Variable speed sign value prediction and confidence modeling
Systems, methods, and apparatuses are disclosed for predicting the value of a variable speed sign (VSS) and determining the predicted value's associated confidence level. Highly assisted driving (HAD) vehicles may read or capture images of the VSS. The speed limit values, images, or videos of the VSS are reported and received by a network and database for analysis. A predicted speed limit value is determined for the variable speed sign from at least a portion of the received traffic data. A confidence level is also calculated for the predicted speed limit value for the variable speed sign.
1. A method comprising:
receiving a set of traffic data from at least one vehicle for a defined time period, the received traffic data comprising captured images, videos, or interpreted speed limit values of a variable speed sign;
determining, using a processor, a predicted speed limit value for the variable speed sign from at least a portion of the received traffic data; and
calculating a confidence level for the predicted speed limit value for the variable speed sign.
2. The method of claim 1 , wherein the determining comprises computing a mode, a median, or a mean of the portion of the received traffic data.
3. The method of claim 1 , further comprising:
prior to the determining of the predicted speed limit value, analyzing the captured images or the videos of the received traffic data to provide analyzed speed limit values of the variable speed sign.
4. The method of claim 1 , further comprising:
reporting the predicted speed limit value and the calculated confidence level to at least one vehicle.
5. The method of claim 1 , further comprising:
comparing the calculated confidence level with a threshold confidence level; and
reporting the predicted speed limit value to at least one vehicle when the calculated confidence level is greater than or equal to the threshold confidence level.
6. The method of claim 1 , wherein the calculated confidence level is a factor of at least one of the following:
a) a number of observations in the set of traffic data, wherein a larger number of observations in the set of traffic data increases the confidence level;
b) a similarity of speed limit values in the observations within the set of traffic data, wherein the confidence level increases with an increase in the similarity of the speed limit values and decreases with a decrease in the similarity of the speed limit values;
c) a previously computed confidence level; or
d) a collected time for each received traffic point within the set of traffic data, wherein the calculated confidence level decreases with an increase in a difference in time between the collected time for each received traffic point and a present time.
7. The method of claim 1 , wherein the calculated confidence level, C t , is:
C
t
=
100
(
1
-
M
2
(
S
t
)
M
1
(
S
t
)
)
wherein:
|M 1 (S t )|=total number of elements in set M 1 (S t ) of speed limit values, and
|M 2 (S t )|=total number of elements in set M 2 (S t ) of speed limit values.
8. The method of claim 1 , wherein the calculated confidence level, C t , is:
C
t
=
100
(
1
-
M
2
(
S
t
)
M
1
(
S
t
)
)
e
(
-
1
β
M
1
(
S
t
)
)
wherein:
|M 1 (S t )|=total number of elements in set M 1 (S t ) of speed limit values,
|M 2 (S t )|=total number of elements in set M 2 (S t ) of speed limit values, and
β is a calibration constant for a user's confidence threshold.
9. The method of claim 1 , further comprising:
updating the set of traffic data with traffic data from vehicles in a second time period;
determining an updated predicted speed limit value for the variable speed sign using the real-time data and at least a portion of the received traffic data; and
calculating an updated confidence level for the predicted speed limit value for the variable speed sign.
10. The method of claim 9 , wherein the portion of the received traffic data used in determining the updated predicted speed limit is from a collection time associated with each traffic data point of the received set of traffic data.
11. The method of claim 9 , wherein when there is no additional traffic data in the second time period, the updated confidence level, C t+1 , is:
C
t
+
1
=
C
t
e
-
(
(
t
+
1
)
-
t
α
)
wherein:
C t =the calculated confidence level at time t, and
α=a calibration constant that defines a confidence level exponential decay-rate.
12. A method comprising:
receiving a set of traffic data from at least one vehicle in a first time period, the received traffic data comprising captured images, videos, or interpreted speed limit values of a variable speed sign;
determining, using a processor, a predicted speed limit value for the variable speed sign from a mode of at least a portion of the received traffic data;
calculating a confidence level for the predicted speed limit value for the variable speed sign;
comparing the calculated confidence level with a user's confidence level; and
reporting the predicted speed limit value to a vehicle when the calculated confidence level is greater than or equal to the user's confidence level.
13. An apparatus 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 at least perform:
receive a set of traffic data from at least one vehicle for a defined time period, the received traffic data comprising captured images, videos, or interpreted speed limit values of a variable speed sign;
determine a predicted speed limit value for the variable speed sign from a portion of the received traffic data; and
calculate a confidence level for the predicted speed limit value for the variable speed sign.
14. The apparatus of claim 13 , wherein the predicted speed limit value is determined through a computation of a mode, a median, or a mean of the portion of the received traffic data.
15. The apparatus of claim 13 , wherein the at least one memory and the computer program code are configured to cause the apparatus to further perform:
report the predicted speed limit value and the calculated confidence level to a vehicle.
16. The apparatus of claim 13 , wherein the at least one memory and the computer program code are configured to cause the apparatus to further perform:
compare the calculated confidence level with a threshold confidence level; and
report the predicted speed limit value to at least one vehicle when the calculated confidence level is greater than or equal to the threshold confidence level.
17. The apparatus of claim 13 , wherein the calculated confidence level, C t , is:
C
t
=
100
(
1
-
M
2
(
S
t
)
M
1
(
S
t
)
)
wherein:
|M 1 (S t )|=total number of elements in set M 1 (S t ) of speed limit values, and
|M 2 (S t )|=total number of elements in set M 2 (S t ) of speed limit values.
18. The apparatus of claim 13 , wherein the calculated confidence level, C t , is:
C
t
=
100
(
1
-
M
2
(
S
t
)
M
1
(
S
t
)
)
e
(
-
1
β
M
1
(
S
t
)
)
wherein:
|M 1 (S t )|=total number of elements in set M 1 (S t ) of speed limit values,
|M 2 (S t )|=total number of elements in set M 2 (S t ) of speed limit values, and
β is a calibration constant for a user's confidence threshold.
19. The apparatus of claim 13 , wherein the at least one memory and the computer program code are configured to cause the apparatus to further perform:
update the set of traffic data with traffic data from vehicles in a second time period;
determine an updated predicted speed limit value for the variable speed sign using the real-time data and at least a portion of the received traffic data; and
calculate an updated confidence level for the predicted speed limit value for the variable speed sign.
20. The apparatus of claim 19 , wherein the portion of the received traffic data used in determining the updated predicted speed limit is from a collection time associated with each traffic data point of the received set of traffic data.
21. The apparatus of claim 19 , wherein when there is no additional traffic data in the second time period, the updated confidence level, C t+1 , is:
C
t
+
1
=
C
t
e
-
(
(
t
+
1
)
-
t
α
)
wherein:
C t =the calculated confidence level at time t, and
α is a calibration constant that defines a confidence level exponential decay-rate.
22. The apparatus of claim 13 , wherein the calculated confidence level is a factor of at least one of the following:
a) a number of observations in the set of traffic data, wherein a larger number of observations in the set of traffic data increases the confidence level;
b) a similarity of speed limit values in the observations within the set of traffic data, wherein the confidence level increases with an increase in the similarity of the speed limit values and decreases with a decrease in the similarity of the speed limit values;
c) a previously computed confidence level; or
d) a collected time for each received traffic point within the set of traffic data, wherein the calculated confidence level decreases with an increase in a difference in time between the collected time for each received traffic point and a present time.