System and method for achieving fast and reliable time-to-contact estimation using vision and range sensor data for autonomous navigation
Described is a robotic system for detecting obstacles reliably with their ranges by a combination of two-dimensional and three-dimensional sensing. In operation, the system receives an image from a monocular video and range depth data from a range sensor of a scene proximate a mobile platform. The image is segmented into multiple object regions of interest and time-to-contact (TTC) value are calculated by estimating motion field and operating on image intensities. A two-dimensional (2D) TTC map is then generated by estimating average TTC values over the multiple object regions of interest. A three-dimensional TTC map is then generated by fusing the range depth data with image. Finally, a range-fused TTC map is generated by averaging the 2D TTC map and the 3D TTC map.
1. A system for estimating a time-to-contact an object using vision and range sensor data for autonomous navigation, comprising:
one or more processors and a memory, the memory having executable instructions encoded thereon, such that upon execution of the instructions, the one or more processors perform operations of:
segmenting an image from a monocular video into multiple object regions of interest, the image being of a scene proximate a mobile platform;
calculating time-to-contact (TTC) values by estimating motion field and operating on image intensities;
generating a two-dimensional (2D) TTC map by estimating average TTC values over the multiple object regions of interest;
fusing range depth data from a range sensor with the image to generate a three-dimensional (3D) TTC map; and
generating a range-fused TTC map by averaging the 2D TTC map and the 3D TTC map.
2. The system as set forth in claim 1 , further comprising operations of:
detecting an object in the range-fused TTC map; and
generating a command to cause a mobile platform to move to avoid contact with the object.
3. The system as set forth in claim 1 , wherein in generating the 3D TTC map, a range data reading is associated with each pixel in the image.
4. The system as set forth in claim 1 , wherein in generating the range-fused TTC map, the range-fused TTC map is generated with range-seeded propagation of disparity and integrated with salient features in foreground objects to generate precise obstacle boundaries and TTC for objects in the scene proximate the mobile platform.
5. The system as set forth in claim 1 , wherein in segmenting an image from the monocular video into multiple object regions of interest, a foreground detector is used to segment objects within a foreground of the image.
6. The system as set forth in claim 1 , wherein a spiking neural network is used to calculate TTC values by estimating motion field and operating on image intensities.
7. The system as set forth in claim 1 , wherein in generating the 3D TTC map, a range data reading is associated with each pixel in the image;
wherein in generating the range-fused TTC map, the range-fused TTC map is generated with range-seeded propagation of disparity and integrated with salient features in foreground objects to generate precise obstacle boundaries and TTC for objects in the scene proximate the mobile platform;
wherein in segmenting an image from the monocular video into multiple object regions of interest, a foreground detector is used to segment objects within a foreground of the image; and
wherein a spiking neural network is used to calculate TTC values by estimating motion field and operating on image intensities.
8. A computer program product for estimating a time-to-contact an object using vision and range sensor data for autonomous navigation, the computer program product comprising:
a non-transitory computer-readable medium having executable instructions encoded thereon, such that upon execution of the instructions by one or more processors, the one or more processors perform operations of:
segmenting an image from a monocular video into multiple object regions of interest, the image being of a scene proximate a mobile platform;
calculating time-to-contact (TTC) values by estimating motion field and operating on image intensities;
generating a two-dimensional (2D) TTC map by estimating average TTC values over the multiple object regions of interest;
fusing range depth data from a range sensor with the image to generate a three-dimensional (3D) TTC map; and
generating a range-fused TTC map by averaging the 2D TTC map and the 3D TTC map.
9. The computer program product as set forth in claim 8 , further comprising instructions for causing the one or more processors to perform operations of:
detecting an object in the range-fused TTC map; and
generating a command to cause a mobile platform to move to avoid contact with the object.
10. The computer program product as set forth in claim 8 , wherein in generating the 3D TTC map, a range data reading is associated with each pixel in the image.
11. The computer program product as set forth in claim 8 , wherein in generating the range-fused TTC map, the range-fused TTC map is generated with range-seeded propagation of disparity and integrated with salient features in foreground objects to generate precise obstacle boundaries and TTC for objects in the scene proximate the mobile platform.
12. The computer program product as set forth in claim 8 , wherein in segmenting an image from the monocular video into multiple object regions of interest, a foreground detector is used to segment objects within a foreground of the image.
13. The computer program product as set forth in claim 8 , wherein a spiking neural network is used to calculate TTC values by estimating motion field and operating on image intensities.
14. The computer program product as set forth in claim 8 , wherein in generating the 3D TTC map, a range data reading is associated with each pixel in the image;
wherein in generating the range-fused TTC map, the range-fused TTC map is generated with range-seeded propagation of disparity and integrated with salient features in foreground objects to generate precise obstacle boundaries and TTC for objects in the scene proximate the mobile platform;
wherein in segmenting an image from the monocular video into multiple object regions of interest, a foreground detector is used to segment objects within a foreground of the image; and
wherein a spiking neural network is used to calculate TTC values by estimating motion field and operating on image intensities.
15. A computer implemented method for estimating a time-to-contact an object using vision and range sensor data for autonomous navigation, the method comprising an act of:
causing one or more processors to execute instructions encoded on a non-transitory computer-readable medium, the one or more processors perform operations of:
segmenting an image from a monocular video into multiple object regions of interest, the image being of a scene proximate a mobile platform;
calculating time-to-contact (TTC) values by estimating motion field and operating on image intensities;
generating a two-dimensional (2D) TTC map by estimating average TTC values over the multiple object regions of interest;
fusing range depth data from a range sensor with the image to generate a three-dimensional (3D) TTC map; and
generating a range-fused TTC map by averaging the 2D TTC map and the 3D TTC map.
16. The method as set forth in claim 15 , further comprising operations of:
detecting an object in the range-fused TTC map; and
generating a command to cause a mobile platform to move to avoid contact with the object.
17. The method as set forth in claim 15 , wherein in generating the 3D TTC map, a range data reading is associated with each pixel in the image.
18. The method as set forth in claim 15 , wherein in generating the range-fused TTC map, the range-fused TTC map is generated with range-seeded propagation of disparity and integrated with salient features in foreground objects to generate precise obstacle boundaries and TTC for objects in the scene proximate the mobile platform.
19. The method as set forth in claim 15 , wherein in segmenting an image from the monocular video into multiple object regions of interest, a foreground detector is used to segment objects within a foreground of the image.
20. The method as set forth in claim 15 , wherein a spiking neural network is used to calculate TTC values by estimating motion field and operating on image intensities.
21. The method as set forth in claim 15 , wherein in generating the 3D TTC map, a range data reading is associated with each pixel in the image;
wherein in generating the range-fused TTC map, the range-fused TTC map is generated with range-seeded propagation of disparity and integrated with salient features in foreground objects to generate precise obstacle boundaries and TTC for objects in the scene proximate the mobile platform;
wherein in segmenting an image from the monocular video into multiple object regions of interest, a foreground detector is used to segment objects within a foreground of the image; and
wherein a spiking neural network is used to calculate TTC values by estimating motion field and operating on image intensities.