Object recognition device
An object recognition device mountable to a vehicle includes a tracking processing unit including a first association processing unit, a second association processing unit, and a vehicle likelihood calculation unit. The first association processing unit extracts, for each tracking target and based on a positional relationship, a provisionally associated cluster which is a cluster assumed to indicate the same target as the tracking target. The second association processing unit calculates a posterior confidence indicating a probability that when it is assumed that the provisionally associated cluster and the tracking target indicate the same target, the target is a vehicle, determines whether to perform association based on the calculated posterior confidence, and calculates the posterior confidence based on the posterior confidence calculated for the tracking target in the previous measurement cycle and on a cluster vehicle likelihood of the provisionally associated cluster calculated by the vehicle likelihood calculation unit.
1 . An object recognition device mountable to a vehicle, the object recognition device comprising:
a radar unit configured to transmit a transmission wave toward a predetermined region around the vehicle and receive a reflected wave of the transmission wave for each measurement cycle set in advance;
a characteristic determination unit configured to determine, for each of reflection points based on a result of reception of the reflected wave, at least one of whether the corresponding one of the reflection points has a vehicle characteristic and whether the corresponding one of the reflection points has a non-vehicle characteristic, the vehicle characteristic being a characteristic observed in a point at which the transmission wave is reflected by a vehicle, and the non-vehicle characteristic being a characteristic observed in a point at which the transmission wave is reflected by an object other than a vehicle;
a cluster forming unit configured to perform clustering of the reflection points to form a cluster which is a set of the reflection points; and
a tracking processing unit configured to perform association of each of tracking targets with a cluster corresponding to a respective one of each of the tracking targets and recognize the cluster as the corresponding one of the tracking targets, the tracking targets each being a target as an object to be tracked, wherein
the tracking processing unit includes
a first association processing unit configured to extract a provisionally associated cluster, in a current measurement cycle, for each of the tracking targets in a previous measurement cycle, wherein the provisionally associated cluster is a cluster assumed to correspond to a respective one of each of the tracking targets, wherein the tracking targets have not yet been subjected to tracking in the current measurement cycle, and the provisionally associated cluster has not yet been associated with a corresponding tracking target in the current measurement cycle, wherein the provisionally associated cluster is extracted based on a positional relationship,
a second association processing unit configured to
calculate a posterior confidence in the current measurement cycle, the posterior confidence indicating a probability that, when it is assumed that the provisionally associated cluster and the corresponding one of the tracking targets indicate the same target, the target is a vehicle, and
determine, based on the calculated posterior confidence, whether to perform the association in which the provisionally associated cluster is associated with the corresponding tracking target,
a target recognition unit configured to recognize the provisionally associated cluster as the corresponding one of the tracking targets in response to a determination by the second association processing unit to perform the association, and
a vehicle likelihood calculation unit configured to calculate, for each cluster and based on the cluster and on a result of determination by the characteristic determination unit, a cluster vehicle likelihood that indicates a likelihood that the cluster indicates a vehicle, and
the second association processing unit is configured to calculate the posterior confidence in the current measurement cycle based on the posterior confidence calculated for the corresponding one of the tracking targets in the previous measurement cycle and on the cluster vehicle likelihood of the cluster calculated by the vehicle likelihood calculation unit.
2 . The object recognition device according to claim 1 , wherein
the vehicle likelihood calculation unit is configured to calculate the cluster vehicle likelihood for each cluster based on at least one of a presence ratio of one or more of the reflection points having the vehicle characteristic in the cluster and a presence ratio of one or more of the reflection points having the non-vehicle characteristic in the cluster.
3 . The object recognition device according to claim 1 , wherein
the characteristic determination unit is configured to determine, for each of the reflection points, whether the corresponding one of the reflection points has the vehicle characteristic and whether the corresponding one of the reflection points has the non-vehicle characteristic.
4 . The object recognition device according to claim 1 , wherein
the tracking processing unit further includes a target management unit configured to exclude, from objects to be tracked, a tracking target for which there is no provisionally associated cluster determined to perform the association by the second association processing unit.
5 . The object recognition device according to claim 1 , wherein
the second association processing unit is configured to
in response to the calculated posterior confidence being higher than a threshold, register the provisionally associated cluster as an association candidate for the corresponding one of the tracking targets.
6 . The object recognition device according to claim 1 , wherein
the second association processing unit is configured to calculate the posterior confidence for all provisionally associated clusters extracted for each of the tracking targets.
7 . An object recognition device mountable to a vehicle, the object recognition device comprising:
a processor;
a non-transitory computer-readable storage medium; and
a set of computer-executable instructions stored on the non-transitory computer-readable storage medium that cause the processor to:
receive reflection point information of reflection points obtained by transmitting a transmission wave toward a predetermined region around the vehicle and receiving a reflected wave of the transmission wave for each measurement cycle set in advance;
determine, for each of the reflection points based on a result of reception of the reflected wave, at least one of whether the corresponding one of the reflection points has a vehicle characteristic and whether the corresponding one of the reflection points has a non-vehicle characteristic, the vehicle characteristic being a characteristic observed in a point at which the transmission wave is reflected by a vehicle, and the non-vehicle characteristic being a characteristic observed in a point at which the transmission wave is reflected by an object other than a vehicle;
perform clustering of the reflection points to form a cluster which is a set of the reflection points; and
perform association of each of tracking targets with a cluster corresponding to a respective one of each of the tracking targets and recognize the cluster as the corresponding one of the tracking targets, the tracking targets each being a target as an object to be tracked, wherein performing of the association of each of tracking targets includes:
extracting a provisionally associated cluster, in a current measurement cycle, for each of the tracking targets in a previous measurement cycle, wherein the provisionally associated cluster is a cluster assumed to correspond to a respective one of each of the tracking targets, wherein the tracking targets have not yet been subjected to tracking in the current measurement cycle, and the provisionally associated cluster has not yet been associated with a corresponding tracking target in the current measurement cycle, wherein the provisionally associated cluster is extracted based on a positional relationship,
calculating a posterior confidence in the current measurement cycle, the posterior confidence indicating a probability that, when it is assumed that the provisionally associated cluster and the corresponding one of the tracking targets indicate the same target, the target is a vehicle, and
determining, based on the calculated posterior confidence, whether to perform the association in which the provisionally associated cluster is associated with the corresponding tracking target,
recognizing the provisionally associated cluster as the corresponding one of the tracking targets in response to a determination to perform the association, and
calculating, for each cluster and based on the cluster and on a result of determination of whether the corresponding one of the reflection points has the vehicle characteristic, a cluster vehicle likelihood that indicates a likelihood that the cluster indicates a vehicle, wherein
the posterior confidence in the current measurement cycle is calculated based on the posterior confidence calculated for the corresponding one of the tracking targets in the previous measurement cycle and on the cluster vehicle likelihood of the cluster.
8 . The object recognition device according to claim 7 , wherein
the set of computer-executable instructions further cause the processor to
calculate the cluster vehicle likelihood for each cluster based on at least one of a presence ratio of one or more of the reflection points having the vehicle characteristic in the cluster and a presence ratio of one or more of the reflection points having the non-vehicle characteristic in the cluster.
9 . The object recognition device according to claim 7 , wherein
the set of computer-executable instructions further cause the processor to
determine, for each of the reflection points, whether the corresponding one of the reflection points has the vehicle characteristic and whether the corresponding one of the reflection points has the non-vehicle characteristic.
10 . The object recognition device according to claim 7 , wherein
the set of computer-executable instructions further cause the processor to
exclude, from objects to be tracked, a tracking target for which there is no provisionally associated cluster determined to perform the association.
11 . The object recognition device according to claim 7 , wherein
the set of computer-executable instructions further cause the processor to
in response to the calculated posterior confidence being higher than a threshold, register the provisionally associated cluster as an association candidate for the corresponding one of the tracking targets.
12 . The object recognition device according to claim 7 , wherein
the set of computer-executable instructions further cause the processor to
calculate the posterior confidence for all provisionally associated clusters extracted for each of the tracking targets.
13 . The object recognition device according to claim 1 , wherein
the second association processing unit is configured to:
determine whether the calculated posterior confidence is higher than a posterior confidence threshold; and
in response to the calculated posterior confidence being higher than the posterior confidence threshold, register the provisionally associated cluster as an association candidate for the corresponding one of the tracking targets.
14 . The object recognition device according to claim 1 , wherein
the second association processing unit is configured to:
determine whether the calculated posterior confidence is lower than or equal to a posterior confidence threshold; and
in response to the calculated posterior confidence being lower than or equal to the posterior confidence threshold, refrain from associating the provisionally associated cluster with the corresponding one of the tracking targets.
15 . The object recognition device according to claim 7 , wherein
the set of computer-executable instructions further cause the processor to:
determine whether the calculated posterior confidence is higher than a posterior confidence threshold; and
in response to the calculated posterior confidence being higher than the posterior confidence threshold, register the provisionally associated cluster as an association candidate for the corresponding one of the tracking targets.
16 . The object recognition device according to claim 7 , wherein
the set of computer-executable instructions further cause the processor to:
determine whether the calculated posterior confidence is lower than or equal to a posterior confidence threshold; and
in response to the calculated posterior confidence being lower than or equal to the posterior confidence threshold, refrain from associating the provisionally associated cluster with the corresponding one of the tracking targets.