Information processing apparatus, information processing method, mobile-object control apparatus, and mobile object
The present technology relates to an information processing apparatus, an information processing method, a program, a mobile-object control apparatus, and a mobile object that make it possible to appropriately set the accuracy in detecting an object. An information processing apparatus includes a first object detector that performs an object detection on the basis of first sensor data from a first sensor; a second object detector that performs an object detection on the basis of second sensor data from a second sensor that differs in type from the first sensor; a tracking section that predicts a state of a target object that is a tracking target, on the basis of a result of the object detection performed on the basis of the first sensor data, and on the basis of a result of the object detection performed on the basis of the second sensor data; and a detection accuracy controller that sets a high-resolution range on the basis of the state of the target object that is predicted on the basis of the result of the object detection performed on the basis of the first sensor data, and on the basis of the result of the object detection performed on the basis of the second sensor data, the high-resolution range being a range in which an object detection is performed with a higher degree of accuracy than in a range other than the high-resolution range. The present technology is applicable to, for example, a system used to track a target object around a vehicle.
1. An information processing apparatus, comprising:
a first object detector configured to perform a first object detection based on first sensor data from a first sensor, wherein
the first sensor is an image sensor, and
the image sensor is in a camera;
a second object detector configured to perform a second object detection based on second sensor data from a second sensor, wherein
the second sensor differs in type from the first sensor;
a tracking section configured to predict a state of a target object that is a tracking target, based on a result of the first object detection and a result of the second object detection; and
a detection accuracy controller configured to set, based on the predicted state of the target object, a high-resolution range to control a degree of accuracy of each of the first object detection associated with the first sensor and the second object detection associated with the second sensor, wherein
the degree of accuracy is higher in the high-resolution range than in a range different from the high-resolution range,
the high-resolution range is a range in which resolution of an output image of the camera is higher than in a range different from the high-resolution range,
the image sensor obtains a first image,
the camera outputs a second image and a third image,
the second image is based on reduction in resolution of the first image,
the third image is based on cutting out, in the set high-resolution range, an image from the first image, and
the first object detector is further configured to perform, based on the second image and the third image, the first object detection.
2. The information processing apparatus according to claim 1 , wherein
the second sensor is a millimeter-wave radar,
the information processing apparatus further comprises a signal processor configured to perform signal processing on output data of the millimeter-wave radar, and
in the signal processing, the degree of accuracy is higher in the high-resolution range than in a range different from the high-resolution range.
3. The information processing apparatus according to claim 2 , wherein
the signal processor is further configured to estimate, based on the output data of the millimeter-wave radar, each of a distance to an object, a speed of the object, and a direction of the object, and
in the estimation of the direction of the object, the degree of accuracy is higher in the high-resolution range than in a range different from the high-resolution range.
4. The information processing apparatus according to claim 3 , wherein
the signal processor is further configured to set an angular resolution that is higher in the high-resolution range than in a range different from the high-resolution range, to estimate the direction of the object.
5. The information processing apparatus according to claim 3 , wherein
the high-resolution range is a distance-and-speed range in which the degree of accuracy in the estimation of the direction of the object is higher than in a range different from the high-resolution range.
6. The information processing apparatus according to claim 1 , wherein
the tracking section is further configured to predict a location of the target object based on the result of the first object detection and the result of the second object detection, and
the detection accuracy controller is further configured to set the high-resolution range based on the predicted location of the target object.
7. The information processing apparatus according to claim 6 , wherein
the detection accuracy controller is further configured to set the high-resolution range based on the predicted location of the target object, and
the predicted location of the target object is at a distance not less than a specific distance.
8. The information processing apparatus according to claim 1 , wherein
the second sensor is a flash LiDAR,
the information processing apparatus further comprises a signal processor that is configured to generate group-of-points data based on output data of the flash LiDAR, and
the degree of accuracy of group-of-points data is higher in the high-resolution range than in a range different from the high-resolution range.
9. The information processing apparatus according to claim 8 , wherein
the tracking section is further configured to predict a location of the target object based on the result of the first object detection and the result of the second object detection, and
the detection accuracy controller is further configured to set the high-resolution range based on the predicted location of the target object.
10. The information processing apparatus according to claim 1 , wherein
the second sensor is one of a millimeter-wave radar or a flash LiDAR.
11. The information processing apparatus according to claim 1 , wherein
the tracking section is further configured to estimate and predict the state of the target object based on the result of the first object detection and the result of the second object detection.
12. The information processing apparatus according to claim 11 , wherein the tracking section uses a Kalman filter or a particle filter.
13. The information processing apparatus according to claim 1 , wherein
each of the first object detector and the second object detector is further configured to perform the first object detection and the second object detection with respect to surroundings of a mobile object, and
the tracking section is further configured to predict the state of the target object in the surroundings of the mobile object.
14. An information processing method, comprising:
in an information processing apparatus:
performing a first object detection based on first sensor data from a first sensor, wherein
the first sensor is an image sensor, and
the image sensor is in a camera;
performing a second object detection based on second sensor data from a second sensor, wherein the second sensor differs in type from the first sensor;
predicting a state of a target object that is a tracking target, based on a result of the first object detection and a result of the second object detection;
setting, based on the predicted state of the target object, a high-resolution range to control a degree of accuracy of each of the first object detection associated with the first sensor and the second object detection associated with the second sensor, wherein
degree of accuracy is higher in the high-resolution range than in a range different from the high-resolution range,
the high-resolution range is a range in which resolution of an output image of the camera is higher than in a range different from the high-resolution range,
the image sensor obtains a first image,
the camera outputs a second image and a third image,
the second image is based on reduction in resolution of the first image, and
the third image is based on cutting out, in the set high-resolution range, an image from the first image; and
performing the first object detection based on the second image and the third image.
15. A non-transitory computer-readable medium having stored thereon, computer-executable instructions which, when executed by a computer, cause the computer to execute operations, the operations comprising:
performing a first object detection based on first sensor data from a first sensor, wherein
the first sensor is an image sensor, and
the image sensor is in a camera;
performing a second object detection based on second sensor data from a second sensor, wherein the second sensor differs in type from the first sensor;
predicting a state of a target object that is a tracking target, based on a result of the first object detection and a result of the second object detection;
setting, based on the predicted state of the target object, a high-resolution range to control a degree of accuracy of each of the first object detection associated with the first sensor and the second object detection associated with the second sensor, wherein
degree of accuracy is higher in the high-resolution range than in a range different from the high-resolution range,
the high-resolution range is a range in which resolution of an output image of the camera is higher than in a range different from the high-resolution range,
the image sensor obtains a first image,
the camera outputs a second image and a third image,
the second image is based on reduction in resolution of the first image, and
the third image is based on cutting out, in the set high-resolution range, an image from the first image; and
performing the first object detection based on the second image and the third image.
16. A mobile-object control apparatus, comprising:
a first object detector configured to perform, based on first sensor data from a first sensor, a first object detection with respect to surroundings of a mobile object, wherein
the first sensor is an image sensor, and
the image sensor is in a camera;
a second object detector configured to perform, based on second sensor data from a second sensor, a second object detection with respect to the surroundings of the mobile object, wherein the second sensor differs in type from the first sensor;
a tracking section configured to estimate and predict a state of a target object that is a tracking target, based on a result of the first object detection and a result of the second object detection;
a detection accuracy controller configured to set, based on the predicted state of the target object, a high-resolution range to control a degree of accuracy of each of the first object detection associated with the first sensor and the second object detection associated with the second sensor, wherein
the degree of accuracy is higher in the high-resolution range than in a range different from the high-resolution range,
the high-resolution range is a range in which resolution of an output image of the camera is higher than in a range different from the high-resolution range,
the image sensor obtains a first image,
the camera outputs a second image and a third image,
the second image is based on reduction in resolution of the first image,
the third image is based on cutting out, in the set high-resolution range, an image from the first image, and
the first object detector is further configured to perform, based on the second image and the third image, the first object detection; and
a movement controller configured to control movement of the mobile object based on the estimated state of the target object.
17. A mobile object, comprising:
a first sensor that is an image sensor, wherein the image sensor is in a camera;
a second sensor that differs in type from the first sensor;
a first object detector configured to perform a first object detection based on first sensor data from the first sensor;
a second object detector configured to perform a second object detection based on second sensor data from the second sensor;
a tracking section configured to estimate and predict a state of a target object that is a tracking target, based on a result of the first object detection and a result of the second object detection;
a detection accuracy controller configured to set, based on the predicted state of the target object, a high-resolution range to control a degree of accuracy of each of the first object detection associated with the first sensor and the second object detection associated with the second sensor, wherein
the degree of accuracy is higher in the high-resolution range than in a range different from the high-resolution range,
the high-resolution range is a range in which resolution of an output image of the camera is higher than in a range different from the high-resolution range,
the image sensor is configured to obtain a first image,
the camera is configured to output a second image and a third image,
the second image is based on reduction in resolution of the first image,
the third image is based on cutting out, in the set high-resolution range, an image from the first image, and
the first object detector is further configured to perform, based on the second image and the third image, the first object detection; and
a movement controller configured to control movement of the mobile object based on the estimated state of the target object.