CN106162330B - A method of extracting caption area in video pictures - Google Patents
A method of extracting caption area in video pictures Download PDFInfo
- Publication number
- CN106162330B CN106162330B CN201610597947.4A CN201610597947A CN106162330B CN 106162330 B CN106162330 B CN 106162330B CN 201610597947 A CN201610597947 A CN 201610597947A CN 106162330 B CN106162330 B CN 106162330B
- Authority
- CN
- China
- Prior art keywords
- point
- corroded
- subtitle
- seed
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/43—Processing of content or additional data, e.g. demultiplexing additional data from a digital video stream; Elementary client operations, e.g. monitoring of home network or synchronising decoder's clock; Client middleware
- H04N21/44—Processing of video elementary streams, e.g. splicing a video clip retrieved from local storage with an incoming video stream, rendering scenes according to MPEG-4 scene graphs
- H04N21/44008—Processing of video elementary streams, e.g. splicing a video clip retrieved from local storage with an incoming video stream, rendering scenes according to MPEG-4 scene graphs involving operations for analysing video streams, e.g. detecting features or characteristics in the video stream
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/43—Processing of content or additional data, e.g. demultiplexing additional data from a digital video stream; Elementary client operations, e.g. monitoring of home network or synchronising decoder's clock; Client middleware
- H04N21/435—Processing of additional data, e.g. decrypting of additional data, reconstructing software from modules extracted from the transport stream
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/47—End-user applications
- H04N21/488—Data services, e.g. news ticker
- H04N21/4884—Data services, e.g. news ticker for displaying subtitles
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N5/00—Details of television systems
- H04N5/222—Studio circuitry; Studio devices; Studio equipment
- H04N5/262—Studio circuits, e.g. for mixing, switching-over, change of character of image, other special effects ; Cameras specially adapted for the electronic generation of special effects
- H04N5/278—Subtitling
Abstract
The method that the invention discloses a kind of to extract caption area in video pictures, it is the following steps are included: S1: removal noise treatment: obtaining gradient image data using maximum value gradient method according to the texture difference of subtitle and background frame: wherein, subtitle parts have neat data, and background frame then will appear irregular noise data；Remove irregular noise data part；S2: caption area is positioned by the way of locking corrosion；S3: headline is determined.The present invention is the efficiency in order to improve the processing of news demolition, a kind of method that caption area is extracted from video pictures according to video caption characteristic for quickly determining suitable title to the news split, while taxonomic revision, the fast search of news video being provided for convenience.Wherein, caption area convenience and high reliablity are determined using the method for locking corrosion.
Description
Technical field
The method that the present invention relates to a kind of to extract caption area in video pictures.
Background technique
For TV station, the news material of this TV station has complete storage management system, it is easy to regard to a news
A piece of news in frequency carries out subsequent use；But the news video in remaining source just needs to carry out demolition processing, i.e., by news
Video split into one by one individual news so as to subsequent use.Subtitle in news video, especially main title can letters
The bright main points for embodying this then news in a capsule.If the caption recognition in video pictures can be come out, can quickly tear open
The individual news being divided into one by one determines suitable title, this has visual classification arrangement, fast search etc. very big
Benefit.
In order to improve the efficiency of news demolition processing, suitable title quickly is determined to the news split, while for side
Just to the taxonomic revision of news video, fast search and invent a kind of extract word from video pictures according to video caption characteristic
The method in curtain region.
Summary of the invention
It is an object of the invention to overcome the deficiencies of the prior art and provide a kind of to extract caption area in video pictures
Method.
The purpose of the present invention is achieved through the following technical solutions: a kind of to extract caption area in video pictures
Method, it the following steps are included:
S1: ladder removal noise treatment: is obtained using maximum value gradient method according to the texture difference of subtitle and background frame
Spend image data: where subtitle parts have neat data, and background frame then will appear irregular noise data；It goes
Except irregular noise data part；
S2: positioning caption area, including following sub-step:
S21: the value of whole region under original state, is set as background by cloth seed, is found and is closed according to gradient image data
Suitable place plants seed, to guarantee seed on word；
S22: the value wait judge the point i.e. G point of corrosion, including following sub-step are calculated using maximum value process:
S221: original graph is calculated separately out in vertical, horizontal, three directions of diagonal line gray scale, tetra- channels R, G, B
Maximum value；
S222: calculating the maximum value in four values, and the maximum value is denoted as the value of G point；
S23: corrosion, including following sub-step are locked:
S231: the point of left and right, the upper and lower four direction of seed in drawing of seeds is successively traversed, judges whether the point is corroded
Judge whether the point is G point: if the point has been corroded, continuing to judge next point；If the point is not corroded, continue
In next step；
S232: compare the size of the value of the point and the first threshold nGsensitive of a setting, if the value of the point is greater than
Threshold value cannot then corrode the point, and as the point has added one layer of lock；If the value of the point be less than threshold value, carry out in next step, carry out to
Left corrosion；
S233:G point is the point that corrosion is waited judging in the seed left side, is had on the right of seed 0 to N number of point being corroded, to the left
Corrosion, including following sub-step:
S2331: judge the quantity of the point being corroded on the right of seed:
(1) when the point not being corroded on the right of seed, then G point can be corroded；
(2) when having 1 to N point being corroded on the right of seed, gray scale, R, G, B tetra- of the point of G point and rightmost are calculated
The maximum value in channel, when the maximum value is less than the second threshold nSensitive of setting, then G point can corrode, otherwise cannot be by
Corrosion adds one layer of lock to the point；
S2332: the point being corroded becomes seed, and return step S2331 is until can not corrosion sites again；
S234: whether monitoring subtitle fringe region reaches Corrosion standards, terminates to corrode if having reached, and otherwise increases rotten
The second threshold nSensitive of erosion repeats above-mentioned steps；
S235: caption data is extracted according to Corrosion results, then noise data is removed further according to character features, finally obtains
One accurate subtitle position region；
S3: headline, including following sub-step are determined:
S31: the continuity Characteristics according to subtitle in video properties screen identical caption area continuous in video
And filtering, and choose picture of the effect preferably as identification subtitle；
S32: being converted to text for subtitle picture, as the headline of this then news.
A kind of method that extracting caption area in video pictures further includes threshold value setting sub-step S0, in beginning
It is preceding that first threshold nGsensitive and second threshold nSensitive are configured respectively.
The N value is 4.
The step S1 further includes following sub-step: according to the feature of news caption profile consistency from top to bottom to video pictures
Lateral striping identification is carried out, and noise removal processing is carried out according to the feature of news caption.
The beneficial effects of the present invention are:
The present invention is the efficiency in order to improve the processing of news demolition, quickly determines suitable title to the news split, together
When taxonomic revision, the fast search of news video are provided for convenience it is a kind of according to video caption characteristic from video pictures
The middle method for extracting caption area.Wherein, caption area convenience and high reliablity are determined using the method for locking corrosion.
Detailed description of the invention
Fig. 1 is the method for the present invention flow chart；
Fig. 2 is that original graph is vertical, horizontal, three direction schematic diagrams of diagonal line；
Fig. 3 is to lock corrosion flow chart.
Specific embodiment
Technical solution of the present invention is described in further detail with reference to the accompanying drawing:
Firstly, news video has a characteristic that
1) texture difference of subtitle and background: generally in order to make subtitle understand easy resolution, subtitle and background are had significantly
Cut-off rule, thus subtitle and the texture difference of background be it is very big, the profile of subtitle can be determined with this.
2) feature of news caption:
A) profile consistency from top to bottom: from profile, background frame is usually irregular pattern-information, and news caption is then
It is pattern neat up and down.
B) profile transformation density is big: from profile, the few profiles of background frame are irregular to be converted very much, and news word
The profile transformation density of curtain is very big.
C) can be with cutting: because there is interval between word and word, each word be that shape is similar in terms of profile in subtitle
Independent individual can be with cutting for caption area；And background frame be difficult occur can be by profile cutting forming shape
Similar individual.
D) generally there are frame characteristics: news caption generally can be all placed on special subtitle bottom plate, have intrinsic colour
Continuous gradation is constant, but with up and down or left and right significant change feature, can according to bottom plate feature remove side program mark
Know data.
3) characteristic of video:
A) continuity: for subtitle, continuous a period of time (for news content subtitle may be several milliseconds may also
It is several seconds, but may be up to even more than ten minutes a few minutes for headline subtitle), for video pictures in variation, subtitle is but
It will not change.That is, there is no variations for subtitle in one section of continuous frame, there is continuity.
B) position is constant: subtitle is to edit in the predetermined region of subtitle bottom plate, and the position of subtitle bottom plate is usually
Constant, so the position of caption area is also fixed and invariable, and the background colour of subtitle bottom plate is gradual change or constant.
Therefore, according to These characteristics, as shown in Figure 1, a kind of method for extracting caption area in video pictures, it includes
Following steps:
S1: removal noise treatment:
(1) gradient image data are obtained using maximum value gradient method according to the texture difference of subtitle and background frame: its
In, subtitle parts have neat data, and background frame then will appear irregular noise data；Remove irregular noise
Data portion；
(2) lateral striping is carried out to video pictures according to the feature of news caption profile consistency from top to bottom and identifies (band
Change identification: carrying out the identification one by one of one fixed width to picture), and carried out at noise removal according to the feature of news caption
Reason.
S2: positioning caption area,
A) according to the region that may have subtitle on the feature location frame out of news caption；
B) lemma is all edited on subtitle bottom plate, and background colour constant according to subtitle bottom plate position is gradual change or not
What is become can determine subtitle bottom plate size, and reposition out caption area；
C) caption area is further determined that using the method for locking corrosion: including following sub-step:
S21: the value of whole region under original state, is set as background (255), according to gradient image data by cloth seed
It finds suitable place and plants seed, to guarantee seed on word；Wherein, the value of seed is 0；
S22: the value wait judge the point i.e. G point of corrosion, including following sub-step are calculated using maximum value process:
S221: calculate separately out original graph vertical, horizontal, three directions (as shown in Figure 2) of diagonal line gray scale, R, G,
The maximum value in tetra- channels B:
GGray scale[x]=Max(abs(a1-a2), abs (b1-b2), abs (c1-c2), abs (d1-d2))；
GR[x]=Max(abs(a1-a2), abs (b1-b2), abs (c1-c2), abs (d1-d2))；
GG[x]=Max(abs(a1-a2), abs (b1-b2), abs (c1-c2), abs (d1-d2))；
GB[x]=Max(abs(a1-a2), abs (b1-b2), abs (c1-c2), abs (d1-d2))；
S222: calculating the maximum value in four values, and the maximum value is denoted as the value of G point:
G [x]=Max (G gray scale [x], GR[x], GG[x], GB[x])
S23: as shown in figure 3, locking corrosion, including following sub-step:
S231: the point of left and right, the upper and lower four direction of seed in drawing of seeds is successively traversed, judges whether the point is corroded
Judge whether the point is G point: being corroded if 0 if the value of the point, continues to judge next point；If the value of the point is not
It is 0, then continues in next step；
S232: compare the size of the value of the point and the first threshold nGsensitive of a setting, if the value of the point is greater than
Threshold value cannot then corrode the point, and as the point has added one layer of lock；If the value of the point be less than threshold value, carry out in next step, carry out to
Left corrosion；
S233:G point is the point that corrosion is waited judging in the seed left side, has 0 to 4 points being corroded on the right of seed, to the left
Corrosion, including following sub-step:
S2331: judge the quantity of the point being corroded on the right of seed:
(1) when the point not being corroded on the right of seed, then G point can be corroded；
(2) when there are 1 to 4 points being corroded on the right of seed, gray scale, R, G, B tetra- of the point of G point and rightmost are calculated
The maximum value in channel, when the maximum value is less than the second threshold nSensitive of setting, then G point can corrode, otherwise cannot be by
Corrosion adds one layer of lock to the point；
S2332: the point being corroded becomes seed, and return step S2331 is until can not corrosion sites again；
S234: whether monitoring subtitle fringe region reaches Corrosion standards, terminates to corrode if having reached, and otherwise increases rotten
The second threshold nSensitive of erosion repeats above-mentioned steps；
S235: caption data is extracted according to Corrosion results, then noise data is removed further according to character features, finally obtains
One accurate subtitle position region；
S3: headline, including following sub-step are determined:
S31: the continuity Characteristics according to subtitle in video properties screen identical caption area continuous in video
And filtering, and choose picture of the effect preferably as identification subtitle；
S32: being converted to text for subtitle picture, as the headline of this then news.
Claims (4)
1. it is a kind of in video pictures extract caption area method, it is characterised in that: it the following steps are included:
S1: gradient map removal noise treatment: is obtained using maximum value gradient method according to the texture difference of subtitle and background frame
As data: where subtitle parts have neat data, and background frame then will appear irregular noise data；Remove nothing
The noise data part of rule；
S2: positioning caption area, including following sub-step:
S21: the value of whole region under original state, is set as background by cloth seed, and it is suitable to be found according to gradient image data
Place plants seed, to guarantee seed on word；
S22: the value wait judge the point i.e. G point of corrosion, including following sub-step are calculated using maximum value process:
S221: original graph is calculated separately out in the maximum value of vertical, horizontal, two diagonal line four directions gray scale, is calculated
Original graph in the maximum value in vertical, horizontal, two diagonal line four directions the channel R, calculate original graph it is vertical, horizontal,
The maximum value in the channel G of two diagonal line four directions calculates original graph in vertical, horizontal, two diagonal line four directions
Channel B maximum value；
S222: calculating the maximum value in four values, and the maximum value is denoted as the value of G point；
S23: corrosion, including following sub-step are locked:
S231: the point of left and right, the upper and lower four direction of seed in drawing of seeds is successively traversed, judges whether the point is corroded and sentences
Whether the point that breaks is G point: if the point has been corroded, continuing to judge next point；If the point is not corroded, continue next
Step；
S232: comparing the size of the value of the point and the first threshold nGsensitive of a setting, if the value of the point is greater than threshold value
The point cannot then be corroded, as the point has added one layer of lock；If the value of the point is less than threshold value, in next step, carry out rotten to the left
Erosion；
S233:G point is the point that corrosion is waited judging in the seed left side, is had on the right of seed 0 to N number of point being corroded, to the left corruption
Erosion, including following sub-step:
S2331: judge the quantity of the point being corroded on the right of seed:
(1) when the point not being corroded on the right of seed, then G point can be corroded；
(2) when having 1 to N point being corroded on the right of seed, gray scale, R, G, B four-way of the point of G point and rightmost are calculated
Maximum value, when the maximum value be less than setting second threshold nSensitive, then G point can corrode, and otherwise cannot be corroded
Add one layer of lock to the point；
S2332: the point being corroded becomes seed, and return step S2331 until can not corrode again；
S234: whether monitoring subtitle fringe region reaches Corrosion standards, terminates to corrode if having reached, and otherwise increases corrosion
Second threshold nSensitive repeats above-mentioned steps；
S235: caption data is extracted according to Corrosion results, then noise data is removed further according to character features, finally obtains one
Accurate subtitle position region；
S3: headline, including following sub-step are determined:
S31: the continuity Characteristics according to subtitle in video properties carry out screening and mistake to identical caption area continuous in video
Filter, and choose picture of the effect preferably as identification subtitle；
S32: being converted to text for subtitle picture, as the headline of this then news.
2. a kind of method for extracting caption area in video pictures according to claim 1, it is characterised in that: further include
Sub-step S0 is arranged in one threshold value, before proceeding respectively to first threshold nGsensitive and second threshold nSensitive
It is configured.
3. a kind of method for extracting caption area in video pictures according to claim 1, it is characterised in that: described
N value is 4.
4. a kind of method for extracting caption area in video pictures according to claim 1, it is characterised in that: described
Step S1 further includes following sub-step: video pictures being carried out with lateral band according to the feature of news caption profile consistency from top to bottom
Change identification, and noise removal processing is carried out according to the feature of news caption.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610597947.4A CN106162330B (en) | 2016-07-27 | 2016-07-27 | A method of extracting caption area in video pictures |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610597947.4A CN106162330B (en) | 2016-07-27 | 2016-07-27 | A method of extracting caption area in video pictures |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106162330A CN106162330A (en) | 2016-11-23 |
| CN106162330B true CN106162330B (en) | 2019-04-12 |
Family
ID=58059925
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610597947.4A Active CN106162330B (en) | 2016-07-27 | 2016-07-27 | A method of extracting caption area in video pictures |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106162330B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108769776B (en) * | 2018-05-31 | 2021-03-19 | 北京奇艺世纪科技有限公司 | Title subtitle detection method and device and electronic equipment |
| CN109348289B (en) * | 2018-11-15 | 2021-08-24 | 北京奇艺世纪科技有限公司 | News program title extraction method and device |
| CN111401368B (en) * | 2020-03-24 | 2023-04-18 | 武汉大学 | News video title extraction method based on deep learning |
| CN111931775B (en) * | 2020-09-28 | 2021-01-08 | 成都索贝数码科技股份有限公司 | Method, system, computer device and storage medium for automatically acquiring news headlines |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101115151A (en) * | 2007-07-10 | 2008-01-30 | 北京大学 | Method for extracting video subtitling |
| CN101453575A (en) * | 2007-12-05 | 2009-06-10 | 中国科学院计算技术研究所 | Video subtitle information extracting method |
| CN102202224A (en) * | 2011-06-22 | 2011-09-28 | 清华大学 | Caption flutter-free method and apparatus used for plane video stereo transition |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100718080B1 (en) * | 2005-07-25 | 2007-05-16 | 삼성전자주식회사 | Broadcast receiving device for displaying closed caption data and method thereof |
| JP4253327B2 (en) * | 2006-03-24 | 2009-04-08 | 株式会社東芝 | Subtitle detection apparatus, subtitle detection method, and pull-down signal detection apparatus |
-
2016
- 2016-07-27 CN CN201610597947.4A patent/CN106162330B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101115151A (en) * | 2007-07-10 | 2008-01-30 | 北京大学 | Method for extracting video subtitling |
| CN101453575A (en) * | 2007-12-05 | 2009-06-10 | 中国科学院计算技术研究所 | Video subtitle information extracting method |
| CN102202224A (en) * | 2011-06-22 | 2011-09-28 | 清华大学 | Caption flutter-free method and apparatus used for plane video stereo transition |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106162330A (en) | 2016-11-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106162330B (en) | A method of extracting caption area in video pictures | |
| CN104766076B (en) | A kind of detection method and device of video image character | |
| US8355079B2 (en) | Temporally consistent caption detection on videos using a 3D spatiotemporal method | |
| CN101527043B (en) | Video picture segmentation method based on moving target outline information | |
| CN106548160A (en) | A kind of face smile detection method | |
| CN102663382A (en) | Video image character recognition method based on submesh characteristic adaptive weighting | |
| CN101510304B (en) | Method, device and pick-up head for dividing and obtaining foreground image | |
| US20050190965A1 (en) | Apparatus and method for determining anchor shots | |
| CN103473551A (en) | Station logo recognition method and system based on SIFT operators | |
| CN101216943A (en) | A method for video moving object subdivision | |
| CN102202224A (en) | Caption flutter-free method and apparatus used for plane video stereo transition | |
| CN101742292B (en) | Image content information-based loop filtering method and filter | |
| CN103020970A (en) | Corn ear image grain segmentation method | |
| CN103198479A (en) | SAR image segmentation method based on semantic information classification | |
| CN104598907A (en) | Stroke width figure based method for extracting Chinese character data from image | |
| Ariki et al. | Extraction of TV news articles based on scene cut detection using DCT clustering | |
| CN106373096A (en) | Multi-feature weight adaptive shadow elimination method | |
| CN102780856B (en) | Method for annotating subtitles in news video | |
| CN106612468A (en) | A video abstract automatic generation system and method | |
| CN103400113A (en) | Method for detecting pedestrian on expressway or in tunnel based on image processing | |
| CN109766892A (en) | A kind of ray detection image tagged information character dividing method based on edge detection | |
| CN102938053A (en) | Sugarcane characteristic extraction and recognition method based on computer vision | |
| CN108492306A (en) | A kind of X-type Angular Point Extracting Method based on image outline | |
| CN100548030C (en) | A kind of news main broadcaster's lens detection method based on the space-time strip pattern analysis | |
| CN104992176B (en) | A kind of Chinese character extracting method towards an inscription on a tablet |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |