# US 2012/0249812 A1 · Method and apparatus for camera motion analysis in video

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2012/0249812 A1
- Document ID: US20120249812A1
- Status: Published application
- Kind: A1
- Application number: 13/355,990
- Filed: 2012-01-23
- Published: 2012-10-04
- Earliest claimed: 2011-03-31
- Assignees: Fujitsu Limited
- Inventors: Fei LI; Rujie Liu; Hao Yu; Takayuki Baba; Yusuke Uehara
- Web page: https://patentgrep.com/patent/US20120249812A1

## CPC

- G06T 7/20: Analysis of motion (motion estimation for coding, decoding, compressing or decompressing digital video signals H04N19/43, H04N19/51)
- G06V 20/48: {Matching video sequences}
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

Embodiments disclose methods and apparatuses for camera motion analysis in a video, wherein one of the methods includes: analyzing, video segments with significant movement characteristics and video segments without any significant movement characteristic; for each segment without any significant movement characteristic, if a first motion type corresponding to a preceding neighboring video segment is different from a second motion type corresponding to a succeeding neighboring video segment, lowering detection criterions of the first type and the second type; judging whether the segment without any significant movement characteristic meets the lowered detection criterions; and merging the segment without any significant movement characteristic with a neighboring video segment according to the result of judgment. With the embodiments of the invention the motion type of a camera in the video can be detected more effectively and accurately, and the photographic intention of a user can be reflected more accurately.

## Claims

### Claim 1

A method for camera motion analysis in a video, comprising:
analyzing, from the video, video segments with significant movement characteristics and video segments without significant movement characteristics;
as for each video segment without significant movement characteristics, if a first motion type corresponding to a preceding neighboring video segment is different from a second motion type corresponding to a succeeding neighboring video segment, lowering detection criterions of the first motion type and the second motion type;
judging whether the video segment without significant movement characteristics meet the lowered detection criterions; and
merging the video segment without significant movement characteristics with neighboring video segments according to the judging result.

### Claim 2 (depends on 1)

The method according to claim 1, wherein merging the video segment without significant movement characteristics with neighboring video segments according to the judging result comprises:
if the video segment without significant movement characteristics meets only one of the lowered detection criterions of the first motion type and the second motion type, merging the video segment without significant movement characteristic with a neighboring video segment corresponding to the met one.

### Claim 3 (depends on 1)

The method according to claim 1, wherein merging the video segment without significant movement characteristics with neighboring video segments according to the judging result comprises:
if the video segment without significant movement characteristics meets the lowered detection criterions of both the first motion type and the second motion type, merging the video segment without significant movement characteristic with the neighboring video segments according to extent of meeting.

### Claim 4 (depends on 1)

The method according to claim 1, wherein merging the video segment without significant movement characteristics with neighboring video segments according to the judging result comprises:
if the video segment without significant movement characteristics meets neither the lowered detection criterion of the first motion type nor that of the second motion type, dividing the video segment without significant movement characteristics into two parts; and
merging the first half part with the preceding neighboring video segment and the second half part with the succeeding neighboring video segment.

### Claim 5

A method for camera motion analysis in a video, comprising:
analyzing, from the video, video segments contained in the video and camera motion types thereof;
finding video segments meeting a preset merging rule in the video segments of the video; and
merging the found video segments and determining the motion type of the merged video segments according to the merging rule.

### Claim 6 (depends on 5)

The method according to claim 5, wherein finding video segments meeting a preset merging rule in the video segments of the video comprises:
finding the video segments meeting the preset merging rule according to neighboring relationship, the motion types and durations of the video segments of the video.

### Claim 7 (depends on 6)

The method according to claim 6, wherein finding the video segments meeting the preset merging rule according to neighboring relationship, the motion types and durations of the video segments of the video comprises:
finding the video segments to be merged according to the following condition: as for a first video segment, a second video segment and a third video segment which are neighboring chronologically, if the motion types of both the first video segment and the third video segment are zooming, the motion type of the second video segment is stillness or translation and the duration thereof is less than a preset threshold, then the first video segment, the second video segment and the third video segment meet the preset merging rule; and
wherein merging the found video segments and determining the motion type of the merged video segments according to the merging rule comprises:
merging the first video segment, the second video segment and the third video segment and determining the motion type of the merged video segments as zooming.

### Claim 8 (depends on 6)

The method according to claim 6, wherein finding the video segments meeting the preset merging rule according to neighboring relationship, the motion types and durations of the video segments of the video comprises:
finding the video segments to be merged according to the following condition: as for a first video segment, a second video segment and a third video segment which are neighboring chronologically, if the motion type of the first video segment is zooming, the motion type of the second video segment is translation and the duration thereof is less than a first preset threshold and the motion type of the third video segment is stillness and the duration thereof is greater than a second preset threshold, then the second video segment and the third video segment meet the preset merging rule; and
wherein merging the found video segments and determining the motion type of the merged video segments according to the merging rule comprises:
merging the second video segment with the third video segment and determining the motion type of the merged video segments as stillness.

### Claim 9 (depends on 6)

The method according to claim 6, wherein finding the video segments meeting the preset merging rule according to neighboring relationship, the motion types and durations of the video segments of the video comprises:
finding the video segments to be merged according to the following condition: as for a first video segment, a second video segment and a third video segment which are neighboring chronologically, if the motion type of the first video segment is zooming, the motion type of the second video segment is stillness and the duration thereof is less than a first preset threshold and the motion type of the third video segment is translation and the duration thereof is greater than a second preset threshold, then the second video segment and the third video segment meet the preset merging rule; and
wherein merging the found video segments and determining the motion type of the merged video segments according to the merging rule comprises:
merging the second video segment with the third video segment and determining the motion type of the merged video segments as translation.

### Claim 10 (depends on 6)

The method according to claim 6, wherein finding the video segments meeting the preset merging rule according to neighboring relationship, the motion types and durations of the video segments of the video comprises:
finding the video segments to be merged according to following conditions: as for a first video segment, a second video segment and a third video segment which are neighboring chronologically, if the motion type of the first video segment is zooming, the motion type of the second video segment is translation in a first direction and the duration thereof is less than a first preset threshold and the motion type of the third video segment is translation in a second direction and the duration thereof is greater than a second preset threshold, wherein the first direction is different from the second direction, then the second video segment and the third video segment meet the preset merging rule; and
wherein merging the found video segments and determining the motion type of the merged video segments according to the merging rule comprises:
merging the second video segment with the third video segment and determining the motion type of the merged video segments as translation in the second direction.

### Claim 11

An apparatus for camera motion analysis in a video, comprising:
an analyzing unit, configured to analyze, from the video, video segments with significant movement characteristics and video segments without significant movement characteristics;
a criterion adjusting unit, configured to, as for each video segment without significant movement characteristics, if a first motion type corresponding to a preceding neighboring video segment is different from a second motion type corresponding to a succeeding neighboring video segment, lower detection criterions of the first motion type and the second motion type;
a judging unit, configured to judge whether the video segment without significant movement characteristics meets the lowered detection criterions; and
a merging unit, configured to merge the video segment without significant movement characteristics with neighboring video segments according to the judging result.

### Claim 12 (depends on 11)

The apparatus according to claim 11, wherein the merging unit comprises:
a first merging subunit, configured to, if the video segment without significant movement characteristics meets the lowered detection criterion of only one of the motion types, merge the video segment without significant movement characteristic with a neighboring video segment corresponding to the met one.

### Claim 13 (depends on 11)

The apparatus according to claim 11, wherein the merging unit comprises:
a second merging subunit, configured to, if the video segment without significant movement characteristics meets the lowered detection criterions of both the first motion type and the second motion type, merge the video segment without significant movement characteristic with neighboring video segments according to extent of meeting.

### Claim 14 (depends on 11)

The apparatus according to claim 11, wherein the merging unit comprises:
a dividing subunit, configured to, if the video segment without significant movement characteristics meets neither the lowered detection criterion of the first motion type nor that of the second motion type, divide the video segment without significant movement characteristics into two parts; and
a third merging subunit, configured to, merge the first half part with the preceding neighboring video segment and the second half part with the succeeding neighboring video segment.
