Title:
IMAGE PROCESSING DEVICE, IMAGE PROCESSING METHOD, AND PROGRAM
Document Type and Number:
WIPO Patent Application WO/2021/048985
Kind Code:
A1
Abstract:
Provided are an image processing device, an image processing method, and a program with which an accurate three-dimensional map can be formed even when a moving object is included in an input image. This image processing device includes: an image acquiring unit that sequentially acquires two-dimensional photographic images; an object type recognition executing unit that applies, to each pixel in the sequentially acquired photographic images, a label indicating the type of an object represented by said pixel; and a three-dimensional map creating unit that executes three-dimensional position recognition for each pixel in the photographic images to create a three-dimensional map on the basis of the sequentially acquired photographic images. The three-dimensional map creating unit limits the three-dimensional position recognition for each pixel in the photographic images in accordance with the label applied to said pixel.
Inventors:
HORIKAWA TSUTOMU (JP)
ONO DAICHI (JP)
YABE HIROYUKI (JP)
ONO DAICHI (JP)
YABE HIROYUKI (JP)
Application Number:
PCT/JP2019/035972
Publication Date:
March 18, 2021
Filing Date:
September 12, 2019
Export Citation:
Assignee:
SONY INTERACTIVE ENTERTAINMENT INC (JP)
International Classes:
G06T7/70
Domestic Patent References:
| WO2015111283A1 | 2015-07-30 |
Foreign References:
| JP2019028861A | 2019-02-21 | |||
| JPH0618221A | 1994-01-25 |
Other References:
TATENO, KEISUKE ET AL.: "CNN-SLAM: Real-time dense monocular SLAM with learned depth prediction", 2017 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, 2017, pages 6565 - 6574, XP033250021, DOI: 10.1109/CVPR.2017.695
MARTIN RUNZ ET AL.: "Co-Fusion: Real-time Segmentation, Tracking and Fusion of Multiple Objects", 2017 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION, 2017, pages 4471 - 4478, XP033127260, DOI: 10.1109/ICRA.2017.7989518
MARTIN RUNZ ET AL.: "Co-Fusion: Real-time Segmentation, Tracking and Fusion of Multiple Objects", 2017 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION, 2017, pages 4471 - 4478, XP033127260, DOI: 10.1109/ICRA.2017.7989518
Attorney, Agent or Firm:
HARUKA PATENT & TRADEMARK ATTORNEYS (JP)
Download PDF: