than the second feature of the second image data, the searching for, the registering of the graph, the repeating, and the calculating of the similarity.

6 . The method of claim 5 , further comprising determining that features included in the graph have a same depth.

7 . The method of claim 6 , further comprising, when performing post-processing on a certain feature, performing the same post-processing on remaining features included in the graph, which have the same depth.

8 . The method of claim 5 , wherein the attribute comprises at least one of a class, an aspect ratio of a feature, a distance, and an angle.

9 . The method of claim 1 , further comprising acquiring location information of the vehicle on the lane map by using a location identification apparatus connected to the vehicle.

10 . The method of claim 1 , further comprising:

displaying, on the lane map, first locations of first group objects determined from a first trajectory of the vehicle which travels on a certain road;

displaying, on the lane map, second locations of second group objects determined from a second trajectory of the vehicle which travels on the certain road; and

determining a same object in the first group objects and the second group objects, aggregating locations of objects determined to be the same object, and displaying final locations of the respective objects on the lane map.

11 . The method of claim 10 , wherein the displaying of the final locations of the respective objects on the lane map comprises constructing a first graph by using the first group objects, constructing a second graph by using the second group objects, and determining a same object in the first group objects and the second group objects via similarity comparison between the first graph and the second graph.

12 . An apparatus for determining a location of an object on a lane map, the apparatus comprising:

a memory storing at least one program; and

a processor configured to perform a calculation by executing the at least one program, wherein the processor is configured to: acquire image data captured by a camera mounted on a vehicle; generate feature information regarding a plurality of features included in the image data by performing certain processing on the image data; determine a plurality of features corresponding to a same object in a plurality of pieces of image data, on the basis of a degree of correlation determined on the basis of an arrangement relationship and a size relationship between the plurality of features included in the image data; determine locations of objects corresponding to the plurality of features on the lane map, on the basis of location information of the vehicle on the lane map and feature location information on the plurality of pieces of image data; and display a final location of an object determined as the same object on the lane map by aggregating the determined locations of the objects corresponding to the plurality of features,

where in search for neighbor feature existing around each of the first feature of the first image data and the second feature of the second image data among the plurality of image data; register a graph including the first feature and the second feature as key features for each of the first neighbor feature and the second neighbor feature based on a result of searching for the first neighbor feature and the second neighbor feature; and determine the first feature of the first image data and the second feature of the second image data indicates the same object based on the similarity between the registered key features.

13 . A non-transitory computer-readable recording medium having recorded thereon a program for executing the method of claim 1 on a computer.