CN101426137B - Video frame type recognizing method and device - Google Patents
Video frame type recognizing method and device Download PDFInfo
- Publication number
- CN101426137B CN101426137B CN2008102037024A CN200810203702A CN101426137B CN 101426137 B CN101426137 B CN 101426137B CN 2008102037024 A CN2008102037024 A CN 2008102037024A CN 200810203702 A CN200810203702 A CN 200810203702A CN 101426137 B CN101426137 B CN 101426137B
- Authority
- CN
- China
- Prior art keywords
- frame
- video
- value
- identified
- threshold
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T1/00—General purpose image data processing
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
Abstract
The invention discloses a video frame type identification method and device, wherein, the method includes: receiving the video frames to be identified and counting the length values of the video frames to be identified; identifying the types of the video frames to be identified based upon the length values of the video frames to be identified. The device includes: a first statistical unit, for receiving the video frames to be identified and counting the length values of the video frames to be identified; an identification unit, for identifying the types of the video frames to be identified based upon the length values of the video frames to be identified. The embodiment of the present invention uses the lengths of the video frames to identify the types of video frames, without analyzing the head information of each video frame, thereby saving the network overhead.
Description
Technical field
The present invention relates to the stream media technology field, particularly a kind of recognition methods of video frame type and device.
Background technology
At present, development along with wireless communication technique and stream media technology, the streaming media video business has become a kind of important application of 3G, commercial 3G network can provide abundant streaming media video program request and net cast business, and the streaming media video business has also obtained increasing user's favor.
In the streaming media video business, frame of video is the least unit of streaming media playing, can be divided into I frame, P frame and B frame according to type.One group of continuous video frames has just constituted a GOP (Group of Pictures, picture group), and GOP is that a string I frame, P frame and the B frame of start frame formed by the I frame.Wherein, what the I frame adopted is intraframe coding method, and P frame and the employing of B frame is inter-frame encoding.That is, the B frame came the nearest I frame in its front or P frame in reference to GOP and comes nearest I frame in its back or P frame in when coding; The P frame comes nearest I frame in its front or P frame in when coding in reference to GOP.When receiving terminal was decoded, coding can independently not decoded with reference to the I frame of B frame or P frame, and B frame and P frame then must be encoded at them and just can be decoded after the reference frame of time institute's reference decodes out.
The importance difference of dissimilar frame of video, thus also different to the influence degree of video quality.When the video defective appearred in reference frame, the video defective will correspondingly appear in the frame of video with reference to this reference frame when coding.Have only when the video defective appears in the B frame, can not cause the video defective of other frame of video.
If wireless network can identify the type of each frame of video, just dissimilar frame of video can be carried out respectively and corresponding analysis of its type and processing, thereby guarantee the better stream media bearing video traffic of wireless network.
Though the header information of each frame of video has comprised the type identification of indicating this video frame type, can accurately read the type of this frame of video by these type identifications, but, at network side the header information of each frame of video is resolved and can take very large network overhead.
Summary of the invention
The embodiment of the invention provides a kind of recognition methods and device of video frame type, can save the network overhead of the type of identification video frame.
The embodiment of the invention discloses a kind of recognition methods of video frame type, comprising:
Receive frame of video to be identified and add up the length value of described frame of video to be identified; According to the length value of described frame of video to be identified, the type of described frame of video to be identified is discerned in the difference that utilizes dissimilar video length values to exist.
The embodiment of the invention also discloses a kind of recognition device of type of frame of video, comprising:
First statistic unit, the length value that is used to receive frame of video to be identified and adds up described frame of video to be identified; Recognition unit, be used for discerning the type of described frame of video to be identified according to the length value of the frame of video to be identified of described first statistic unit statistics, wherein, the type that the length value of described frame of video to be identified according to described first statistic unit statistics is discerned described frame of video to be identified is: according to the length value of described frame of video to be identified, the type of described frame of video to be identified is discerned in the difference that utilizes dissimilar video length values to exist.
By the embodiment of the invention described above as can be seen, because the length of dissimilar frame of video is different, network side utilizes the length of frame of video to come the type of identification video frame, compared with resolving the type that header information comes the identification video frame, more saves network overhead.
Description of drawings
Fig. 1 is the flow chart of the recognition methods of a kind of video frame type of disclosing of the embodiment of the invention one;
Fig. 2 is the flow chart of the recognition methods of a kind of video frame type of disclosing of the embodiment of the invention two;
Fig. 3 is the subpackage and carrying schematic diagram of frame of video in the prior art;
Fig. 4 determines the flow chart of method for current measurement period among the present invention;
Fig. 5 is the structure chart of the recognition device of a kind of video frame type of disclosing of the embodiment of the invention three.
Embodiment
Below in conjunction with accompanying drawing the embodiment of the invention is described in detail.
Embodiment one
See also Fig. 1, Fig. 1 is the flow chart of the recognition methods of a kind of video frame type of disclosing of the embodiment of the invention one, may further comprise the steps:
Step 101: receive frame of video to be identified and add up the length value of described frame of video to be identified;
Frame of video is broken down into service data unit SDU at network side and transmits in network, so its length value can obtain by the length value of all SDU of statistics frame of video.
Specific implementation, for example: receive service data unit SDU; Identify the SDU that belongs to same frame of video; With the described length value addition summation that belongs to the SDU of same frame of video.
Step 102: the type of discerning described frame of video to be identified according to the length value of described frame of video to be identified.
The inventor finds that there is certain difference in the length value of dissimilar frame of video, can utilize the type of this difference identification frame.For example, under the identical situation of frame per second, the maximum of B frame is generally less than the minimum value of P frame, and the maximum of P frame is generally less than the minimum value of I frame.Therefore the type that can utilize the size of this frame of video to come the identification video frame.
Its embodiment, for example:
Determine the first threshold and second threshold value, described first threshold is the numerical value between B frame length value and P frame length value, and described second threshold value is the numerical value between P frame length value and I frame length value;
When the length value of described frame of video is less than or equal to first threshold, judge that described frame of video is the B frame; Perhaps, when the length value of described frame of video during, judge that described frame of video is the I frame more than or equal to second threshold value; Perhaps, when the length value of described frame of video during, judge that described frame of video is the P frame greater than described first threshold and less than described second threshold value.
First threshold does not have at current code check to remain unchanged under the situation of change, for getting involved a value between the B frame maximum and P frame minimum value within this cycle.
In order to obtain first threshold, the type of identification B frame can realize: determine current period by the following method; Add up first mean value of the frame of video length that receives in the described current period; First coefficient relatively be multiply by the size of the resultant numerical value of described first mean value and default B frame max-thresholds, the minimum value among both is made as described first threshold.When the length value of described frame of video during less than described first threshold, judge that described frame of video is the B frame, wherein, 0＜the first coefficient＜0.6, and H.263, successively decrease successively among the MPEG4, H.264 coded system.
Second threshold value does not have at code check to remain unchanged under the situation of change yet, for getting involved a value between the I frame minimum value and P frame maximum within this cycle.
In order to obtain second threshold value, the type of identification I frame can realize: determine current period by the following method; Add up first mean value of the frame of video length that receives in the described current period; Relatively second coefficient multiply by the size of the resulting numerical value of described first mean value and default I frame minimum threshold, maximum among both is made as described second threshold value, when the length value of described frame of video during greater than described second threshold value, judge that described frame of video is the I frame, wherein, 2＜the second coefficients＜4, and H.263, increase progressively successively among the MPEG4, H.264 coded system.
Wherein, current period is relevant with the situation of change of code check usually, and when code check changed, current period need change accordingly, so that can carry out the identification of video frame type more accurately.Can be made as the finish time of one-period the zero hour of current period, can be made as a code check change moment before the finish time of current period, perhaps shorter.
Definite process of current period comprises, for example: for described current period is provided with an initial period value; B frame number that estimation receives in the described initial period continuously and the I frame number that receives continuously; If the B frame number of described continuous reception is updated to described current period greater than the 3rd threshold value
The time of described t1 for beginning from described initial period timing to be continued during greater than the 3rd threshold value to the B frame number of finding to receive continuously, described the 3rd threshold value is the maximum number that occurs the B frame under the constant situation of code check continuously; Perhaps, if the I frame number of described continuous reception is updated to described current period greater than the 4th threshold value
Wherein, described t2 is the time that is continued when the I frame number of reception is greater than the 4th threshold value continuously from the initial period timing to discovery, and described the 4th threshold value is the maximum number that occurs the I frame under the constant situation of code check continuously.
Wherein, B frame number that receives continuously in the described estimation initial period and the I frame number that receives continuously comprise: second mean value of the frame of video length that statistics received in the described initial period; Statistics in the described initial period, with the length that receives continuously less than the number of the frame of video of the product of first coefficient and described second mean value number as the B frame; Statistics in the described initial period, with the length that receives continuously greater than the number of the frame of video of the product of second coefficient and described second mean value number as the I frame.
For first coefficient and second coefficient, can determine a suitable value according to experience.First coefficient is relevant with the size and the coded system of second coefficient, for the different video file that adopts with a kind of coded format, their first coefficient equates generally speaking, second coefficient also equates generally speaking, therefore also can be in advance the header information of frame of video of video file by resolving the different coding mode, determine pairing first coefficient of different coding mode and second coefficient.And I frame minimum threshold be I frame in a video file length the minimum value that can occur, and B frame max-thresholds refers to the maximum that the length of B frame in the video file can occur, corresponding to different video file with a kind of coded system, its I frame minimum threshold should be identical with B frame max-thresholds, and these two values can obtain by the file of resolving the different coding mode in advance.
The embodiment of the invention can also obtain the first threshold and second threshold value by alternate manner.Such as, can be earlier by resolving the mode of header information, obtain the maximum and the minimum value of P frame, the minimum value of I frame, the maximum of B frame, between the maximum of the minimum value of I frame and P frame, select a value as first threshold then, get a value between the minimum value of the maximum of B frame and P frame as second threshold value, and then calculate first coefficient and second coefficient.
As can be seen from the above-described embodiment, because the length of dissimilar frame of video is different, network side utilizes the length of frame of video to come the type of identification video frame, does not need the header information of each frame of video is resolved, thereby has saved network overhead.
Embodiment two
See also Fig. 2, Fig. 2 is the flow chart of the recognition methods of a kind of video frame type of disclosing of the embodiment of the invention two, and this method may further comprise the steps:
Step 201: determine current period.
Wherein, current period is relevant with the situation of change of code check usually, and when code check changed, current period need change accordingly, so that can carry out the identification of video frame type more accurately.
See also Fig. 4, the definite method for current period specifically may further comprise the steps.
Step 401: for described current period is provided with an initial period value;
Wherein, described initial period value can be to go up the one-period value, or predefined system default value, and this system default value is that reference video file GOP length and frame per second are comprehensively determined, generally select about 1s, but the embodiment of the invention is not limited to this value.
Step 402: parameter i and parameter j are made as 0 respectively, and wherein, parameter i is used to add up the number that occurs the B frame continuously, and parameter j is used to add up the number that occurs the I frame continuously;
Step 403: the mean value of received frame of video length value in the statistics initial period value;
Wherein, the computing formula of the mean value of received frame of video length value is in the initial period value: N
MeanThe total quantity of received frame of video in total length value/initial period of received SDU in the '=initial period value.
Step 404: the length N of adding up the frame of video that is received in the initial period successively
Frame';
This step is a cyclic process.Since the initial period, calculate the length value of the frame of video that receives successively, after a frame of video disposes, continue to receive next frame of video, all videos frame in the initial period is processed finishing all.If in the cyclic process, parameter i greater than the 4th threshold value, illustrates the code check change greater than the 3rd threshold value or parameter j, then upgrades the value of current period, and ends the circulation of this step.
The length value of frame of video can be obtained by the SDU addition that will belong to same frame of video.Except the method for using video frame boundary identification identifies SDU that belongs to each frame of video that receives in the current measurement period and the combination of the carrying out frame of video, can also insert the frame identification with identical time stamp in the RTP bag that transmitting terminal splits same frame of video, network side carries out the combination of frame of video according to frame identification.
Step 405: whether the length of judging the frame of video that is received in the initial period less than the product of the mean value of first coefficient and the frame of video length that receives in the initial period, if, enter step 406, if not, enter step 408;
Step 406: i adds 1 with parameter, and j is made as 0;
Step 407: judge that whether parameter i is greater than the 3rd threshold value, if enter step 411, if not, return step 404, calculate the length value of next frame of video, continue to judge next frame of video, all frame of video in the described initial period all judgement finish;
Wherein, the 3rd threshold value determines comprehensively that by the file type and the coded quantization factor preferred value of present embodiment is 6, but the invention is not restricted to this value.
Step 408: judge that this frame of video is whether greater than the product of the mean value of second coefficient and the frame of video length value that receives in the initial period, if enter step 409; If not, return step 404, continue to judge next frame of video, all frame of video in the described initial period all judgement finish;
Step 409: j adds 1 with parameter, and i is made as 0;
Step 410: judge that whether parameter j is greater than the 4th threshold value, if enter step 412; If not, return step 405, continue to judge next frame of video, all frame of video in the described initial period all judgement finish;
Wherein, the 4th threshold value determines comprehensively that by the file type and the coded quantization factor preferred value of present embodiment is 6, but the invention is not restricted to this value.
Step 411: current period is
Finish this circulation;
Wherein, the frame per second of frame of video is the playback rate of frame of video, and the number of the frame of video of playing in the unit interval is used f
RateExpression as parameter i during greater than the 3rd threshold value, thinks that code check shakes, and marked change can take place in the code check change size of frame of video afterwards, at this moment, need upgrade accordingly the value of current period, and can not adopt the initial period value.
Step 412: current period is
Finish this circulation;
Wherein,, think that code check shakes, need change accordingly current period, and can not adopt the initial period value as parameter j during greater than the 4th threshold value.
When above-mentioned steps 411,412, in any one step determined the value of current period, if step 411,412 is not carried out then used the determined current period value of step 401, turn back in the flow process of identification video frame type, continue execution in step 202.
Step 202: network side receives SDU (Service Data Unit, service data unit) in determined current period;
Wherein, the professional host-host protocol that uses of streaming media video has RTP (Real-time TransportProtocol, RTP), RTCP (RTP Control Protocol, RTP Control Protocol) and RTSP (Real-Time Streaming Protocol, real-time stream media protocol), the agreement of carrying audio frequency, video data is RTP.Transmitting terminal is divided into the plurality of RTP bag with a frame of video, and the RTP bag of these application layers transmits at network side as carrier with SDU, sees also Fig. 3, and Fig. 3 is the subpackage and carrying schematic diagram of frame of video in the prior art.
Step 203: the mean value of the frame of video length value that receives in the statistics current period;
Can be according to the total length value of the SDU that receives in the current period and the total quantity of frame of video, wherein, the computing formula of the mean value of the frame of video length that receives in the current period is: N
MeanThe total quantity of the frame of video that receives in total length value/current period of the SDU that receives in=the current period.
For example, in current period, network side receives 2 frame of video, and the total length of SDU is 10byte, then N
Mean=10/2=5byte.
Step 204: the length value of adding up each frame of video in the current period successively.
Method by video frame boundary identification identifies the SDU that belongs to each frame of video that receives in the current period, and the length value of the SDU of each frame of video is carried out addition, obtains the length value N of each frame of video
Jrame
In the above-mentioned steps, except the method for using video frame boundary identification identifies SDU that belongs to each frame of video that receives in the current period and the combination of the carrying out frame of video, can also insert the frame identification with identical time stamp in the RTP bag that transmitting terminal splits same frame of video, network side carries out the combination of frame of video according to frame identification.
This step is the process of a circulation, after the type identification of a frame of video finishes, continues to calculate the length value of next frame of video, and identification types, and all video frame types are all discerned and finished in current period.
Step 205: whether the mean value of judging a frame of video length that receives in the current period is less than or equal to first threshold, if, enter step 206, if not, enter step 207;
Step 206: judge that this frame of video is the B frame, return step 204, continue to judge the next frame of video that is received in the current period, all frame of video judgements that receive up to current period finish;
Wherein, the type of all frame of video that received in to current period judge finish after, get back to step 201, continue to determine the next cycle value, and continue to judge the type of the frame of video that in next cycle, is received, up to the type of all frame of video all judge finish till.
Step 207: whether judge this frame of video more than or equal to second threshold value, if, enter step 208, if not, enter step 209;
Step 208: judge that this frame of video is the I frame, return step 204, continue to judge the next frame of video that is received in the current period, all frame of video judgements that receive up to current period finish;
Wherein, the type of all frame of video that received in to current period judge finish after, get back to step 201, continue to determine the next cycle value, and continue to judge the type of the frame of video that in next cycle, is received, up to the type of all frame of video all judge finish till.
Step 209: judge that this frame of video is the P frame, return step 204, continue to judge the next frame of video that is received in the current period, all frame of video judgements that receive up to current period finish.
Wherein, the type of all frame of video that received in to current period judge finish after, get back to step 201, continue to determine the next cycle value, and continue to judge the type of the frame of video that in next cycle, is received, up to the type of all frame of video all judge finish till.
Identify the type of frame of video when network side after, can also carry out priority treatment to the high frame of video of priority according to the priority that sets for every type frame of video in advance, thereby guarantee the transmission quality of the frame of video that priority level is high.
As can be seen from the above-described embodiment, the length of dissimilar frame of video is different, and network side is with the length N of each frame of video of obtaining in the current measurement period
FrameCompare with the first threshold and second threshold value respectively, work as N
FrameDuring less than first threshold, judge and described N
FrameCorresponding frame of video is the B frame, works as N
FrameDuring greater than second threshold value, judge and described N
FrameCorresponding frame of video is the I frame, when above-mentioned two conditions do not satisfy, judges and described N
FrameCorresponding frame of video is the P frame, thereby has realized the identification of video frame type on the basis of saving network overhead, and then network side can carry out respectively and the corresponding processing of frame type frame of video according to the type of frame of video.Simultaneously, network side carries out and corresponding analysis of its type and processing dissimilar frame of video according to the type of frame of video, guarantees the better stream media bearing video traffic of wireless network.
Embodiment three
See also Fig. 5, the structure chart of the recognition device of a kind of video type that Fig. 5 the invention process three discloses, this device comprises first statistic unit 501 and recognition unit 502.Operation principle below in conjunction with this device is further introduced its internal structure and interactively.
First statistic unit 501 is used to add up the length value of frame of video to be identified;
Wherein, first statistic unit 501 comprises: receiving element is used to receive service data unit SDU; Recognition unit is used to identify the SDU that belongs to same frame of video; Sum unit is used for the length value addition summation that belongs to the SDU of same frame of video with described.
Recognition unit 502 is used for discerning according to the length value of the frame of video to be identified of described first statistic unit 501 statistics the type of described frame of video to be identified.
Recognition unit 502 further comprises:
First determining unit 5021, second determining unit 5022, frame type judging unit 5023.
First determining unit 5021 is used for determining first threshold that described first threshold is the numerical value between B frame length value and P frame length value;
Wherein, under a concrete enforcement scene, described first determining unit 5021 also can comprise: the current period determining unit is used for determining current period; Second statistic unit is used to add up the mean value of the frame of video length that receives in the determined current period of described current period determining unit as first mean value; First is provided with the unit, is used for the size that comparison first coefficient multiply by the resultant numerical value of described first mean value and default B frame max-thresholds, and the minimum value among both is made as described first threshold.
Second determining unit 5022 is used for determining second threshold value that described second threshold value is the numerical value between P frame length value and I frame length value;
Wherein, under a concrete enforcement scene, described second determining unit 5022 can also comprise: the current period determining unit is used for determining current period; Second statistic unit, the mean value that is used to add up the frame of video length that receives in the determined current period of described current period determining unit is average as first; Second is provided with the unit, is used for the size that comparison second coefficient multiply by the resulting numerical value of described first mean value and default I frame minimum threshold, and the maximum among both is made as described second threshold value.
Above-mentioned current period determining unit comprises: default unit is used to described current period that an initial period value is set; Evaluation unit is used to estimate B frame number that receives continuously and the I frame number that receives continuously in the described initial period; First updating block is if the B frame number of continuous reception that is used for the estimation of described evaluation unit is updated to described current period greater than the 3rd threshold value
The time of described t1 for beginning from described initial period timing to be continued during greater than the 3rd threshold value to the B frame number of finding to receive continuously, described the 3rd threshold value is the maximum number that occurs the B frame under the constant situation of code check continuously; Second updating block is if the I frame number of continuous reception that is used for the estimation of described evaluation unit is updated to described current period greater than the 4th threshold value
Wherein, described t2 is the time that is continued when the I frame number of reception is greater than the 4th threshold value continuously from the initial period timing to discovery, and described the 4th threshold value is the maximum number that occurs the I frame under the constant situation of code check continuously.
Under a concrete enforcement scene, above-mentioned evaluation unit can also comprise: the 3rd statistic unit, and the mean value that is used to add up the frame of video length that receives in the described initial period is as second mean value; The 4th statistic unit is used for statistics in the described initial period, with the length that receives continuously less than the number of the frame of video of the product of first coefficient and described second mean value number as the B frame; The 5th statistic unit is used for statistics in the described initial period, with the length that receives continuously greater than the number of the frame of video of the product of second coefficient and described second mean value number as the I frame.
Frame type judging unit 5023 is used for judging that when institute is to be identified when stating frame of video and being less than or equal to described first threshold described frame of video to be identified is the B frame; When described frame of video to be identified during, judge that described frame of video to be identified is the I frame more than or equal to described second threshold value; When described frame of video to be identified during, judge that described frame of video to be identified is the P frame greater than described first threshold and less than described second threshold value.
As can be seen from the above-described embodiment, added up the length of frame of video when first statistic unit after, utilize the frame of video length of first statistic unit statistics that the type of frame of video is judged by recognition unit, when described frame of video is less than or equal to first threshold, judge that described frame of video is the B frame; When described frame of video during, judge that described frame of video is the I frame more than or equal to second threshold value; When described frame of video during, judge that described frame of video is the P frame greater than described first threshold and less than described second threshold value.Avoided network the header information of each frame of video to be resolved the waste network overhead problem of being brought.Simultaneously, network side carries out and corresponding analysis of its type and processing dissimilar frame of video according to the type of frame of video, guarantees the better stream media bearing video traffic of wireless network.
The embodiment of the invention also can be applicable to video quality assessment, fields such as video quality guarantee.
One of ordinary skill in the art will appreciate that all or part of step that realizes in the foregoing description method is to instruct relevant hardware to finish by program, described program can be stored in the computer read/write memory medium, the length value that this program comprises the steps: to receive frame of video to be identified and adds up described frame of video to be identified; Discern the type of described frame of video to be identified according to the length value of described frame of video to be identified.Described storage medium, as: ROM/RAM, magnetic disc, CD etc.
The above only is a preferred implementation of the present invention; should be pointed out that for those skilled in the art, under the prerequisite that does not break away from the principle of the invention; can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.
Claims (9)
1. the recognition methods of a video frame type is characterized in that, described method comprises:
Receive frame of video to be identified and add up the length value of described frame of video to be identified;
According to the length value of described frame of video to be identified, the type of described frame of video to be identified is discerned in the difference that utilizes the length value of dissimilar frame of video to exist;
Wherein, the difference that exists of the described length value that the utilizes dissimilar frame of video type of discerning described frame of video to be identified comprises:
Determine the first threshold and second threshold value, described first threshold is the numerical value between B frame length value and P frame length value, and described second threshold value is the numerical value between P frame length value and I frame length value;
When the length value of described frame of video to be identified is less than or equal to described first threshold, judge that described frame of video to be identified is the B frame; Perhaps,
When the length value of described frame of video to be identified during, judge that described frame of video to be identified is the I frame more than or equal to described second threshold value; Perhaps,
When the length value of described frame of video to be identified during, judge that described frame of video to be identified is the P frame greater than described first threshold and less than described second threshold value.
2. method according to claim 1 is characterized in that, described definite first threshold comprises:
Determine current period;
The mean value of adding up the frame of video length that receives in the described current period is as first mean value;
Relatively first coefficient multiply by resultant numerical value of described first mean value and default B frame max-thresholds, with the minimum value among both as described first threshold.
3. method according to claim 1 is characterized in that, described definite second threshold value comprises:
Determine current period;
The mean value of adding up the frame of video length that receives in the described current period is as first mean value;
Relatively second coefficient multiply by resulting numerical value of described first mean value and default I frame minimum threshold, with the maximum among both as described second threshold value.
4. according to claim 2 or 3 any described methods, it is characterized in that described definite current period comprises:
For described current period is provided with an initial period value;
B frame number that estimation receives in the described initial period continuously and the I frame number that receives continuously;
If the B frame number of described continuous reception is updated to described current period greater than the 3rd threshold value
Described t
1Be the time that begins from described initial period timing to be continued during greater than the 3rd threshold value to the B frame number of finding to receive continuously, described the 3rd threshold value is the maximum number that occurs the B frame under the constant situation of code check continuously; Perhaps,
If the I frame number of described continuous reception is updated to described current period greater than the 4th threshold value
Wherein, described t
2Be the time that begins from described initial period timing to be continued during greater than the 4th threshold value to the I frame number of finding to receive continuously, described the 4th threshold value is the maximum number that occurs the I frame under the constant situation of code check continuously.
5. method according to claim 4 is characterized in that, B frame number that described estimation receives in the described initial period continuously and the I frame number that receives continuously comprise:
The mean value of the frame of video length value that statistics received in the described initial period is second mean value;
Statistics in the described initial period, with the length value that receives continuously less than the number of the frame of video of the product of first coefficient and described second mean value number as the B frame;
Statistics in the described initial period, with the length value that receives continuously greater than the number of the frame of video of the product of second coefficient and described second mean value number as the I frame.
6. the recognition device of the type of a frame of video is characterized in that, comprising:
First statistic unit, the length value that is used to receive frame of video to be identified and adds up described frame of video to be identified;
Recognition unit, be used for discerning the type of described frame of video to be identified according to the length value of the frame of video to be identified of described first statistic unit statistics, wherein, the type that the length value of described frame of video to be identified according to described first statistic unit statistics is discerned described frame of video to be identified is: according to the length value of described frame of video to be identified, the type of described frame of video to be identified is discerned in the difference that utilizes the length value of dissimilar frame of video to exist;
Wherein, described recognition unit comprises:
First determining unit is used for determining first threshold that described first threshold is the numerical value between B frame length value and P frame length value;
Second determining unit is used for determining second threshold value that described second threshold value is the numerical value between P frame length value and I frame length value;
The frame type judging unit is used for when the length value of described frame of video to be identified is less than or equal to described first threshold, judges that described frame of video to be identified is the B frame; Perhaps, when the length value of described frame of video to be identified during, judge that described frame of video to be identified is the I frame more than or equal to described second threshold value; Perhaps, when the length value of described frame of video to be identified during, judge that described frame of video to be identified is the P frame greater than described first threshold and less than described second threshold value.
7. device according to claim 6 is characterized in that, described first determining unit comprises:
The current period determining unit is used for determining current period;
Second statistic unit is used to add up the mean value of the frame of video length value that receives in the determined current period of described current period determining unit as first mean value;
First is provided with the unit, is used for the size that comparison first coefficient multiply by the resultant numerical value of described first mean value and default B frame max-thresholds, and the minimum value among both is made as described first threshold.
8. device according to claim 6 is characterized in that, described second determining unit comprises:
The current period determining unit is used for determining current period;
Second statistic unit is used to add up the mean value of the frame of video length that receives in the determined current period of described current period determining unit as first mean value;
Second is provided with the unit, is used for the size that comparison second coefficient multiply by the resulting numerical value of described first mean value and default I frame minimum threshold, and the maximum among both is made as described second threshold value.
9. according to claim 7 or 8 any described devices, it is characterized in that described current period determining unit comprises:
Default unit is used to described current period that an initial period value is set;
Evaluation unit is used to estimate B frame number that receives continuously in the initial period that is provided with in described default unit and the I frame number that receives continuously;
First updating block is if the B frame number of continuous reception that is used for the estimation of described evaluation unit is updated to described current period greater than the 3rd threshold value
Described t
1Be the time that begins from described initial period timing to be continued during greater than the 3rd threshold value to the B frame number of finding to receive continuously, described the 3rd threshold value is the maximum number that occurs the B frame under the constant situation of code check continuously;
Second updating block is if the I frame number of continuous reception that is used for the estimation of described evaluation unit is updated to described current period greater than the 4th threshold value
Wherein, described t2 is the time that is continued when the I frame number of reception is greater than the 4th threshold value continuously from the initial period timing to discovery, and described the 4th threshold value is the maximum number that occurs the I frame under the constant situation of code check continuously.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008102037024A CN101426137B (en) | 2008-11-25 | 2008-11-25 | Video frame type recognizing method and device |
| PCT/CN2009/075123 WO2010060376A1 (en) | 2008-11-25 | 2009-11-25 | Video frame type identification method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008102037024A CN101426137B (en) | 2008-11-25 | 2008-11-25 | Video frame type recognizing method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101426137A CN101426137A (en) | 2009-05-06 |
| CN101426137B true CN101426137B (en) | 2011-08-03 |
Family
ID=40616429
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2008102037024A Active CN101426137B (en) | 2008-11-25 | 2008-11-25 | Video frame type recognizing method and device |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN101426137B (en) |
| WO (1) | WO2010060376A1 (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102752022A (en) * | 2012-06-29 | 2012-10-24 | 清华大学深圳研究生院 | Power line communication method and frame response method and device thereof |
| CN102752023A (en) * | 2012-06-29 | 2012-10-24 | 清华大学深圳研究生院 | Electric terminal equipment and system based on wireless HART (Highway Addressable Remote Transducer) and power wire communication |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101426137B (en) * | 2008-11-25 | 2011-08-03 | 上海华为技术有限公司 | Video frame type recognizing method and device |
| CN102547300B (en) * | 2010-12-17 | 2015-01-21 | 华为技术有限公司 | Method for detecting frame types and device |
| CN103634698B (en) * | 2012-08-21 | 2014-12-03 | 华为技术有限公司 | Methods for detecting frame type and frame size of video stream and apparatus |
| CN104283699A (en) * | 2013-07-01 | 2015-01-14 | 中兴通讯股份有限公司 | Method and device for determining service types |
| CN108023869B (en) * | 2016-10-28 | 2021-03-19 | 海能达通信股份有限公司 | Parameter adjusting method and device for multimedia communication and mobile terminal |
| CN111191484A (en) * | 2018-11-14 | 2020-05-22 | 普天信息技术有限公司 | Method and device for recognizing human speaking in video image |
| CN112019852A (en) * | 2019-05-29 | 2020-12-01 | 杭州海康威视数字技术股份有限公司 | Target type frame identification method and device and electronic equipment |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1964250A (en) * | 2006-11-23 | 2007-05-16 | 华为技术有限公司 | Data frame type identification method and device |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2005322291A (en) * | 2004-05-07 | 2005-11-17 | Matsushita Electric Ind Co Ltd | Reproducing unit and reproducing method |
| CN101426137B (en) * | 2008-11-25 | 2011-08-03 | 上海华为技术有限公司 | Video frame type recognizing method and device |
-
2008
- 2008-11-25 CN CN2008102037024A patent/CN101426137B/en active Active
-
2009
- 2009-11-25 WO PCT/CN2009/075123 patent/WO2010060376A1/en active Application Filing
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1964250A (en) * | 2006-11-23 | 2007-05-16 | 华为技术有限公司 | Data frame type identification method and device |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102752022A (en) * | 2012-06-29 | 2012-10-24 | 清华大学深圳研究生院 | Power line communication method and frame response method and device thereof |
| CN102752023A (en) * | 2012-06-29 | 2012-10-24 | 清华大学深圳研究生院 | Electric terminal equipment and system based on wireless HART (Highway Addressable Remote Transducer) and power wire communication |
| CN102752022B (en) * | 2012-06-29 | 2014-09-10 | 清华大学深圳研究生院 | Power line communication method and frame response method and device thereof |
| CN102752023B (en) * | 2012-06-29 | 2014-10-22 | 清华大学深圳研究生院 | Electric terminal equipment and system based on wireless HART (Highway Addressable Remote Transducer) and power wire communication |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101426137A (en) | 2009-05-06 |
| WO2010060376A1 (en) | 2010-06-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101426137B (en) | Video frame type recognizing method and device | |
| US9025673B2 (en) | Temporal quality metric for video coding | |
| CN104125429B (en) | The adjusting method and device of video data transmission | |
| US20130188742A1 (en) | Method and apparatus for encoder assisted-frame rate up conversion (ea-fruc) for video compression | |
| CN102480618B (en) | Realize H264 video code model and play the method and system optimized | |
| US9030565B2 (en) | Method, apparatus, and system for evaluating video quality | |
| CN101389029B (en) | Method and apparatus for video image encoding and retrieval | |
| JP2007529125A (en) | System and method for compressing media signals using different codecs | |
| Dou et al. | Edge computing-enabled deep learning for real-time video optimization in IIoT | |
| CN109660805B (en) | Audio and video synchronous optimization method, storage medium, equipment and system in decoding and playing | |
| CN113438501B (en) | Video compression method, apparatus, computer device and storage medium | |
| WO2010051834A1 (en) | Method and system for determining a quality value of a video stream | |
| KR20150003316A (en) | Method and device for evaluating media quality | |
| CN109511011B (en) | YouTube DASH encrypted video-oriented fingerprint database construction method | |
| CN102130886A (en) | Network video streaming media system, transmission processing method, transmitting end and receiving end | |
| WO2023174254A1 (en) | Video posting method and apparatus, and device and storage medium | |
| CN114339454A (en) | Audio and video synchronization method and device, electronic device and storage medium | |
| CN102075784A (en) | Video quality assessment method under condition of comprehensively considering compression and packet loss impairment | |
| US11638051B2 (en) | Real-time latency measurement of video streams | |
| CN111447451A (en) | Image coding and decoding method and device | |
| US20080080379A1 (en) | Network device and frame processing method thereof | |
| Ververidis et al. | Evaluation of quality of experience for video streaming over dynamic spectrum access systems | |
| CN105187688A (en) | Method and system for carrying out synchronization on real-time video and audio collected by mobile phone | |
| Hsu et al. | On the accuracy and complexity of rate-distortion models for fine-grained scalable video sequences | |
| KR101373414B1 (en) | Packet-level video streaming prioritization scheme and system enhancing quality of h.264/avc over mpeg-2 transport stream |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |