Method for image segmentation

Patent metadata

Field	Value
Publication number	US-9807316-B2
Application number	US-201514838378-A
Country	US
Kind code	B2
Filing date	Aug 28, 2015
Priority date	Sep 4, 2014
Publication date	Oct 31, 2017
Grant date	Oct 31, 2017

Abstract

An image segment method is provided in this disclosure. The method is suitable for an electronic apparatus including a first camera and a motion sensor. The method includes steps of: providing at least one pre-defined model mask; fetching pose data from the motion sensor, the pose data being related to an orientation or a position of the first camera; adjusting one of the at least one pre-defined model mask into an adaptive model mask according to the pose data; and, extracting an object from an image captured by the first camera according to the adaptive model mask.

First claim

What is claimed is:

1. A method, suitable for an electronic apparatus comprising a first camera and a motion sensor, the method comprising:
providing at least one pre-defined model mask;
fetching pose data from the motion sensor, the pose data being related to an orientation or a position of the first camera;
adjusting one of the at least one pre-defined model mask into an adaptive model mask according to the pose data; and
extracting an object from an image captured by the first camera according to the adaptive model mask;
wherein the adjusting step comprises at least one of the following:
when the pose data indicates that the first camera is moved downward, the pre-defined model mask is adjusted such that a bypassing area of the adaptive model mask is larger in a lower region and narrower in an upper region;
when the pose data indicates that the first camera is moved upward, the pre-defined model mask is adjusted such that the bypassing area of the adaptive model mask is larger in the upper region and narrower in the lower region; and
when the pose data indicates that the first camera is moved to the right side, the pre-defined model mask is adjusted such that the bypassing area of the adaptive model mask is larger on the right side and narrower on the left side.

2. The method of claim 1 , wherein the electronic apparatus further comprises a second camera, the method comprises:
capturing another image by the second camera simultaneous to the image captured by the first camera; and
blending the object from the image captured by the first camera into the image captured by the second camera.

3. The method of claim 2 , wherein the first camera and the second camera are disposed on opposite surfaces of the electronic apparatus, the image captured by the first camera is a selfie image.

4. The method of claim 1 , wherein the pre-defined model mask is shaped corresponding to a front view of a human, the front view at least covers a head and an upper body of the human, and a right half and a left half of the pre-defined model mask are substantially mirrored to each other.

5. The method of claim 1 , wherein plural pre-defined model masks are provided, the pre-defined model masks are shaped in different types of humans, the pre-defined model masks are differentiated by a hair style, a weight, a size or a skeleton topology.

6. The method of claim 1 , wherein the pre-defined model mask is shaped corresponding to a front view of multiple people, the front view at least covers heads and upper bodies of these people.

7. The method of claim 1 , wherein the pose data fetched from the motion sensor indicates a rotation angle of the orientation of the first camera or a shifting distance of the position of the first camera, the adaptive model mask is calculated from the pre-defined model mask in correspondence with the rotation angle and the shifting distance.

8. A method, suitable for an electronic apparatus comprising a first camera and a motion sensor, the method comprising:
capturing a first image by the first camera;
estimating an object pose from the first image;
fetching pose data from the motion sensor, the pose data being related to a position or an orientation of the first camera when the first image is captured;
analyzing a camera pose according to the pose data;
tracking a relative motion according to the camera pose and the object pose;
providing a pre-defined model mask;
adjusting the pre-defined model mask into an adaptive model mask according to the relative motion; and
adaptively extracting the object from the first image according to the adaptive model mask;
wherein the adjusting step comprises at ;east one of the following:
when the relative motion indicates that the first camera is moved downward, the pre-defined model mask is adjusted such that a bypassing area of the adaptive model mask is larger in a lower region and narrower in an upper region;
when the relative motion indicates that the first camera is moved upward, the pre-defined model mask is adjusted such that the bypassing area of the adaptive model mask is larger in the upper region and narrower in the lower region; and
when the relative motion indicates that the first camera is moved to the right side, the pre-defined model mask is adjusted such that the bypassing area of the adaptive model mask is larger on the right side and narrower on the left side.

9. The method of claim 8 , further comprising:
providing a plurality of pre-defined model masks;
selecting one of the pre-defined model masks according to the first image; and
adjusting the selected pre-defined model mask into the adaptive model mask according to the relative motion.

10. The method of claim 9 , wherein the pre-defined model masks are shaped in different types of humans, the pre-defined model masks are differentiated by a hair style, a weight, a size or a skeleton topology, in the step of selecting one of the pre-defined model masks according to the first image, the method further comprises:
comparing an object within the first image with at least one of the pre-defined model masks.

11. The method of claim 8 , wherein the electronic apparatus further comprises a second camera, the method comprises:
capturing a second image by the second camera simultaneous to the first image captured by the first camera; and
blending the object from the first image captured by the first camera into the second image captured by the second camera.

12. The method of claim 11 , wherein the first camera and the second camera are disposed on opposite surfaces of the electronic apparatus, the first image captured by the first camera is a selfie image.

13. The method of claim 8 , wherein the pre-defined model masks is shaped corresponding to a front view of multiple people, the front view at least covers heads and upper bodies of these people.

14. The method of claim 8 , wherein the relative motion indicates a rotation angle or a shifting distance between the camera pose and the object pose, the adaptive model mask is calculated from the pre-defined model mask in correspondence with the relative motion.

15. A method, suitable for an electronic apparatus comprising a first camera and a motion sensor, the method comprising:
providing a first model mask;
capturing a first image by the first camera, the first image comprising an object;
extracting the object from the first image according to the first model mask;
capturing a second image by the first camera;
fetching pose data from the motion sensor, the pose data being related to a position or an orientation of the first camera when the second image is captured;
tracking a relative motion from the first image to the second image according to the pose data and the object from the first image;
dynamically providing a second model mask according to the relative motion; and
adaptively extracting another object from the second image according to the second model mask;
wherein the dynamically providing step comprises at least one of the following:
when the relative motion indicates that the first camera is moved downward, the second model mask is adjusted such that a bypassing area of the adaptive model mask is larger in a lower region and narrower in an upper region;
when the relative motion indicates that the first camera is moved upward, the second model mask is adjusted such that the bypassing area of the adaptive model mask is larger in the upper region and narrower in the lower region;
and when the relative motion indicates that the first camera is moved to the right side, the second model mask is adjusted such tha

Assignees

• Htc Corp

Inventors

• Jen Cheng-Lung
• Tai Pol-Lin
• Wu Jing-Lung
• Chueh Hsin-Ti

Classifications

• H04N23/635

  Region indicators; Field of view indicators · CPC title

• H04N23/611

  where the recognised objects include parts of the human body · CPC title

• H04N23/61

  based on recognised objects · CPC title

• G06F18/22

  Matching criteria, e.g. proximity measures · CPC title

• H04N5/2621

  Cameras specially adapted for the electronic generation of special effects during image pickup, e.g. digital cameras, camcorders, video cameras having integrated special effects capability · CPC title