Summary of the invention
It is not easy the technical issues of recognizing in order to solve nationality where image scene in the prior art, the present invention provides one kind
Nationality's information intelligent recognition methods.
For this purpose, the present invention needs following inventive point important everywhere:
(1) in the subsequent processing of film, to head audio fragment and the corresponding nationality's information of head video content
Consistency analysis is carried out, to judge whether the audio file of mispairing different language;
(2) in nationality's information extraction to head video content, high-precision customization image procossing mechanism is used;
(3) by the extraction and analysis to the target in image to be processed, judge the wire-frame image vegetarian refreshments in image to be processed
Sum avoids when excessive with the wire-frame image vegetarian refreshments in the image to be processed using the filtering bad to profile filter effect
Device, and select airspace smothing filtering mode;
(4) when the wire-frame image vegetarian refreshments in the image to be processed is few, selection is more preferably smart using inhibition low-frequency effect
Fine filtering wave mode.
According to an aspect of the present invention, a kind of nationality's information intelligent recognition methods is provided, this method includes using a kind of state
Nationality information intelligent identifying system identifies nationality's information, and nationality's information intelligent identifying system includes: touch screen, compiles with audio
It collects device to connect with video editor, be visited for providing user to the editor of audio editing device and the editor of video editor for user
Ask interface;Audio editing device extracts head from the audio data to be edited for receiving audio data to be edited
Audio fragment, the audio data to be edited and the video data to be edited belong to same film;Video editor,
For receiving video data to be edited, multiple head images are extracted from the video data to be edited, wherein described
Multiple timestamps of multiple head images of the extraction of video editor keep equidistant time interval.
More specifically, in nationality's information intelligent identifying system, further includes:
Time set device is connect with the video editor, for set the video editor extraction it is multiple
Equidistant time interval between multiple timestamps of head image;Multiplicity evaluation apparatus is connect with the video editor,
For receiving the multiple head image, identifies the scene of each head image, will there is the head image for repeating scene to draw
It is one group, to obtain multiple groups head image, using the minimum image of entropy in each group of head image as target image, with output
The corresponding multiple target images of multiple groups head image.
More specifically, in nationality's information intelligent identifying system, further includes:
Objective extraction equipment is connect with the multiplicity evaluation apparatus, for receiving the target image, to the target
Each target in image extracts, and to obtain corresponding each target subgraph, each target subgraph only includes group
At each pixel of respective objects, each profile of its profile for corresponding to target of composition is determined based on each target subgraph
Pixel.
More specifically, in nationality's information intelligent identifying system, further includes:
Edge analysis equipment is connect with the Objective extraction equipment, for receiving each target in the target image
The corresponding each wire-frame image vegetarian refreshments of subgraph adds up the corresponding each contour pixel of each target subgraph in the target image
Point exports the profile point sum to obtain profile point sum;Filtering triggering equipment, connect with the edge analysis equipment,
For receiving the profile point sum, and when the profile point sum is less than default points threshold value, the first triggering letter is issued
Number, and when the profile point sum is more than default points threshold value, issue the second trigger signal;Combined processing equipment, respectively
Equipment is triggered with the filtering and the Objective extraction equipment is connect, including logarithm operation unit, Fourier transform unit, high pass
Filter unit, inverse Fourier transform unit and exponent arithmetic unit, for receiving described first in first processing equipment
When trigger signal, to the target image successively execute logarithm operation, Fourier transformation, high-pass filtering, inverse Fourier transform and
Exponent arithmetic, to obtain and export signal conversion image corresponding with the target image;Single treatment equipment, respectively with it is described
Filtering triggering equipment is connected with the Objective extraction equipment, for when receiving first trigger signal, to the target
Image executes airspace the disposal of gentle filter, converts image to obtain signal corresponding with the target image;Power supply management device,
It is connect respectively with the combined processing equipment and the single treatment equipment, in the combined processing equipment or described single
When processing equipment enters battery saving mode, controls and the electricity of the combined processing equipment or the single treatment equipment is supplied;State
Nationality judges equipment, connect with the single treatment equipment, for receiving each signal conversion image, converts to each signal
Image executes following operation: the signal being converted the performer region that the depth of field is most shallow in image and converts image from the signal
In split, based on the judgement of various countries' standardized human body's profile and the most shallow performer region of the depth of field in signal conversion image
The title human body contour outline that is best suitable for of profile corresponding to nationality using as the corresponding nationality of signal conversion image;Nationality's system
Equipment is counted, judges that equipment is connect with the nationality, for receiving the corresponding each nationality of each signal conversion image, is based on
Each nationality judges the corresponding video nationality classification of video data to be edited;Consistency analysis equipment, respectively with it is described
Nationality counts equipment and connects with the audio editing device, for carrying out constituent analysis to head audio fragment, with determination described
The ownership nationality of music in head audio fragment, to the ownership nationality of the music in the head audio fragment with it is described to be edited
The corresponding video nationality classification of video data carry out consistency analysis, to obtain corresponding analysis result;Wherein, in the state
Nationality counts in equipment, judges that the corresponding video nationality classification of video data to be edited includes: by institute based on each nationality
The highest nationality of the frequency of occurrences is stated in each nationality as the corresponding video nationality classification of video data to be edited.
More specifically, in nationality's information intelligent identifying system: in the combined processing equipment, the Fourier
Converter unit is connect with the logarithm operation unit and the high pass filter unit respectively, the inverse Fourier transform unit and institute
High pass filter unit connection is stated, the exponent arithmetic unit is connect with the inverse Fourier transform unit.
More specifically, in nationality's information intelligent identifying system: the combined processing equipment is receiving described
When two trigger signals, battery saving mode is switched to from operating mode.
More specifically, in nationality's information intelligent identifying system: the single treatment equipment is receiving described
When one trigger signal, battery saving mode is switched to from operating mode.
More specifically, in nationality's information intelligent identifying system: in the consistency analysis equipment, when described point
When analysing result more than limitation, nationality's corresponding instruction is issued, otherwise, sending nationality, which misplaces, to be instructed.
Specific embodiment
Embodiment of the present invention is described in detail below with reference to accompanying drawings.
The important technology of audio-video editor is characterized in that the sound of the technology in addition to turning video with picture, outstanding profession regards
Frequency is edited, and be should also have and is added various Multi-media Materials for original image, realizes the displaying of the videography produced and cyclopentadienyl, example
Such as, match music for picture, add MTV subtitle effect, various photograph transition transition special effectives etc., these are all outstanding audio-videos
Edit the notable feature that must have.
In order to overcome above-mentioned deficiency, the present invention builds a kind of nationality's information intelligent recognition methods, and this method includes using one
Nationality's information intelligent identifying system is planted to identify nationality's information.Nationality's information intelligent identifying system can effectively solve the problem that accordingly
The technical issues of.
Fig. 1 is according to the video editor of nationality's information intelligent identifying system shown in embodiment of the present invention to be edited
Video data in multiple head images for extracting.Multiple head images are continuous on a timeline.
The nationality's information intelligent identifying system shown according to an embodiment of the present invention includes:
Touch screen is connect with audio editing device and video editor, for for user to audio editing device editor and view
The editor of frequency editing machine provides user's access interface;
Audio editing device extracts slice from the audio data to be edited for receiving audio data to be edited
Head audio fragment, the audio data to be edited and the video data to be edited belong to same film;
Video editor extracts more for receiving video data to be edited from the video data to be edited
A head image, wherein multiple timestamps of multiple head images of the extraction of the video editor are kept for the equidistant time
Interval.
Then, continue that the specific structure of nationality's information intelligent identifying system of the invention is further detailed.
In nationality's information intelligent identifying system, further includes:
Time set device is connect with the video editor, for set the video editor extraction it is multiple
Equidistant time interval between multiple timestamps of head image;
Multiplicity evaluation apparatus is connect with the video editor, for receiving the multiple head image, is identified each
The scene of a head image will there is the head image for repeating scene to divide one group into, to obtain multiple groups head image, by each group
The minimum image of entropy is as target image in head image, to export the corresponding multiple target figures of multiple groups head image
Picture.
In nationality's information intelligent identifying system, further includes:
Objective extraction equipment is connect with the multiplicity evaluation apparatus, for receiving the target image, to the target
Each target in image extracts, and to obtain corresponding each target subgraph, each target subgraph only includes group
At each pixel of respective objects, each profile of its profile for corresponding to target of composition is determined based on each target subgraph
Pixel.
In nationality's information intelligent identifying system, further includes:
Edge analysis equipment is connect with the Objective extraction equipment, for receiving each target in the target image
The corresponding each wire-frame image vegetarian refreshments of subgraph adds up the corresponding each contour pixel of each target subgraph in the target image
Point exports the profile point sum to obtain profile point sum;
Filtering triggering equipment, connect with the edge analysis equipment, for receiving the profile point sum, and in the wheel
When wide point sum is less than default points threshold value, the first trigger signal is issued, and in the profile point sum more than preset
When number threshold value, the second trigger signal is issued;
Combined processing equipment triggers equipment with the filtering respectively and the Objective extraction equipment is connect, including logarithm fortune
Unit, Fourier transform unit, high pass filter unit, inverse Fourier transform unit and exponent arithmetic unit are calculated, for described
When first processing equipment receives first trigger signal, logarithm operation is successively executed to the target image, Fourier becomes
It changes, high-pass filtering, inverse Fourier transform and exponent arithmetic, to obtain and export signal transition diagram corresponding with the target image
Picture;
Single treatment equipment triggers equipment with the filtering respectively and the Objective extraction equipment is connect, for receiving
When to first trigger signal, airspace the disposal of gentle filter is executed to the target image, to obtain and the target image
Corresponding signal converts image;
Power supply management device is connect with the combined processing equipment and the single treatment equipment respectively, for described
When combined processing equipment or the single treatment equipment enter battery saving mode, control is to the combined processing equipment or described single
The electricity of processing equipment is supplied;
Nationality judges equipment, connect with the single treatment equipment, for receiving each signal conversion image, to each
A signal conversion image executes following operation: the signal is converted the depth of field is most shallow in image performer region from the letter
It is split in number conversion image, converts in image with the signal that the depth of field is most shallow drills based on the judgement of various countries' standardized human body's profile
Nationality corresponding to the title human body contour outline that the profile of member region is best suitable for is corresponding to convert image as the signal
Nationality;
Nationality counts equipment, judges that equipment is connect with the nationality, respectively corresponds for receiving each signal conversion image
Each nationality, the corresponding video nationality classification of video data to be edited is judged based on each nationality;
Consistency analysis equipment counts equipment with the nationality respectively and the audio editing device is connect, for head
Audio fragment carries out constituent analysis, with the ownership nationality of the music in the determination head audio fragment, to the head audio
The ownership nationality of music in segment video nationality classification corresponding with the video data to be edited carries out consistency analysis,
To obtain corresponding analysis result;
Wherein, it is counted in equipment in the nationality, judges that video data to be edited is corresponding based on each nationality
Video nationality's classification includes: that the highest nationality of the frequency of occurrences in each nationality is corresponding as video data to be edited
Video nationality's classification.
In nationality's information intelligent identifying system: in the combined processing equipment, the Fourier transform unit
It is connect respectively with the logarithm operation unit and the high pass filter unit, the inverse Fourier transform unit and the high pass are filtered
The connection of wave unit, the exponent arithmetic unit are connect with the inverse Fourier transform unit.
In nationality's information intelligent identifying system: the combined processing equipment is receiving second trigger signal
When, battery saving mode is switched to from operating mode.
In nationality's information intelligent identifying system: the single treatment equipment is receiving first trigger signal
When, battery saving mode is switched to from operating mode.
In nationality's information intelligent identifying system: in the consistency analysis equipment, when the analysis result is super
When crossing limitation, nationality's corresponding instruction is issued, otherwise, issues nationality's dislocation instruction.
In addition, in nationality's information intelligent identifying system: the consistency analysis equipment can type selecting be general array
Logical device GAL.
General array logic chip (Generic Array Logic www.husoon.com) device is 1985
LATTICE company invent at first it is erasable remove, may be programmed, the PLD of settable encrypted bits.Representative GAL chip has
GAL16V8, GAL20, both GAL can almost emulate all types of PAL devices.In practical application, GAL device is to PAL
Device simulation has 100% compatibility, so GAL can almost replace PAL device entirely, and may replace most of SSI, MSI
Digital integrated electronic circuit, thus be widely applied.
The maximum differential of GAL and PAL is that the export structure of GAL can be defined by the user, and is a kind of programmable output knot
Structure.Two kinds of base type GAL16V8 (20 pin) GAL20V8 (24 pin) of GAL can replace ten kinds of PAL devices of tree, thus claim
To use programmable circuit bitterly.And the output of PAL is defined by producer, is just secured after chip is selected, user can not change.
Using nationality's information intelligent identifying system of the invention, it is not easy to distinguish for nationality where image scene in the prior art
The technical issues of recognizing believes head audio fragment and the corresponding nationality of head video content in the subsequent processing of film
Breath carries out consistency analysis, to judge whether the audio file of mispairing different language;In nationality's information to head video content
In extraction, high-precision customization image procossing mechanism is used;By the extraction and analysis to the target in image to be processed, sentence
The wire-frame image vegetarian refreshments sum to break in image to be processed, when excessive with the wire-frame image vegetarian refreshments in the image to be processed, avoids making
With the filter bad to profile filter effect, and select airspace smothing filtering mode;And in the image to be processed
When wire-frame image vegetarian refreshments is few, selection is asked using low-frequency effect more preferably finer filter mode is inhibited to solve above-mentioned technology
Topic.
It is understood that although the present invention has been disclosed in the preferred embodiments as above, above-described embodiment not to
Limit the present invention.For any person skilled in the art, without departing from the scope of the technical proposal of the invention,
Many possible changes and modifications all are made to technical solution of the present invention using the technology contents of the disclosure above, or are revised as
With the equivalent embodiment of variation.Therefore, anything that does not depart from the technical scheme of the invention are right according to the technical essence of the invention
Any simple modifications, equivalents, and modifications made for any of the above embodiments still fall within the range of technical solution of the present invention protection
It is interior.