CN104598921A - Video preview selecting method and device - Google Patents
Video preview selecting method and device Download PDFInfo
- Publication number
- CN104598921A CN104598921A CN201410852257.XA CN201410852257A CN104598921A CN 104598921 A CN104598921 A CN 104598921A CN 201410852257 A CN201410852257 A CN 201410852257A CN 104598921 A CN104598921 A CN 104598921A
- Authority
- CN
- China
- Prior art keywords
- sectional drawing
- gray
- score value
- value
- face
- 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.)
- Pending
Links
Abstract
The invention discloses a video preview selecting method and device. The method comprises receiving a video and performing a set number of random screenshots on the video; performing face number detection and gray level computation on every screenshot, and according to the face number and the gray level computing result of every screenshot, assigning scores for every screenshot; according to the scores of the screenshots, selecting one of the screenshots as the preview of the video. The video preview selecting method achieves modeling design from the aspects of the face number, the clear degree, the color richness and the color fluctuation of the screenshots to digitize various influence factors, combines the influence factors through weighting, and finally evaluates the screen shots by quantizing the influence factors, thereby providing a high-practicality screenshot evaluating mechanism and being capable of automatically screening the screenshots to accurately and stably screen out reasonable video previews.
Description
Technical field
The application relates to digital image processing techniques field, is specifically related to choosing method and the selecting device of a kind of video preview figure.
Background technology
Along with the fast development of multimedia technology and Internet technology, increasing multimedia messages is propagated on network.Because video can carry abundanter, more lively information compared to other medias such as text, image and sound, receive and accept extensively and like.Current video website is watched for user containing a large amount of videos, and user is when browsing webpage, usually wishes in the video library of magnanimity, to find oneself interested video to watch in the short time.Therefore add that preview graph becomes a kind of method improving browse efficiency for each video.
At present for the method for video interpolation preview graph mainly contains following several, the first intercepts a frame picture, as the preview graph of this video exactly at random in video.This is the scheme that current each large video website generally adopts.This method simple practical, but there are some problems.Such as, owing to being random intercepting, may there is black, Quan Bai, the situation such as fuzzy entirely in the preview graph obtained, can not show the content of video well, the object finally making user understand video content by preview graph cannot realize.Another kind is the method for artificial screening, and the random multiframe of intercepting in video picture, goes out the preview graph of subjective best picture as this video by editorial staff's artificial screening afterwards usually.Although this method can obtain ideal preview graph, because screening process needs manual intervention, in the face of the video of magnanimity, the cost of labor of this screening and time cost are all very high.
How to be automatically performed by computer, accurately, stably filter out rational video preview figure, and reach the effect close to artificial screening, just become technical matters urgently to be resolved hurrily.
Summary of the invention
The object of the application is the choosing method and the selecting device that provide a kind of video preview figure, is automatically performed by computer, accurately, stably filters out rational video preview figure.
In order to solve the problems of the technologies described above, this application discloses the choosing method of a kind of video preview figure, comprising: receiver, video also carries out the random sectional drawing of predetermined quantity to video; Respectively every width sectional drawing is carried out to detection and the gray count of face number, according to face number and the gray count result of described every secondary sectional drawing, for every secondary sectional drawing gives score value; The preview graph of a sectional drawing as described video is chosen according to the score value of every width sectional drawing.
Further, described gray count, comprising: the readability detecting every width sectional drawing, calculates the rich color degree in every width sectional drawing and color fluctua degree; Described gray count result, comprising: the readability of every width sectional drawing, rich color degree and color fluctua degree.
Further, be that every width sectional drawing gives score value, comprise: the score value of described every width sectional drawing, with the score value of face number, the score value of readability, the score value positive correlation of rich color degree, with the score value negative correlation of described color fluctua degree; Wherein, the number of described face less and described face in described sectional drawing, present position is the closer to middle part, the score value of the face number of described sectional drawing is higher; The width at the edge in described sectional drawing is less, and the score value of the readability of described sectional drawing is higher; The lowest gray value of described sectional drawing and the difference of the highest gray-scale value larger, the score value of the rich color degree of described sectional drawing is higher; The relative standard deviation of the gray level image of described sectional drawing larger, the score value of the color fluctua degree of described sectional drawing is higher.
Further, for arbitrary sectional drawing, identify the type of described arbitrary sectional drawing, for the sectional drawing type identified, for score value, the score value of readability, the score value of rich color degree, the score value of color fluctua degree of face number, choose corresponding weight allocation plan, and according to score value, the score value of readability, the score value of rich color degree, the score value of color fluctua degree of face number, in conjunction with the weight in weight allocation plan being often kind of score value configuration, for described arbitrary sectional drawing gives score value.
Further, every width sectional drawing is carried out to the detection of face number, comprising: for every width sectional drawing, feature detection is carried out by the face characteristic operator preset, identify the face in image, wherein, described face characteristic operator is for detecting the position proportional relation in face between each key feature points; Be the score value that described sectional drawing gives face number according to the face identified, wherein, the number of described face less and described face in described sectional drawing, present position is the closer to middle part, the score value of the face number of described sectional drawing is higher.
Further, detect the readability of every width sectional drawing, comprising: for every width sectional drawing, by the edge detection operator preset, rim detection computing is carried out to sectional drawing, the width at the edge detected is identified; According to the score value that the width at the edge identified is sectional drawing imparting readability, wherein, in described sectional drawing, the width at edge is less, and the score value of the readability of described sectional drawing is higher.
Further, calculate the rich color degree in every width sectional drawing, comprise: for every width sectional drawing, it is converted to gray level image by coloured image, obtain the gray-scale value of each pixel in described gray level image, statistics draws its intensity histogram diagram data, has the pixel of lowest gray value and the highest gray-scale value in gray level image according to its grey level histogram data search; According to the score value that described lowest gray value and the highest gray-scale value are described sectional drawing imparting rich color degree, wherein, the difference of described lowest gray value and the highest gray-scale value is larger, and the score value of the rich color degree of described sectional drawing is higher.
Further, calculate the color fluctua degree in every width sectional drawing, comprising: be positioned at the pixel number of each gray-scale value between the two according to described lowest gray value and the highest gray-scale value by intercepting in described intensity histogram diagram data; According to by the total number of pixel intercepted out in described intensity histogram diagram data and the total number by the gray-scale value intercepted out in described intensity histogram diagram data, calculate the mean value of pixel distribution, according to the relative standard deviation of gray level image described in by the pixel number of each gray-scale value intercepted out in described intensity histogram diagram data and described mean value calculation; Be the score value that described sectional drawing gives color fluctua degree according to the relative standard deviation of described gray level image, wherein, the relative standard deviation of the gray level image of described sectional drawing larger, the score value of the color fluctua degree of described sectional drawing is higher.
Further, there is in gray level image according to its grey level histogram data search the pixel of lowest gray value and the highest gray-scale value, comprise: from first gray-scale value of described intensity histogram diagram data, pixel number corresponding to described gray-scale value is successively inquired about, and recording pixel point number is greater than first gray-scale value of 0 as lowest gray value; From last gray-scale value of described intensity histogram diagram data, pixel number corresponding to described gray-scale value is successively inquired about, and recording pixel point number is greater than first gray-scale value of 0 as the highest gray-scale value; Wherein, described intensity histogram diagram data presses the ascending order arrangement of gray-scale value.
In order to solve the problems of the technologies described above, itself please further disclose the selecting device of a kind of video preview figure, comprising: screen capture module, video be carried out to the random sectional drawing of predetermined quantity for receiver, video; Assignment module, is respectively used to the detection and the gray count that every width sectional drawing are carried out to face number, according to face number and the gray count result of described every secondary sectional drawing, for every secondary sectional drawing gives score value; Choose module, for choosing the preview graph of a sectional drawing as described video according to the score value of every width sectional drawing.
Further, the gray count that described assignment module is carried out, comprising: the readability detecting every width sectional drawing, calculates the rich color degree in every width sectional drawing and color fluctua degree; The described gray count result obtained, comprising: the readability of every width sectional drawing, rich color degree and color fluctua degree.
Further, described assignment module is the score value that described every width sectional drawing is given, with the score value of face number, the score value of readability, the score value positive correlation of rich color degree, with the score value negative correlation of described color fluctua degree; Wherein, the number of described face less and described face in described sectional drawing, present position is the closer to middle part, the score value of the face number of described sectional drawing is higher; The width at the edge in described sectional drawing is less, and the score value of the readability of described sectional drawing is higher; The lowest gray value of described sectional drawing and the difference of the highest gray-scale value larger, the score value of the rich color degree of described sectional drawing is higher; The relative standard deviation of the gray level image of described sectional drawing larger, the score value of the color fluctua degree of described sectional drawing is higher.
Further, described assignment module, for arbitrary sectional drawing, identify the type of described arbitrary sectional drawing, for the sectional drawing type identified, for score value, the score value of readability, the score value of rich color degree, the score value of color fluctua degree of face number, choose corresponding weight allocation plan, and according to score value, the score value of readability, the score value of rich color degree, the score value of color fluctua degree of face number, in conjunction with the weight in weight allocation plan being often kind of score value configuration, for described arbitrary sectional drawing gives score value.
Further, described assignment module, for every width sectional drawing, carry out feature detection by the face characteristic operator preset, identify the face in image, wherein, described face characteristic operator is for detecting the position proportional relation in face between each key feature points; Be the score value that sectional drawing gives face number according to the face identified, wherein, the number of described face less and described face in described sectional drawing, present position is the closer to middle part, the score value of the face number of described sectional drawing is higher.
Further, described assignment module, for every width sectional drawing, for the edge detection operator by presetting, carrying out rim detection computing to sectional drawing, identifying the width at the edge detected; According to the score value that the width at the edge identified is sectional drawing imparting readability, wherein, the width at the edge in described sectional drawing is less, and the score value of the readability of described sectional drawing is higher.
Further, described assignment module, for every width sectional drawing, for it is converted to gray level image by coloured image, obtains the gray-scale value of each pixel in described gray level image; Gray-scale value ordered series of numbers is counted according to the span of gray-scale value, add up the pixel number of each gray-scale value in described gray-scale value ordered series of numbers according to pixels all in described gray level image and gray-scale value thereof, search the pixel in described gray level image with lowest gray value and the highest gray-scale value according to described gray-scale value ordered series of numbers and corresponding pixel number; According to the score value that described lowest gray value and the highest gray-scale value are described sectional drawing imparting rich color degree, wherein, the difference of described lowest gray value and the highest gray-scale value is larger, and the score value of the rich color degree of described sectional drawing is higher.
Further, described assignment module, for according to described lowest gray value and the highest gray-scale value by the pixel number intercepting part gray-scale value ordered series of numbers and the corresponding each gray-scale value of described part gray-scale value ordered series of numbers be arranged between the two in described gray-scale value ordered series of numbers; According to total number of gray-scale value in total number of pixel corresponding in described part gray-scale value ordered series of numbers and described part gray-scale value ordered series of numbers, calculate the mean value of the pixel distribution on described part gray-scale value ordered series of numbers, the relative standard deviation of gray level image according to the pixel number of each gray-scale value in described part gray-scale value ordered series of numbers and described mean value calculation; Be the score value that described sectional drawing gives color fluctua degree according to the relative standard deviation of described gray level image, wherein, the relative standard deviation of the gray level image of described sectional drawing larger, the score value of the color fluctua degree of described sectional drawing is higher.
Further, described assignment module, for from first gray-scale value of described gray-scale value ordered series of numbers, pixel number corresponding to described gray-scale value is successively inquired about, and recording pixel point number is greater than first gray-scale value of 0 as lowest gray value; From last gray-scale value of described gray-scale value ordered series of numbers, pixel number corresponding to described gray-scale value is successively inquired about, and recording pixel point number is greater than first gray-scale value of 0 as the highest gray-scale value; Wherein, described intensity histogram diagram data presses the ascending order arrangement of gray-scale value.
Compared with prior art, the application can obtain and comprise following technique effect:
The application carries out Modeling and Design from the angle of the face number of sectional drawing, readability, rich color degree and color fluctua degree, by various factors datumization, and influence factor is combined by weighting, the most above-mentioned influence factor mode is by quantifying evaluated sectional drawing, thus provide the very strong sectional drawing evaluation mechanism of a kind of practicality, and by this mechanism, automatic screening is carried out to sectional drawing, accurately, stably filter out rational video preview figure.
Accompanying drawing explanation
Accompanying drawing described herein is used to provide further understanding of the present application, and form a application's part, the schematic description and description of the application, for explaining the application, does not form the improper restriction to the application.In the accompanying drawings:
Fig. 1 is the method flow diagram of the embodiment of the present application one.
Fig. 2 is the method flow diagram of the embodiment of the present application two.
Fig. 3 is the method flow diagram of the embodiment of the present application three.
Fig. 4 is the method flow diagram of the embodiment of the present application four.
Fig. 5 is the method flow diagram of the embodiment of the present application five.
Fig. 6 is the structure drawing of device of the embodiment of the present application six.
Embodiment
Drawings and Examples will be coordinated below to describe the embodiment of the application in detail, by this to the application how application technology means solve technical matters and the implementation procedure reaching technology effect can fully understand and implement according to this.
As employed some vocabulary to censure specific components in the middle of instructions and claim.Those skilled in the art should understand, and hardware manufacturer may call same assembly with different noun.This specification and claims are not used as with the difference of title the mode distinguishing assembly, but are used as the criterion of differentiation with assembly difference functionally." comprising " as mentioned in the middle of instructions and claim is in the whole text an open language, therefore should be construed to " comprise but be not limited to "." roughly " refer to that in receivable error range, those skilled in the art can solve the technical problem within the scope of certain error, reach described technique effect substantially.In addition, " couple " or " electric connection " one word comprise directly any and indirectly electric property coupling means at this.Therefore, if describe a first device in literary composition to be coupled to one second device, then represent described first device and directly can be electrically coupled to described second device, or be indirectly electrically coupled to described second device by other device or the means that couple.Instructions subsequent descriptions is implement the better embodiment of the application, and right described description is for the purpose of the rule that the application is described, and is not used to the scope limiting the application.The protection domain of the application is when being as the criterion depending on the claims person of defining.
Also it should be noted that, term " comprises ", " comprising " or its other variant any are intended to contain comprising of nonexcludability, thus make to comprise the process of a series of key element, method, commodity or system and not only comprise those key elements, but also comprise other key element clearly do not listed, or also comprise by the intrinsic key element of this process, method, commodity or system.When not more restrictions, the key element limited by statement " comprising ... ", and be not precluded within process, method, commodity or the system comprising described key element and also there is other identical element.Described in the embodiment of the present application, " transcoding " can also be called " coding ".
embodiment one
The choosing method of a kind of video preview figure that the embodiment of the present application discloses, as shown in Figure 1, it comprises the following steps:
Step S100, receiver, video also carries out the sectional drawing of predetermined quantity to video;
Sectional drawing can adopt random interception way, and also can intercept according to anchor-frame position, the application does not limit this.
For the ease of enough good video preview figure finally can be found, need to ensure certain sample size, sample size can not be allowed again excessive simultaneously, cause being calculated to be merits and demerits high, in practical operation, predetermined quantity is chosen to be 16 sectional drawings, and certain the application does not limit this.
Step S102, carries out detection and the gray count of face number respectively to every width sectional drawing, according to face number and the gray count result of described every secondary sectional drawing, for every secondary sectional drawing gives score value.
Described gray count, comprising: the readability detecting every width sectional drawing, calculates the rich color degree in every width sectional drawing and color fluctua degree; Described gray count result, comprising: the readability of every width sectional drawing, rich color degree and color fluctua degree.
For every width sectional drawing, computational analysis can be carried out according to picture, the score value of the score value of face number, the score value of readability, the score value of rich color degree and color fluctua degree is obtained according to different analysis strategies; Then do computing according to these score values, finally obtain the score value of every width sectional drawing.
In general, the score value score of described every width sectional drawing, with the score value sj positive correlation of the score value sf of face number, the score value sm of readability and rich color degree, with color fluctua degree sc negative correlation, is obtained by following formula 1.1:
Score=sf+2 × sj-1+sm-sc formula 1.1
For calculating the realization flow of score value of face number see embodiment two.For calculating the realization flow of score value of readability see embodiment three.For calculating the realization flow of score value of rich color degree see embodiment four.For calculating the realization flow of score value of color fluctua degree see embodiment five.
For arbitrary sectional drawing, identify the type of described arbitrary sectional drawing, for the sectional drawing type identified, for score value, the score value of readability, the score value of rich color degree, the score value of color fluctua degree of face number, choose corresponding weight allocation plan, and according to score value, the score value of readability, the score value of rich color degree, the score value of color fluctua degree of face number, in conjunction with the weight in weight allocation plan being often kind of score value configuration, for described arbitrary sectional drawing gives score value, improve accuracy, the objectivity of choosing preview graph.
Such as, for the sectional drawing of non-animation video, some sectional drawings may be scene environment classes, and some sectional drawings may be figure image classes, for from different classifications, the emphasis of sectional drawing is different, but can effectively preview graph use; But for figure image class sectional drawing, rich color degree possibility can not be very high, but the score value of face number may be higher, for scene environment class sectional drawing, rich color degree may be very high, but the score value of face number may be lower, if that is not sub-category, same weight is used for all marks, the quality of sectional drawing can not be gone out by actual response like this.Therefore can be directed to different sectional drawing types, corresponding weight scheme is set, make scoring more objective.Such as figure image class sectional drawing, image content is exactly mainly face close shot, so the quality of facial image directly affects the quality of sectional drawing, therefore the score value of face number, the score value of readability obviously should obtain larger weight, and the weight of the score value of the score value of rich color degree, color fluctua degree should correspondingly reduce.Again such as scene environment class sectional drawing, image content is exactly mainly environment distant view, the overall content of distant view and colour match more can highlight the quality of picture, therefore the score value of face number, the score value of readability are not obviously very important, and the weight of the score value of rich color degree, the score value of color fluctua degree should correspondingly improve.Above-mentioned example is only illustrative explanation, does not limit the weight allocation plan of the application.
Step S104, chooses the preview graph of a sectional drawing as described video according to the score value of every width sectional drawing.
In this application, more meet the requirements because the scoring of the score value of the score value of the score value of face number, readability, rich color degree is all image, these score values are higher.And the score value of color fluctua degree is image more meets the requirements, this score value is lower.Therefore, with the score value of face number, the score value of readability, the score value positive correlation of rich color degree and higher with the score value of every width sectional drawing of the score value negative correlation of color fluctua degree, show that sectional drawing more meets the requirements, therefore, using the preview graph of sectional drawing the highest for score value as described video.
Certainly, more meet the requirements if the scoring of the score value of the score value of the score value of face number, readability, rich color degree is all image, these score values are lower.The score value of color fluctua degree is that image more meets the requirements, and this score value is higher.That is just using the preview graph of sectional drawing minimum for score value as described video.
embodiment two
What the embodiment of the present application disclosed is in the step S102 of Fig. 1 for the realization flow of score value calculating face number, and as shown in Figure 2, it comprises the following steps:
Step S1020, for every width sectional drawing, carry out feature detection by the face characteristic operator preset, identify the face in image, wherein, described face characteristic operator is for detecting the position proportional relation in face between each key feature points;
By the face characteristic operator preset, feature detection is carried out to sectional drawing, tentatively identifies in sectional drawing the part and background parts that meet face characteristic; By the face characteristic operator preset, again feature detection is carried out to the part meeting face characteristic in the sectional drawing tentatively identified, identifies the part wherein meeting face characteristic, continuous iteration, until be accurately identified the part meeting face characteristic.
Step S1022 is the score value that sectional drawing gives face number according to the face identified, wherein, the number of described face less and described face in described sectional drawing, present position is the closer to middle part, the score value of the face number of described sectional drawing is higher.
For the sectional drawing of non-animation class video, preferably face is there is in described sectional drawing, certainly, the face of face is not The more the better, because too much face can cause visual unclear in a jumble, less and be positioned in the middle of screen giving prominence to the key points of sectional drawing is described, and composition is simply clear for face, and such sectional drawing is suitable as preview graph.
Facial features localization uses Viola, P.A. and Jones, M.J 2004 propose based on Haar-like feature, the Face datection algorithm of Adaboost sorter and Cascade cascade classifier.After completing Face datection, according to following rule, the scoring obtaining this module is quantized to image.If the scope up and down of face is ft, fb, fl, fr; Final score is sf.
If there is no face in image, then sf=0;
If there is 1 face in image, and the region conforms formula 2.1 of face, then sf score is as shown in formula 2.2.
sf＝1.2+2*max((fr-fl),(fb-ft))/height+(fr-fl)/|(fl+fr)/2-width/2|
Formula 2.2
If there are 2 faces in image, then judge each human face region whether coincidence formula 2.1, if meet, then current score is designated as sfi (i=0,1), as shown in formula 2.3, otherwise sfi=0.And final score sf is as shown in formula 2.4.
Sf
i=2*max ((fr-fl), (fb-ft))/height formula 2.3
If there are more than 2 faces in image, then judge each human face region whether coincidence formula 2.1, if meet, then current score is designated as sfi (i=0,1 ... n.n is positive integer), as shown in formula 2.3, otherwise sfi=0.And final score sf is as shown in formula 2.5.
Sf=0.4+max (sf
i) formula 2.5
Need to illustrate, above-mentioned formula is just in order to preferably describe the present embodiment, and it does not make restriction to the protection domain of the application, and other modes also can realize the application.
embodiment three
What the embodiment of the present application disclosed is in the step S102 of Fig. 1 for the realization flow of score value calculating readability, and as shown in Figure 3, it comprises the following steps:
Step S1120, for every width sectional drawing, by the edge detection operator preset, carries out rim detection computing to sectional drawing, identifies the width at the edge detected;
For every width sectional drawing, after carrying out rim detection computing, convert described sectional drawing to shade of gray image, identify that in the shade of gray image of described face, gradient gray-scale value is higher than the connected component of a threshold value, described connected component is edge.
Edge detection operator is first difference operator, adopts sobel (Sobel) operator to realize.Sobel operator comprises horizontal sobel operator and longitudinal sobel operator, is all the matrix of 3x3.Horizontal sobel operator and longitudinal sobel operator are made planar convolution with sectional drawing respectively, horizontal and longitudinal brightness difference approximate value can be drawn respectively, thus obtain shade of gray image.Certainly, other operators also can be used to carry out rim detection, and such as canny operator, the application is not restricted to this.
Step S1122, be the score value that sectional drawing gives readability according to the width at the edge identified, wherein, the width at the edge in described sectional drawing is less, and the score value of the readability of described sectional drawing is higher.
If the width at edge is very wide, illustrate that current sectional drawing is a bit fuzzy, such image is obviously second-rate.Can arrange the threshold value of multiple width, the corresponding score value of each threshold value, score value takes from the floating number between 0 ~ 1.Threshold value is less, and score value is higher.Go and threshold value comparison according to the width identified, search corresponding score value according to residing threshold interval, this score value is exactly the score value of the readability of described sectional drawing.
Certainly, also have a kind of mode, be by the width at edge divided by a constant, thus the width at edge is converted to the score value of the floating number between 0 ~ 1 as readability, this constant can get value larger in the width of sectional drawing and the length of sectional drawing.
embodiment four
What the embodiment of the present application disclosed is in the step S102 of Fig. 1 for the realization flow of score value calculating rich color degree, and as shown in Figure 4, it comprises the following steps:
Step S1220, for every width sectional drawing, is converted to gray level image by it by coloured image, obtains the gray-scale value of each pixel in described gray level image;
The conversion of gray level image can according to such as under type: the gray-scale value calculating each pixel according to R, G, B component of each pixel, thus coloured image is converted to gray level image, wherein, in described coloured image, each pixel stores R, G, B component of respective pixel point.
The gray-scale value gray of each pixel according to R, G, B component of each pixel, can be obtained by following formula 4.1:
Gray=R*0.299+G*0.587+B*0.114 formula 4.1
Need to illustrate, above-mentioned formula is just in order to preferably describe the present embodiment, and it does not make restriction to the protection domain of the application, and other modes also can realize the application.
Step S1222, statistics draws its intensity histogram diagram data, has the pixel of lowest gray value and the highest gray-scale value in gray level image according to its grey level histogram data search.
Specifically, gray-scale value ordered series of numbers is counted according to the span of gray-scale value, add up the pixel number of each gray-scale value in described gray-scale value ordered series of numbers according to pixels all in described gray level image and gray-scale value thereof, search the pixel in described gray level image with lowest gray value and the highest gray-scale value according to described gray-scale value ordered series of numbers and corresponding pixel number.
Search when there is the pixel of lowest gray value and the highest gray-scale value in described gray level image, from first gray-scale value of described gray-scale value ordered series of numbers, pixel number corresponding to described gray-scale value is successively inquired about, and recording pixel point number is greater than first gray-scale value of 0 as lowest gray value; From last gray-scale value of described gray-scale value ordered series of numbers, pixel number corresponding to described gray-scale value is successively inquired about, and recording pixel point number is greater than first gray-scale value of 0 as the highest gray-scale value.
Described gray-scale value span is the darkest gray-scale value of 0 ~ 255,0 representative, the gray-scale value that 255 representatives are the brightest, forms one by the ascending tactic gray-scale value ordered series of numbers 0 ~ 255 of gray-scale value according to the span of described gray-scale value.
Be described with an example.In described gray level image, detect and whether there is the pixel that gray-scale value is 0, if had, be the number record hist [0] of the pixel of 0 by gray-scale value, if do not had, gray-scale value is the number of the pixel of 0 is 0, be recorded in 0 in hist [0], add up the pixel number of each gray-scale value in described gray-scale value ordered series of numbers by similar mode, for gray-scale value non-existent in described gray level image, corresponding pixel number is 0, obtains hist [0] ~ hist [255] like this.
From array first element, traversal array, if currency hist [i] (0≤i≤255) are greater than 0, then stop traversal, and utilizes currency i to be exactly lowest gray value, carry out standardization obtain sb=i/255 to lowest gray value.In like manner, from last element of array, traversal array, if currency hist [i] (0≤i≤255) are greater than 0, then stop traversal, and utilizes currency i to be exactly the highest gray-scale value, carry out standardization obtain sw=i/255 to the highest gray-scale value.
Step S1224, be the score value that described sectional drawing gives rich color degree according to described lowest gray value and the highest gray-scale value, wherein, the difference of described lowest gray value and the highest gray-scale value is larger, and the score value of the rich color degree of described sectional drawing is higher.
Connect above-mentioned example.The score value sj=sw-sb of rich color degree.Like this, the difference of the highest gray-scale value and lowest gray value is larger, and illustrate that, from the brightest more to the darkest gray-scale value, corresponding color is abundanter, the image quality of sectional drawing is better, and such sectional drawing is obviously more suitable for as preview graph.
embodiment five
What the embodiment of the present application disclosed is in the step S102 of Fig. 1 for the realization flow of score value calculating color fluctua degree, and as shown in Figure 5, it comprises the following steps:
Step S1320, for every width sectional drawing, is converted to gray level image by it by coloured image, obtains the gray-scale value of each pixel in described gray level image.
The conversion of gray level image can according to such as under type: the gray-scale value calculating each pixel according to R, G, B component of each pixel, thus coloured image is converted to gray level image, wherein, in described coloured image, each pixel stores R, G, B component of respective pixel point.The gray-scale value gray of each pixel according to R, G, B component of each pixel, can be obtained by above-mentioned formula 4.1.Need to illustrate, above-mentioned formula is just in order to preferably describe the present embodiment, and it does not make restriction to the protection domain of the application, and other modes also can realize the application.
Step S1322, statistics draws its intensity histogram diagram data, has the pixel of lowest gray value and the highest gray-scale value in gray level image according to its grey level histogram data search.
Specifically, gray-scale value ordered series of numbers is counted according to the span of gray-scale value, add up the pixel number (this is exactly intensity histogram diagram data) of each gray-scale value in described gray-scale value ordered series of numbers according to pixels all in described gray level image and gray-scale value thereof, search the pixel in described gray level image with lowest gray value and the highest gray-scale value according to described gray-scale value ordered series of numbers and corresponding pixel number.
Search when there is the pixel of lowest gray value and the highest gray-scale value in described gray level image, from first gray-scale value of described gray-scale value ordered series of numbers, pixel number corresponding to described gray-scale value is successively inquired about, and recording pixel point number is greater than first gray-scale value of 0 as lowest gray value; From last gray-scale value of described gray-scale value ordered series of numbers, pixel number corresponding to described gray-scale value is successively inquired about, and recording pixel point number is greater than first gray-scale value of 0 as the highest gray-scale value.
Described gray-scale value span is the darkest gray-scale value of 0 ~ 255,0 representative, the gray-scale value that 255 representatives are the brightest, forms one by the ascending tactic gray-scale value ordered series of numbers 0 ~ 255 of gray-scale value according to the span of described gray-scale value.
Be described with an example.In described gray level image, detect and whether there is the pixel that gray-scale value is 0, if had, be the number record hist [0] of the pixel of 0 by gray-scale value, if do not had, gray-scale value is the number of the pixel of 0 is 0, be recorded in 0 in hist [0], add up the pixel number of each gray-scale value in described gray-scale value ordered series of numbers by similar mode, for gray-scale value non-existent in described gray level image, corresponding pixel number is 0, obtains hist [0] ~ hist [255] like this.
From array first element, traversal array, if currency hist [i] (0≤i≤255) are greater than 0, then stop traversal, and utilizes currency i to be exactly lowest gray value.In like manner, from last element of array, traversal array, if currency hist [i] (0≤i≤255) are greater than 0, then stop traversal, and utilizes currency i to be exactly the highest gray-scale value.
Step S1324, is positioned at the pixel number of each gray-scale value between the two according to described lowest gray value and the highest gray-scale value by intercepting in described intensity histogram diagram data.
Specifically, according to described lowest gray value and the highest gray-scale value by the pixel number intercepting part gray-scale value ordered series of numbers and the corresponding each gray-scale value of described part gray-scale value ordered series of numbers be arranged between the two in described gray-scale value ordered series of numbers.
Connect above-mentioned example.According to lowest gray value and the highest gray-scale value, statistics data between the two in hist [0] ~ hist [255], these data are exactly the pixel number of corresponding each gray-scale value in part gray-scale value ordered series of numbers between the two, obviously, gray-scale value between lowest gray value and the highest gray-scale value not necessarily can find in gray level image, and the pixel number of these gray-scale values that can not find is 0.
Step S1326, according to by the total number of pixel intercepted out in described intensity histogram diagram data and the total number by the gray-scale value intercepted out in described intensity histogram diagram data, calculate the mean value of pixel distribution, according to the relative standard deviation of gray level image described in by the pixel number of each gray-scale value intercepted out in described intensity histogram diagram data and described mean value calculation.
Specifically, according to total number of gray-scale value in total number of pixel corresponding in described part gray-scale value ordered series of numbers and described part gray-scale value ordered series of numbers, calculate the mean value of the pixel distribution on described part gray-scale value ordered series of numbers, the relative standard deviation of gray level image according to the pixel number of each gray-scale value in described part gray-scale value ordered series of numbers and described mean value calculation.
Connect above-mentioned example.The pixel number of corresponding each gray-scale value in the part gray-scale value ordered series of numbers found above is sued for peace, thus obtain total number of pixel, by the total number of the total number of pixel divided by gray-scale value in described part gray-scale value ordered series of numbers, just obtain the mean value ave of the pixel distribution on described part gray-scale value ordered series of numbers.
The standard deviation sd of gray level image is used for the degree of the pixel number deviation average ave weighing each gray-scale value on described part gray-scale value ordered series of numbers, standard deviation sd is less, on described part gray-scale value ordered series of numbers, the pixel number deviation average ave of each gray-scale value is fewer, and vice versa.
And the size of relative standard deviation is weighed by the multiplying power relation of standard deviation and mean value.
Therefore, the relative standard deviation that formula 5.1 can be utilized to calculate image divides sc:
Sc=sd/ave formula 5.1
Need to illustrate, above-mentioned formula is just in order to preferably describe the present embodiment, and it does not make restriction to the protection domain of the application, and other modes also can realize the application.
Step S1328 is the score value that described sectional drawing gives color fluctua degree according to the relative standard deviation of described gray level image, and wherein, the relative standard deviation of the gray level image of described sectional drawing larger, the score value of the color fluctua degree of described sectional drawing is higher.
COLOR COMPOSITION THROUGH DISTRIBUTION score value is exactly the size of relative standard deviation.The relative standard deviation of the gray level image of described sectional drawing larger, illustrate that the pixel number deviation average ave of interior (i.e. corresponding described part gray-scale value ordered series of numbers) each gray-scale value of the chromatic zones of gray level image is less, the color fluctua degree of key diagram picture is inviolent, this image color is soft is easy to viewing, be very suitable for as preview graph, therefore the score value of color fluctua degree is higher.
embodiment six
The selecting device of a kind of video preview figure that the embodiment of the present application also discloses, as shown in c Fig. 6, comprising: screen capture module 600, assignment module 602 and choose module 604.
Screen capture module 600, carries out the random sectional drawing of predetermined quantity for receiver, video to video;
Assignment module 602, is respectively used to the detection and the gray count that every width sectional drawing are carried out to face number, according to face number and the gray count result of described every secondary sectional drawing, for every secondary sectional drawing gives score value; The gray count that described assignment module is carried out, comprising: the readability detecting every width sectional drawing, calculates the rich color degree in every width sectional drawing and color fluctua degree; The described gray count result obtained, comprising: the readability of every width sectional drawing, rich color degree and color fluctua degree;
Choose module 604, for choosing the preview graph of a sectional drawing as described video according to the score value of every width sectional drawing.In this application, more meet the requirements because the scoring of the score value of the score value of the score value of face number, readability, rich color degree is all image, these score values are higher.And the score value of color fluctua degree is image more meets the requirements, this score value is lower.Therefore, with the score value of face number, the score value of readability, the score value positive correlation of rich color degree and higher with the score value of every width sectional drawing of the score value negative correlation of color fluctua degree, show that sectional drawing more meets the requirements, therefore, using the preview graph of sectional drawing the highest for score value as described video.Certainly, more meet the requirements if the scoring of the score value of the score value of the score value of face number, readability, rich color degree is all image, these score values are lower.The score value of color fluctua degree is that image more meets the requirements, and this score value is higher.That is just using the preview graph of sectional drawing minimum for score value as described video.
When carrying out the detection of face number to every width sectional drawing, described assignment module 602, for every width sectional drawing, feature detection is carried out by the face characteristic operator preset, identify the face in image, wherein, described face characteristic operator is for detecting the position proportional relation in face between each key feature points; Be the score value that sectional drawing gives face number according to the face identified, wherein, the number of described face less and described face in described sectional drawing, present position is the closer to middle part, the score value of the face number of described sectional drawing is higher.
When identifying the readability of every width sectional drawing, for every width sectional drawing, described assignment module 602, by the edge detection operator preset, is carried out rim detection computing to sectional drawing, is identified the width at the edge detected; According to the score value that the width at the edge identified is sectional drawing imparting readability, wherein, the width at the edge in described sectional drawing is less, and the score value of the readability of described sectional drawing is higher.For every width sectional drawing, after described assignment module 602 pairs of sectional drawings carry out rim detection computing, convert described sectional drawing to shade of gray image, identify that in the shade of gray image of described face, gradient gray-scale value is higher than the connected component of a threshold value, described connected component is the edge of face.
When calculating the rich color degree in every width sectional drawing, for every width sectional drawing, it is converted to gray level image by coloured image by described assignment module 602, obtain the gray-scale value of each pixel in described gray level image, statistics draws its intensity histogram diagram data, has the pixel of lowest gray value and the highest gray-scale value in gray level image according to its grey level histogram data search; According to the score value that described lowest gray value and the highest gray-scale value are described sectional drawing imparting rich color degree, wherein, the difference of described lowest gray value and the highest gray-scale value is larger, and the score value of the rich color degree of described sectional drawing is higher.
When calculating the color fluctua degree in every width sectional drawing, described assignment module 602 is positioned at the pixel number of each gray-scale value between the two according to described lowest gray value and the highest gray-scale value by intercepting in described intensity histogram diagram data; According to by the total number of pixel intercepted out in described intensity histogram diagram data and the total number by the gray-scale value intercepted out in described intensity histogram diagram data, calculate the mean value of pixel distribution, according to the relative standard deviation of gray level image described in by the pixel number of each gray-scale value intercepted out in described intensity histogram diagram data and described mean value calculation; Be the score value that described sectional drawing gives color fluctua degree according to the relative standard deviation of described gray level image, wherein, the relative standard deviation of the gray level image of described sectional drawing larger, the score value of the color fluctua degree of described sectional drawing is higher.According to described gray-scale value ordered series of numbers and corresponding pixel number search in described gray level image, there is the pixel of lowest gray value and the highest gray-scale value time, described assignment module 602 is from first gray-scale value of described intensity histogram diagram data, pixel number corresponding to described gray-scale value is successively inquired about, and recording pixel point number is greater than first gray-scale value of 0 as lowest gray value; From last gray-scale value of described intensity histogram diagram data, pixel number corresponding to described gray-scale value is successively inquired about, and recording pixel point number is greater than first gray-scale value of 0 as the highest gray-scale value.
Described assignment module 602 is the score value of every width sectional drawing, with the score value of face number, the score value of readability, the score value positive correlation of rich color degree, with the score value negative correlation of described color fluctua degree.
The technical scheme of this device and functional character, the connected mode of each module, corresponding with the characteristic sum technical scheme described in previous embodiment one to five, weak point refers to previous embodiment one to five.
Above-mentioned explanation illustrate and describes the embodiment of the application, but as previously mentioned, be to be understood that the application is not limited to the form disclosed by this paper, should not regard the eliminating to other embodiments as, and can be used for other combinations various, amendment and environment, and can in invention contemplated scope described herein, changed by the technology of above-mentioned instruction or association area or knowledge.And the change that those skilled in the art carry out and change do not depart from the spirit and scope of the application, then all should in the protection domain of the application's claims.
Claims (14)
1. a choosing method of video preview figure, is characterized in that, comprising:
Receiver, video also carries out the random sectional drawing of predetermined quantity to video;
Respectively every width sectional drawing is carried out to detection and the gray count of face number, according to face number and the gray count result of described every secondary sectional drawing, for every secondary sectional drawing gives score value;
The preview graph of a sectional drawing as described video is chosen according to the score value of every width sectional drawing.
2. choosing method as claimed in claim 1, is characterized in that,
Described gray count, comprising: the readability detecting every width sectional drawing, calculates the rich color degree in every width sectional drawing and color fluctua degree;
Described gray count result, comprising: the readability of every width sectional drawing, rich color degree and color fluctua degree.
3. choosing method as claimed in claim 2, is characterized in that, for every width sectional drawing gives score value, comprises further:
The score value of described every width sectional drawing, with the score value of face number, the score value of readability, the score value positive correlation of rich color degree, with the score value negative correlation of described color fluctua degree;
Wherein, the number of described face less and described face in described sectional drawing, present position is the closer to middle part, the score value of the face number of described sectional drawing is higher; The width at the edge in described sectional drawing is less, and the score value of the readability of described sectional drawing is higher; The lowest gray value of described sectional drawing and the difference of the highest gray-scale value larger, the score value of the rich color degree of described sectional drawing is higher; The relative standard deviation of the gray level image of described sectional drawing larger, the score value of the color fluctua degree of described sectional drawing is higher.
4. choosing method as claimed in claim 3, is characterized in that,
For arbitrary sectional drawing, identify the type of described arbitrary sectional drawing, for the sectional drawing type identified, for score value, the score value of readability, the score value of rich color degree, the score value of color fluctua degree of face number, choose corresponding weight allocation plan, and according to score value, the score value of readability, the score value of rich color degree, the score value of color fluctua degree of face number, in conjunction with the weight being often kind of score value configuration, be that described arbitrary sectional drawing gives score value in weight allocation plan.
5. choosing method as claimed in claim 1, is characterized in that, every width sectional drawing is carried out to the detection of face number, comprising:
For every width sectional drawing, carry out feature detection by the face characteristic operator preset, identify the face in image, wherein, described face characteristic operator is for detecting the position proportional relation in face between each key feature points;
Be the score value that described sectional drawing gives face number according to the face identified, wherein, the number of described face less and described face in described sectional drawing, present position is the closer to middle part, the score value of the face number of described sectional drawing is higher.
6. choosing method as claimed in claim 2, is characterized in that, detect the readability of every width sectional drawing, comprising:
For every width sectional drawing, by the edge detection operator preset, rim detection computing is carried out to sectional drawing, the width at the edge detected is identified;
According to the score value that the width at the edge identified is sectional drawing imparting readability, wherein, in described sectional drawing, the width at edge is less, and the score value of the readability of described sectional drawing is higher.
7. choosing method as claimed in claim 2, is characterized in that, calculate the rich color degree in every width sectional drawing, comprising:
For every width sectional drawing, it is converted to gray level image by coloured image, obtain the gray-scale value of each pixel in described gray level image, statistics draws its intensity histogram diagram data, has the pixel of lowest gray value and the highest gray-scale value in gray level image according to its grey level histogram data search;
According to the score value that described lowest gray value and the highest gray-scale value are described sectional drawing imparting rich color degree, wherein, the difference of described lowest gray value and the highest gray-scale value is larger, and the score value of the rich color degree of described sectional drawing is higher.
8. choosing method as claimed in claim 7, is characterized in that, calculate the color fluctua degree in every width sectional drawing, comprising:
The pixel number of each gray-scale value is between the two positioned at by intercepting in described intensity histogram diagram data according to described lowest gray value and the highest gray-scale value;
According to by the total number of pixel intercepted out in described intensity histogram diagram data and the total number by the gray-scale value intercepted out in described intensity histogram diagram data, calculate the mean value of pixel distribution, according to the relative standard deviation of gray level image described in by the pixel number of each gray-scale value intercepted out in described intensity histogram diagram data and described mean value calculation;
Be the score value that described sectional drawing gives color fluctua degree according to the relative standard deviation of described gray level image, wherein, the relative standard deviation of the gray level image of described sectional drawing larger, the score value of the color fluctua degree of described sectional drawing is higher.
9. choosing method as claimed in claim 7, is characterized in that having the pixel of lowest gray value and the highest gray-scale value in gray level image according to its grey level histogram data search, comprise further:
From first gray-scale value of described intensity histogram diagram data, pixel number corresponding to described gray-scale value is successively inquired about, and recording pixel point number is greater than first gray-scale value of 0 as lowest gray value;
From last gray-scale value of described intensity histogram diagram data, pixel number corresponding to described gray-scale value is successively inquired about, and recording pixel point number is greater than first gray-scale value of 0 as the highest gray-scale value;
Wherein, described intensity histogram diagram data presses the ascending order arrangement of gray-scale value.
10. a selecting device of video preview figure, is characterized in that, comprising:
Screen capture module, carries out the random sectional drawing of predetermined quantity for receiver, video to video;
Assignment module, is respectively used to the detection and the gray count that every width sectional drawing are carried out to face number, according to face number and the gray count result of described every secondary sectional drawing, for every secondary sectional drawing gives score value;
Choose module, for choosing the preview graph of a sectional drawing as described video according to the score value of every width sectional drawing.
11. selecting devices as claimed in claim 10, is characterized in that,
The gray count that described assignment module is carried out, comprising: the readability detecting every width sectional drawing, calculates the rich color degree in every width sectional drawing and color fluctua degree; The described gray count result obtained, comprising: the readability of every width sectional drawing, rich color degree and color fluctua degree.
12. selecting devices as claimed in claim 11, is characterized in that,
Described assignment module is the score value that described every width sectional drawing is given, with the score value of face number, the score value of readability, the score value positive correlation of rich color degree, with the score value negative correlation of described color fluctua degree; Wherein, the number of described face less and described face in described sectional drawing, present position is the closer to middle part, the score value of the face number of described sectional drawing is higher; The width at the edge in described sectional drawing is less, and the score value of the readability of described sectional drawing is higher; The lowest gray value of described sectional drawing and the difference of the highest gray-scale value larger, the score value of the rich color degree of described sectional drawing is higher; The relative standard deviation of the gray level image of described sectional drawing larger, the score value of the color fluctua degree of described sectional drawing is higher.
13. selecting devices as claimed in claim 12, is characterized in that,
Described assignment module, for arbitrary sectional drawing, identify the type of described arbitrary sectional drawing, for the sectional drawing type identified, for score value, the score value of readability, the score value of rich color degree, the score value of color fluctua degree of face number, choose corresponding weight allocation plan, and according to score value, the score value of readability, the score value of rich color degree, the score value of color fluctua degree of face number, in conjunction with the weight in weight allocation plan being often kind of score value configuration, for described arbitrary sectional drawing gives score value.
14. selecting devices as claimed in claim 10, is characterized in that,
Described assignment module, for every width sectional drawing, carry out feature detection by the face characteristic operator preset, identify the face in image, wherein, described face characteristic operator is for detecting the position proportional relation in face between each key feature points; Be the score value that sectional drawing gives face number according to the face identified, wherein, the number of described face less and described face in described sectional drawing, present position is the closer to middle part, the score value of the face number of described sectional drawing is higher.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410852257.XA CN104598921A (en) | 2014-12-31 | 2014-12-31 | Video preview selecting method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410852257.XA CN104598921A (en) | 2014-12-31 | 2014-12-31 | Video preview selecting method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104598921A true CN104598921A (en) | 2015-05-06 |
Family
ID=53124693
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410852257.XA Pending CN104598921A (en) | 2014-12-31 | 2014-12-31 | Video preview selecting method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104598921A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106686403A (en) * | 2016-12-07 | 2017-05-17 | 腾讯科技（深圳）有限公司 | Video preview generation method, device, server and system |
| WO2019205603A1 (en) * | 2018-04-26 | 2019-10-31 | 北京大米科技有限公司 | Image fuzziness measurement method and apparatus, computer device and readable storage medium |
| CN110996169A (en) * | 2019-07-12 | 2020-04-10 | 北京达佳互联信息技术有限公司 | Method, device, electronic equipment and computer-readable storage medium for clipping video |
| CN112312199A (en) * | 2019-07-31 | 2021-02-02 | 杭州海康威视数字技术股份有限公司 | Image processing method, device and equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080046406A1 (en) * | 2006-08-15 | 2008-02-21 | Microsoft Corporation | Audio and video thumbnails |
| CN101807198A (en) * | 2010-01-08 | 2010-08-18 | 中国科学院软件研究所 | Video abstraction generating method based on sketch |
| CN101853286A (en) * | 2010-05-20 | 2010-10-06 | 上海全土豆网络科技有限公司 | Intelligent selection method of video thumbnails |
| CN103020947A (en) * | 2011-09-23 | 2013-04-03 | 阿里巴巴集团控股有限公司 | Image quality analysis method and device |
-
2014
- 2014-12-31 CN CN201410852257.XA patent/CN104598921A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080046406A1 (en) * | 2006-08-15 | 2008-02-21 | Microsoft Corporation | Audio and video thumbnails |
| CN101807198A (en) * | 2010-01-08 | 2010-08-18 | 中国科学院软件研究所 | Video abstraction generating method based on sketch |
| CN101853286A (en) * | 2010-05-20 | 2010-10-06 | 上海全土豆网络科技有限公司 | Intelligent selection method of video thumbnails |
| CN103020947A (en) * | 2011-09-23 | 2013-04-03 | 阿里巴巴集团控股有限公司 | Image quality analysis method and device |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106686403A (en) * | 2016-12-07 | 2017-05-17 | 腾讯科技（深圳）有限公司 | Video preview generation method, device, server and system |
| CN106686403B (en) * | 2016-12-07 | 2019-03-08 | 腾讯科技（深圳）有限公司 | A kind of video preview drawing generating method, device, server and system |
| WO2019205603A1 (en) * | 2018-04-26 | 2019-10-31 | 北京大米科技有限公司 | Image fuzziness measurement method and apparatus, computer device and readable storage medium |
| CN110996169A (en) * | 2019-07-12 | 2020-04-10 | 北京达佳互联信息技术有限公司 | Method, device, electronic equipment and computer-readable storage medium for clipping video |
| CN112312199A (en) * | 2019-07-31 | 2021-02-02 | 杭州海康威视数字技术股份有限公司 | Image processing method, device and equipment |
| CN112312199B (en) * | 2019-07-31 | 2022-07-29 | 杭州海康威视数字技术股份有限公司 | Image processing method, device and equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Dev et al. | Categorization of cloud image patches using an improved texton-based approach | |
| CN105118048B (en) | The recognition methods of reproduction certificate picture and device | |
| Sun et al. | Photo assessment based on computational visual attention model | |
| KR100799557B1 (en) | Method for discriminating a obscene video using visual features and apparatus thereof | |
| CN109580656B (en) | Mobile phone light guide plate defect detection method and system based on dynamic weight combination classifier | |
| CN104766076B (en) | A kind of detection method and device of video image character | |
| US8340412B2 (en) | Image processing | |
| CN101833664A (en) | Video image character detecting method based on sparse expression | |
| CN104598921A (en) | Video preview selecting method and device | |
| CN103530638A (en) | Method for matching pedestrians under multiple cameras | |
| JP7078175B2 (en) | Inspection equipment and method | |
| CN108876756A (en) | The measure and device of image similarity | |
| CN112163572A (en) | Method and device for identifying object | |
| CN100583145C (en) | Method for evaluating adjustable dimension fidelity based on content relevant image | |
| CN104202596A (en) | Image color-cast detection method and system applied to intelligent terminal | |
| Li et al. | Identifying photorealistic computer graphics using second-order difference statistics | |
| CN104581379A (en) | Video preview image selecting method and device | |
| CN111833347A (en) | Transmission line damper defect detection method and related device | |
| CN104410867A (en) | Improved video shot detection method | |
| CN112287884B (en) | Examination abnormal behavior detection method and device and computer readable storage medium | |
| CN105959685B (en) | A kind of compression bit rate Forecasting Methodology based on video content and cluster analysis | |
| Narwaria et al. | An objective method for high dynamic range source content selection | |
| CN113255423A (en) | Method and device for extracting color scheme from video | |
| CN114519689A (en) | Image tampering detection method, device, equipment and computer readable storage medium | |
| WO2007072347A2 (en) | System and method for processing video |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20150506 |