Image processing device and image processing method
AssigneeOMRON Corporation
InventorTatsuki Sawada
Life of this patent
Estimating the term…
Abstract
An image processing apparatus includes a tracking target setter that sets a tracking target in a first frame of a video, a first feature tracker that tracks the tracking target in a second frame based on a first feature of the tracking target set in the first frame, a second feature tracker that tracks the tracking target in the second frame based on a second feature of the tracking target set in the first frame, a tracking manager that mixes a tracking result obtained by the first feature tracker with a tracking result obtained by the second feature tracker at a predetermined mixing ratio, and an output unit that outputs a detection position of the tracking target in the second frame based on a mixing result obtained by the tracking manager.
Claims11
An image processing apparatus, comprising:
- a memory configured to store instructions;
- and
- a processor configured to execute the instructions stored in the memory to thereby implement:
- a tracking target setter configured to set a tracking target in a first frame of a video;
- a first feature tracker configured to track the tracking target in a second frame based on a first feature of the tracking target set in the first frame;
- a second feature tracker configured to track the tracking target in the second frame based on a second feature of the tracking target set in the first frame;
- a tracking manager configured to mix a tracking result obtained by the first feature tracker with a tracking result obtained by the second feature tracker at a predetermined mixing ratio;
- and
- an output unit configured to output a detection position of the tracking target in the second frame based on a mixing result obtained by the tracking manager, wherein
- the first feature is a color feature, and the second feature is a shape feature,
- the first feature tracker obtains a color distance between a color of the tracking target in the first frame and a color of the tracking target in the second frame, and generates a likelihood map indicating a likelihood of a position of the tracking target in the second frame based on the color distance, and
- the second feature tracker obtains a difference in shape between an image in the second frame and an image obtained by shifting the position of the tracking target in the first frame, and generates a likelihood map indicating a likelihood of the position of the tracking target in the second frame based on the difference.
An image processing method implementable with a computer, the method comprising:
- setting a tracking target in a first frame of a video;
- tracking the tracking target in a second frame based on a first feature of the tracking target set in the first frame;
- tracking the tracking target in the second frame based on a second feature of the tracking target set in the first frame;
- mixing a tracking result obtained based on the first feature with a tracking result obtained based on the second feature at a predetermined mixing ratio;
- and
- outputting a detection position of the tracking target in the second frame based on a mixing result obtained by mixing the tracking result obtained based on the first feature with the tracking result obtained based on the second feature at the predetermined mixing ratio, wherein
- the first feature is a color feature, and the second feature is a shape feature,
- in the tracking the tracking target in the second frame based on the first feature, a color distance between a color of the tracking target in the first frame and a color of the tracking target in the second frame is obtained, and a likelihood map indicating a likelihood of a position of the tracking target in the second frame is generated based on the obtained color distance, and
- in the tracking the tracking target in the second frame based on the second feature, a difference in shape between an image in the second frame and an image obtained by shifting the position of the tracking target in the first frame is obtained, and a likelihood map indicating a likelihood of the position of the tracking target in the second frame is generated based on the obtained difference.
Dependencies are inferred from the claim text.