Railroad light detection
The technology relates to controlling a vehicle based on a railroad light's activation status. In one example, one or more processors receive images of a railroad light. The one or more processors determine, based on the images of the railroad light, the illumination status of a pair of lights of the railroad light over a period of time as the vehicle approaches the railroad light. The one or more processors determine based on the illumination status of the pair of lights, a confidence level, wherein the confidence level indicates the likelihood the railroad light is active. The vehicle is controlled as it approaches the railroad light based on the confidence level.
1. A computing device comprising:
one or more processors, wherein the one or more processors are configured to:
receive images of a pair of lights of a railroad light assembly;
analyze the images of the pair of lights of the railroad light assembly in order to modify, based on one or more illumination patterns of the pair of lights of the railroad light assembly, a confidence level that indicates a likelihood that the railroad light assembly is active;
control a vehicle to autonomously take a first action when the modified confidence level satisfies a first threshold condition; and
control the vehicle to autonomously take a second action, different from the first action, when the modified confidence level does not satisfy the first threshold condition.
2. The computing device of claim 1 , wherein the one or more processors are further configured to determine, based on one or more road features, whether the vehicle has reached a location that is a predefined threshold distance away from the railroad light assembly.
3. The computing device of claim 2 , wherein the one or more processors are further configured to analyze the images of the pair of lights of the railroad light assembly when the vehicle is determined to have reached the location that is the predefined threshold distance away from the railroad light assembly.
4. The computing device of claim 1 , wherein the one or more processors are further configured to:
update, based on the modified confidence level, a trajectory of the vehicle; and
control the vehicle as the vehicle approaches the railroad light assembly based on the updated trajectory.
5. The computing device of claim 1 , wherein the one or more processors are further configured to:
categorize image data of the images of the pair of lights of the railroad light assembly by color and shape;
determine a brightness level of portions of the image data categorized as red and circular; and
determine whether the brightness level satisfies or fails to satisfy a second threshold condition indicating that at least one light of the pair of lights is illuminated or not illuminated, respectively.
6. The computing device of claim 5 , wherein the one or more processors use a sliding window based correlation filter to determine the brightness level.
7. The computing device of claim 5 , wherein the one or more processors use a convolutional neural network to determine whether the brightness level satisfies or fails to satisfy the second threshold condition.
8. The computing device of claim 1 , wherein the modification includes increasing the confidence level when one of the one or more illumination patterns indicates an alternating flash rate consistent with a typical railroad light assembly indicating a number of alternating flashes per time unit between each light of the pair of lights.
9. The computing device of claim 1 , wherein the modification includes decreasing the confidence level when one of the one or more illumination patterns does not indicate an alternating flash rate consistent with a typical railroad light assembly indicating a number of alternating flashes per time unit between each light of the pair of lights.
10. The computing device of claim 9 , wherein the modification includes decreasing the confidence level when at least one of the following is detected: i) no illumination of either of the pair of lights of the railroad light assembly; ii) illumination of one or both of the pair of lights has not been detected for a period of time inconsistent with a typical railroad light assembly; and iii) simultaneous illumination of both of the pair of lights of the railroad light assembly.
11. The computing device of claim 1 , wherein the one or more processors are further configured to:
control the vehicle to autonomously take a third action, different from the first action and the second action, when the modified confidence level is within a predefined value of a predetermined confidence threshold value based on the first threshold condition,
wherein the predetermined confidence threshold value is a value of the modified confidence level indicating that the railroad light assembly is active.
12. The computing device of claim 8 , wherein the alternating flash rate consistent with a typical railroad light assembly is between 45 and 65 times per minute.
13. A system comprising:
one or more sensors configured to capture images of a pair of lights of a railroad light assembly; and
a computing device including one or more processors, wherein the one or more processors are configured to:
receive the images of the pair of lights of the railroad light assembly from the one or more sensors;
analyze the images of the pair of lights of the railroad light assembly in order to modify, based on one or more illumination patterns of the pair of lights of the railroad light assembly, a confidence level that indicates a likelihood that the railroad light assembly is active;
control a vehicle to autonomously take a first action when the modified confidence level satisfies a first threshold condition; and
control the vehicle to autonomously take a second action, different from the first action, when the modified confidence level does not satisfy the first threshold condition.
14. The system of claim 13 , wherein the one or more processors are further configured to determine, based on one or more road features, whether the vehicle has reached a location that is a predefined threshold distance away from the railroad light assembly.
15. The system of claim 14 , wherein the one or more processors are further configured to analyze the images of the pair of lights of the railroad light assembly when the vehicle is determined to have reached the location that is the predefined threshold distance away from the railroad light assembly.
16. The system of claim 13 , wherein the one or more processors are further configured to:
update, based on the modified confidence level, a trajectory of the vehicle; and
control the vehicle as the vehicle approaches the railroad light assembly based on the updated trajectory.
17. The system of claim 13 , wherein the one or more processors are further configured to:
categorize image data of the images of the pair of lights of the railroad light assembly by color and shape;
determine a brightness level of portions of the image data categorized as red and circular; and
determine whether the brightness level satisfies or fails to satisfy a second threshold condition indicating that at least one light of the pair of lights is illuminated or not illuminated, respectively.
18. The system of claim 17 , wherein the one or more processors use a sliding window based correlation filter to determine the brightness level; and
wherein the one or more processors use a convolutional neural network to determine whether the brightness level satisfies or fails to satisfy the second threshold condition.
19. The system of claim 13 , wherein the modification includes increasing the confidence level when one of the one or more illumination patterns indicates an alternating flash rate consistent with a typical railroad light assembly indicating a number of alternating flashes per time unit between each light of the pair of lights.
20. The system of claim 19 , wherein the alternating flash rate consistent with a typical railroad light assembly is between 45 and 65 times per minute.