CN111669656B

CN111669656B - Method and device for determining wonderful degree of video clip

Info

Publication number: CN111669656B
Application number: CN202010569015.5A
Authority: CN
Inventors: 牛文娟; 阳任科
Original assignee: Beijing QIYI Century Science and Technology Co Ltd
Current assignee: Beijing QIYI Century Science and Technology Co Ltd
Priority date: 2020-06-19
Filing date: 2020-06-19
Publication date: 2022-12-02
Anticipated expiration: 2040-06-19
Also published as: CN111669656A

Abstract

The embodiment of the invention provides a method and a device for determining the wonderful degree of a video clip, which relate to the technical field of data processing, wherein the method comprises the following steps: counting the playing times, fast forward times and backspacing times of each video clip in the video to be processed; and determining the wonderful degree of each video clip according to the playing times, the fast forward times, the backspacing times and the time sequence value of each video clip in the video to be processed in a mode that the playing times and the backspacing times generate positive influence on the wonderful degree of the video clips and the fast forward times generate negative influence on the wonderful degree of the video clips. By applying the scheme provided by the embodiment of the invention to determine the wonderful degree of the video clip, the accuracy of determining the wonderful degree of the obtained video clip can be improved.

Description

Method and device for determining wonderful degree of video clip

Technical Field

The present invention relates to the field of data processing technologies, and in particular, to a method and an apparatus for determining a highlight level of a video clip.

Background

When a user watches a video, the user generally tends to watch a highlight in the video, and for this reason, the video platform often marks the highlight degree for each video clip in the video, so that the user can conveniently watch the highlight in the video.

In the prior art, before marking the highlight degree for each video segment, when determining the highlight degree of each video segment, the highlight degree of each video segment in the video is determined mainly in a manual determination manner. However, the mode of manually determining the wonderful degree of the video segments is easily influenced by subjective factors of workers, and it is difficult to objectively reflect the general opinion of the wonderful degree of the video segments by a large number of users, so that the accuracy of the wonderful degree of each video segment determined is low.

Disclosure of Invention

The embodiment of the invention aims to provide a method and a device for determining the wonderful degree of a video clip so as to improve the accuracy of the wonderful degree of the determined video clip. The specific technical scheme is as follows:

in a first aspect of the present invention, there is provided a method for determining a highlight level of a video segment, the method including:

counting the playing times, fast forward times and backspacing times of each video clip in the video to be processed, wherein the playing times of each video clip are as follows: the total times of playing the video clip by each user playing the video to be processed without fast forward playing or rewinding in the process of playing the video clip, wherein the fast forward times of each video clip are as follows: the total times of fast-forwarding and playing the video clips by each user, and the backspacing times of each video clip are as follows: the total times of playing the video clip by each user back;

and determining the wonderful degree of each video clip according to the playing times, the fast forwarding times, the backspacing times and the time sequence value of each video clip in the video to be processed in a mode that the playing times and the backspacing times have positive influence on the wonderful degree of the video clips and the fast forwarding times have negative influence on the progress degree of the video clips.

In an embodiment of the present invention, the determining the highlight degree of each video segment according to the play times, fast forward times, rewind times of each video segment, and a time sequence value of each video segment in the video to be processed in a manner that the play times and rewind times have a positive influence on the highlight degree of the video segment and the fast forward times have a negative influence on the highlight degree of the video segment includes:

calculating a first adjustment parameter for adjusting the order of the times to be adjusted of each video clip to the same order of magnitude as a preset reference time, wherein the times to be adjusted are as follows: the number of times except the preset reference number of times in the playing number of times, the fast forward number of times and the backspacing number of times is as follows: any one of the farmhouse times, the fast forward times and the rollback times;

calculating a second adjustment parameter for adjusting the fast forward times of each video segment according to the time sequence value, wherein the second adjustment parameter is: the parameter is used for reducing the influence of fast forward playing of the video clips of the video to be processed on the wonderful degree by each user;

and determining the wonderful degree of each video clip according to the playing times, the fast forward times, the backspacing times, the first adjusting parameter and the second adjusting parameter of each video clip in a mode that the playing times and the backspacing times have positive influence on the wonderful degree of the video clips and the fast forward times have negative influence on the wonderful degree of the video clips.

In an embodiment of the present invention, the calculating a first adjustment parameter for adjusting the order of magnitude of the to-be-adjusted times of each video clip to be the same as the preset reference times includes:

calculating a first adjustment parameter for adjusting the number of times to be adjusted of each video clip by the following formula:

wherein, A is _i A first adjustment parameter for adjusting the ith number of times to be adjusted, i represents the type of the number of times to be adjusted, j is the time sequence value of the video clip, n is the total number of the video clips in the video to be processed, b _i,j Represents the ith number of times to be adjusted of the jth video segment, c _j Indicating a preset reference number of times of the jth video segment.

In an embodiment of the present invention, the calculating a second adjustment parameter for adjusting the fast forwarding times of each video segment according to the time sequence value includes:

determining a median of the time sequence values of the video segments of the video to be processed as an intermediate order;

and calculating a second adjusting parameter for adjusting the fast forward times of each video clip according to the intermediate sequence and the time sequence value of each video clip in the video to be processed.

In an embodiment of the present invention, the calculating a second adjustment parameter for adjusting the fast forwarding times of each video segment according to the intermediate order and the time sequence value of each video segment in the video to be processed includes:

calculating a second adjustment parameter for adjusting the number of fast-forwarding times for each video segment according to the following first expression:

wherein t is a time sequence value of each video segment in the video to be processed, f (t) is a second adjustment parameter corresponding to the video segment with the time sequence value of t, μ is the intermediate order, and a, b and σ are preset parameters.

In an embodiment of the present invention, the determining the highlight of each video segment according to the play count, the fast forward count, the rewind count, the first adjustment parameter and the second adjustment parameter of each video segment in a manner that the play count and the rewind count have a positive influence on the highlight of the video segment and the fast forward count has a negative influence on the highlight of the video segment includes:

determining the wonderful degree of each video clip according to the following second expression:

wherein, quality is the wonderful degree of each video clip, back is the backspace frequency, skip is the fast forward frequency, play is the Play frequency, w ₁ A first preset weight, w, corresponding to the rollback number ₂ A second preset weight, w, corresponding to the fast forward times ₃ And setting alpha and beta as the first adjusting parameters and f (t) as the second adjusting parameters corresponding to the video clip with the time sequence value t as the third preset weight corresponding to the playing times.

In an embodiment of the present invention, before determining the highlight of each video segment according to the play count, the fast forward count, the rewind count, and the time sequence value of each video segment in the video to be processed, in a manner that the play count and the rewind count have a positive influence on the highlight of the video segment and the fast forward count has a negative influence on the highlight of the video segment, the method further includes:

determining a first target video clip with incomplete times in each video clip, wherein the incomplete times are as follows: one or more items which are not successfully obtained in the playing times, the fast forwarding times and the backspacing times;

calculating a correction value of the incomplete times according to the times corresponding to the incomplete times in a second target video clip, wherein the second target video clip is as follows: the absolute value of the difference value of the time sequence values between the first target video clip and the first target video clip is minimum, and the video clip of the times corresponding to the incomplete times is successfully obtained;

and correcting the incomplete times into the correction value.

In an embodiment of the present invention, in a manner that when the to-be-processed video includes a video segment belonging to a slice head and/or a video segment belonging to a slice tail, the number of times of playing and the number of times of rewinding have a positive influence on the wonderness of the video segment, and the number of times of fast forwarding has a negative influence on the wonderness of the video segment, after determining the wonderness of each video segment according to the number of times of playing, the number of times of fast forwarding, the number of times of rewinding, and a time sequence value of each video segment in the to-be-processed video, the method further includes:

correcting the wonderful degree of each video clip belonging to the leader according to the wonderful degree of a second preset number of video clips which are closest to the leader in time sequence and do not belong to the leader in the video to be processed;

and/or

And correcting the wonderful degree of each video clip belonging to the leader according to the wonderful degree of a third preset number of video clips which are closest to the leader in the time sequence order and do not belong to the leader in the video to be processed.

In one embodiment of the present invention, the accuracy of the calculated wonderful degree is judged in the following manner:

obtaining the sample playing times, the sample fast forwarding times and the sample backspacing times of each sample video clip in the sample video;

calculating a first sample adjustment parameter for adjusting the times to be adjusted of each sample video clip;

calculating a second sample adjustment parameter for adjusting the sample fast forward times of each sample video clip;

determining the wonderful degree of each sample video clip as the wonderful degree of the sample according to the sample playing times, the sample fast forward times, the sample backspacing times, the first sample adjusting parameter and the second sample adjusting parameter;

for each sample video segment in the sample video, determining whether the sample highlight degree of the sample video segment is greater than or equal to the sample highlight degree of each sample video segment in the first video segment and greater than or equal to the highlight degree of each sample video segment in the second video segment, if so, the sample video segment is a peak video segment, wherein the first video segment is: according to the time sequence, a video clip composed of a first preset number of consecutive sample video clips located before the sample video clip, wherein the second video clip is: according to the time sequence order, a video clip consisting of a second preset number of continuous sample video clips located behind the sample video clip, wherein the first video clip is adjacent to the sample video clip, and the second video clip is adjacent to the sample video clip;

for each sample video, calculating a sample ratio of the number of peak video segments belonging to a sample segment to the number of sample segments, wherein the sample segments are: a sample video segment which is marked as a highlight video segment in the sample video in advance;

and judging the accuracy of the calculated wonderful degree according to the sample proportion of each sample video.

In a second aspect of the present invention, there is also provided an apparatus for determining a highlight level of a video clip, the apparatus including:

the number obtaining module is used for counting the playing number, the fast forward number and the backspacing number of each video clip in the video to be processed, wherein the playing number of each video clip is as follows: the total times of playing the video clip by each user playing the video to be processed without fast forward playing or rewinding in the process of playing the video clip, wherein the fast forward times of each video clip are as follows: the total times of fast-forwarding and playing the video clips by each user, and the backspacing times of each video clip are as follows: the total times of playing the video clip by each user back;

and the degree determining module is used for determining the wonderful degree of each video clip according to the playing times, the fast forward times and the backspacing times of each video clip and the time sequence value of each video clip in the video to be processed in a mode that the playing times and the backspacing times have positive influence on the wonderful degree of the video clip and the fast forward times have negative influence on the wonderful degree of the video clip.

In an embodiment of the present invention, the degree determining module includes:

a first parameter calculating sub-module, configured to calculate a first adjustment parameter for adjusting the order of magnitude of the to-be-adjusted times of each video segment to the same order of magnitude as a preset reference time, where the to-be-adjusted times are: the number of times except the preset reference number of times in the playing number of times, the fast forward number of times and the backoff number of times is as follows: any one of the play times, fast forward times and rollback times;

a second parameter calculation sub-module, configured to calculate, according to the time sequence value, a second adjustment parameter for adjusting the fast forward times of each video segment, where the second adjustment parameter: the parameter is used for reducing the influence of fast forward playing of the video clips of the video to be processed on the wonderful degree by each user;

and the degree determining submodule is used for determining the wonderful degree of each video segment according to the playing times, the fast forward times, the backspacing times, the first adjusting parameter and the second adjusting parameter of each video segment in a mode that the playing times and the backspacing times have positive influence on the wonderful degree of the video segments and the fast forward times have negative influence on the wonderful degree of the video segments.

In an embodiment of the present invention, the first parameter calculating sub-module is specifically configured to:

wherein, A is _i A first adjustment parameter for adjusting the ith number of times to be adjusted, i represents the type of the number of times to be adjusted, j is the time sequence value of the video clip, n is the total number of the video clips in the video to be processed, b _i,j Represents the ith number of times to be adjusted of the jth video clip, c _j Indicating a preset reference number of times of the jth video clip.

In an embodiment of the present invention, the second parameter calculating sub-module includes:

an order determination unit configured to determine a median of timing values of video segments of the video to be processed as an intermediate order;

and the second parameter calculation unit is used for calculating a second adjustment parameter for adjusting the fast forward times of each video clip according to the intermediate sequence and the time sequence value of each video clip in the video to be processed.

In an embodiment of the present invention, the second parameter calculating unit is specifically configured to:

calculating a second adjustment parameter for adjusting the number of fast-forwarding times of each video segment according to the following first expression:

wherein t is a time sequence value of each video clip in the video to be processed, f (t) is a second adjustment parameter corresponding to the video clip with the time sequence value of t, μ is the intermediate order, and a, b and σ are preset parameters.

In an embodiment of the present invention, the degree determining submodule is specifically configured to:

In one embodiment of the present invention, the apparatus further comprises:

a target segment determining module, configured to determine, before the degree determining module determines the highlight degree of each video segment, a first target video segment with the number of the incomplete times in each video segment, where the number of the incomplete times is: one or more items which are not successfully obtained in the playing times, the fast forwarding times and the backspacing times;

the correction value estimation module is used for calculating the correction value of the incomplete times according to the times corresponding to the incomplete times in a second target video clip, wherein the second target video clip is as follows: the absolute value of the difference value of the time sequence values between the first target video clip and the first target video clip is minimum, and the video clip of the times corresponding to the incomplete times is successfully obtained;

and the frequency correction module is used for correcting the incomplete frequency into the correction value.

In one embodiment of the present invention, the apparatus further comprises:

the film head correcting module is used for correcting the wonderful degree of each video clip belonging to the film head according to the wonderful degree of a second preset number of video clips which are closest to the film head in time sequence order and do not belong to the film head in the video to be processed after the wonderful degree of each video clip is determined by the degree determining module;

and/or

And the film trailer correcting module is used for correcting the wonderful degree of each video clip belonging to the film head according to the wonderful degree of a third preset number of video clips which are closest to the film trailer in time sequence order and do not belong to the film trailer in the video to be processed after the wonderful degree of each video clip is determined by the degree determining module.

In an embodiment of the present invention, the apparatus further includes an accuracy determining module, configured to determine an accuracy of the calculated wonderful degree;

the accuracy judging module comprises:

the sample frequency obtaining submodule is used for obtaining the sample playing frequency, the sample fast forward frequency and the sample backspacing frequency of each sample video clip in the sample video;

the first sample parameter calculation submodule is used for calculating a first sample adjustment parameter for adjusting the times to be adjusted of each sample video clip;

the second sample parameter calculation submodule is used for calculating a second sample adjustment parameter for adjusting the sample fast forwarding times of each sample video clip;

the sample degree determining submodule is used for determining the wonderful degree of each sample video clip as the wonderful degree of the sample according to the sample playing times, the sample fast forward times, the sample backspacing times, the first sample adjusting parameter and the second sample adjusting parameter;

the peak fragment determining submodule is configured to determine, for each sample video fragment in a sample video, whether a sample highlight of the sample video fragment is greater than or equal to a sample highlight of each sample video fragment in a first video fragment and greater than or equal to a highlight of each sample video fragment in a second video fragment, and if yes, the sample video fragment is a peak video fragment, where the first video fragment is: according to the time sequence, a video clip composed of a first preset number of consecutive sample video clips located before the sample video clip, wherein the second video clip is: according to the time sequence, a video clip composed of a second preset number of continuous sample video clips is positioned behind the sample video clip, the first video clip is adjacent to the sample video clip, and the second video clip is adjacent to the sample video clip;

a sample proportion calculation submodule, configured to calculate, for each sample video, a sample proportion between the number of peak video segments belonging to a sample segment and the number of sample segments, where a sample segment is: a sample video segment which is marked as a highlight video segment in the sample video in advance;

and the accuracy judgment sub-module is used for judging the accuracy of the wonderful degree obtained by calculation according to the sample proportion of each sample video.

The third aspect of the present invention further provides an electronic device, which includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory complete communication with each other through the communication bus;

a memory for storing a computer program;

a processor configured to implement the method steps of any one of the first aspect when executing a program stored in the memory.

In a further aspect of the present invention, there is also provided a computer-readable storage medium having a computer program stored therein, which when executed by a processor, performs the method steps of any one of the above first aspects.

In a further aspect of the present invention, there is also provided a computer program product comprising instructions which, when run on a computer, cause the computer to perform the method steps of any one of the above first aspects.

The user may generate a plurality of different operations during the playing of the video, including playing the video without fast forwarding or rewinding, fast forwarding, rewinding, and the like. Since a user often fast-forwards playing an unattractive video clip while viewing the video clip, the video clip fast-forwarded by the user is often less wonderful to the user. When a user views an attractive video clip, the user often does not fast forward or rewind the video clip for watching again, so that the video clip which is not fast forward or rewind the video clip is more wonderful for the user. In view of the above situation, when determining the highlight degree of a video segment by applying the scheme provided by the embodiment of the present invention, the highlight degree of each video segment can be determined according to the number of times of playing, the number of times of fast forwarding, and the number of times of rewinding of each video segment in the video to be processed. And because the playing times, the fast forward times and the rollback times are generated in the process of playing the video to be processed by a large number of users, compared with a mode of subjective determination by a worker, the scheme provided by the embodiment of the invention can reflect the common opinions of the large number of users on each video clip, thereby improving the accuracy of the wonderful degree of each determined video clip.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the prior art descriptions will be briefly described below.

Fig. 1 is a flowchart illustrating a method for determining a highlight level of a video segment according to a first embodiment of the present invention;

fig. 2 is a flowchart illustrating a second method for determining the highlight level of a video segment according to an embodiment of the present invention;

fig. 3 is a flowchart illustrating a method for determining a highlight level of a third video segment according to an embodiment of the present invention;

fig. 4A is a flowchart illustrating a fourth method for determining a highlight level of a video segment according to an embodiment of the present invention;

fig. 4B is a flowchart illustrating a method for determining a highlight level of a fifth video segment according to an embodiment of the present invention;

fig. 5 is a schematic flow chart of a method for determining the accuracy of the wonderful degree according to the embodiment of the present invention;

fig. 6 is a schematic structural diagram of a first apparatus for determining the highlight level of a video segment according to an embodiment of the present invention;

fig. 7 is a schematic structural diagram of a second apparatus for determining the highlight level of a video segment according to an embodiment of the present invention;

fig. 8 is a schematic structural diagram of a third apparatus for determining the highlight level of a video segment according to an embodiment of the present invention;

fig. 9 is a schematic structural diagram of a fourth apparatus for determining the highlight level of a video segment according to an embodiment of the present invention;

fig. 10 is a schematic structural diagram of a device for determining a highlight level of a fifth video segment according to an embodiment of the present invention;

fig. 11 is a schematic structural diagram of an accuracy determining module according to an embodiment of the present invention;

fig. 12 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be described below with reference to the drawings in the embodiments of the present invention.

In order to solve the problem that the accuracy of determining the wonderness degree of each obtained video clip is low when the wonderness degree of the video clip is determined in the prior art, embodiments of the present invention provide a method and an apparatus for determining the wonderness degree of the video clip.

In an embodiment of the present invention, a method for determining a highlight level of a video segment is provided, including:

counting the playing times, fast forward times and backspacing times of each video clip in the video to be processed, wherein the playing times of each video clip are as follows: the total times of playing the video clip by each user playing the video to be processed without fast forward playing or rewinding in the process of playing the video clip, wherein the fast forward times of each video clip are as follows: the total number of fast forward playing the video segments by each user is as follows: and each user backs the total times of playing the video clip.

And determining the wonderful degree of each video clip according to the playing times, the fast forward times, the backspacing times and the time sequence value of each video clip in the video to be processed in a mode that the playing times and the backspacing times generate positive influence on the wonderful degree of the video clips and the fast forward times generate negative influence on the wonderful degree of the video clips.

As can be seen from the above, the user may perform a variety of different operations during the playing of the video, including playing the video without fast forwarding or rewinding, fast forwarding, rewinding, and so on. Since a user often fast-forwards playing an unattractive video clip while viewing the video clip, the video clip fast-forwarded by the user is often less wonderful to the user. When a user views an attractive video clip, the user often does not fast forward or rewind the video clip for watching again, so that the video clip which is not fast forward or rewind the video clip is more wonderful for the user. In view of the above situation, when the scheme provided by the embodiment of the present invention is applied to determine the highlight degree of a video clip, the highlight degree of each video clip can be determined according to the number of times of playing, the number of times of fast forwarding, and the number of times of rewinding of each video clip in the video to be processed. And because the playing times, the fast forward times and the backspacing times are generated in the process of playing the videos to be processed by a large number of users, compared with a mode determined by a worker subjectively, the scheme provided by the embodiment of the invention can reflect the common opinions of the large number of users on each video clip, thereby improving the accuracy of the wonderful degree of each video clip determined.

The following describes a method and an apparatus for determining the highlight level of a video segment according to an embodiment of the present invention with specific embodiments.

Referring to fig. 1, a flow chart of a first method for determining the highlight level of a video segment is provided, and specifically, the method includes the following steps S101 to S102.

S101: and counting the playing times, the fast forward times and the backspacing times of each video clip in the video to be processed.

Wherein, the playing times of each video clip are as follows: the total times of playing the video clip without fast forward playing or rewinding in the process of playing the video clip by each user playing the video to be processed are as follows: the total number of fast forward playing the video segments by each user is as follows: and each user backs the total times of playing the video clip.

Specifically, the video to be processed may be divided into video segments with the same duration according to a preset duration, for example, the preset duration may be 1 second, 1 minute, and the like. In addition, the video clips can also be video clips with different preset durations in the video to be processed.

Due to the limitation of the visual characteristics of human eyes, a user cannot easily acquire video contents in a video clip with the time length of less than 1 second in detail, so that the time dimension of 1 second belongs to a time unit with shorter time length in the video playing field, and when the video clip is a clip obtained by dividing a video to be processed by taking 1 second as a unit, the wonderful degree determined by the scheme provided by the embodiment of the invention is the wonderful degree of the video clip with the time length of 1 second, and the wonderful degree is more accurate in the time dimension.

In addition, fast-forward playing a video clip indicates: in the process of playing the video clip, part or all of the content of the video clip is fast-forwarded and skipped by a user, so that the video clip is played quickly. The backspacing playing video clip represents that: in the process of playing the video clip, the user retreats the video clip to the starting position of the video clip or a certain position of the played content for playing again.

In an embodiment of the present invention, the playing times, fast forward times and rollback times of each video segment generated by each user playing the video to be processed within the preset time period may be obtained.

Wherein, the preset time period may be: and the time period from the time when the video to be processed can be played by the user to the time when the playing times, the fast forward times and the backspacing times of each video clip in the video to be processed are requested to be obtained. The preset time period may be: and requesting to obtain the playing times, the fast forward times and the backspacing times of each video clip in the video to be processed, wherein the time is 1 week, 30 days, 1 year and the like before.

S102: and determining the wonderful degree of each video clip according to the playing times, the fast forward times, the backspacing times and the time sequence value of each video clip in the video to be processed in a mode that the playing times and the backspacing times have positive influence on the wonderful degree of the video clips and the fast forward times have negative influence on the wonderful degree of the video clips.

The time sequence value of the video clip in the video to be processed may be: the timing value of the video clip in the aforementioned video to be processed in the forward timing sequence from front to back is, for example, the timing value of the video clip at the front end in the forward timing sequence in the video to be processed is 1, the timing value of the video clip adjacent to the video clip at the front end in the timing sequence is 2, and so on.

The time sequence value of the video clip in the video to be processed may also be: for example, the timing value of the video clip positioned at the rearmost end in the forward timing sequence in the video to be processed is 1, the timing value of the video clip adjacent to the rearmost end in the timing sequence is 2, and so on.

Specifically, for each video clip, the more the playing times of the video clip, the more the times that the video clip is not fast-forwarded and not rewound, that is, in the process of playing the to-be-processed video, the fast-forwarding and rewinding operations are not performed, the more the times that the video clip is completely played in a time sequence order, the higher the wonderful degree of the video clip.

The more the video segment is rewound, the more the user rewinds to the starting position of the video segment or a position of the played content, the more the video segment is rewarded by the user, the higher the highlight of the video segment is, and the highlight of the video segment rewarded by the user is higher than the highlight of the video segment which is not rewarded for fast forward play.

The more fast-forwarding times of the video clip, the more times the video clip is fast-forwarded by the user and not watched, and the less wonderful the video clip.

In an embodiment of the present invention, the number of playing times, the number of backspacing times, and the number of fast forwarding times may be weighted, and a result of the weighted calculation is determined as a highlight degree of each video segment. The higher the number of plays and the number of rewind, the higher the wonderful degree, so the weighted value of the number of plays and the number of rewind can be a positive value, and the higher the number of fast-forwards, the lower the wonderful degree, so the weighted value of the number of rewind can be a negative value.

In addition, in another embodiment of the present invention, the wonderful degree calculation model obtained by training the initial model according to the sample play times, the sample rewind times and the sample fast forward times of the sample video clip may be used to calculate the wonderful degree of the video clip, and the play times, the rewind times and the fast forward times of the video clip may be input into the trained wonderful degree calculation model to obtain the wonderful degree of the video clip.

Meanwhile, the video contents of the video clips in the video to be processed are different, and the attractions of different video contents to the user are different, for example, the plot points of the video contents in the video clip at the front end of the video to be processed may be more, the attractions to the user are higher, the plot points of the video contents in the video clip in the middle of the video to be processed may be less, and the attractions to the user are less. For example, since the head portion and the tail portion of the video to be processed are often the head song and the tail song, and the user habitually fast-forwards plays the head portion and the tail portion of the video to be processed, the user fast-forwards the head portion and the tail portion of the video to be processed no matter whether the head portion and the tail portion are wonderful or not, the wonderful degree cannot be accurately obtained according to the fast-forwarding times of the video clips in the head portion and the tail portion, the fast-forwarding times of the video clips are related to the time sequence value of the video clips in the video to be processed, and the fast-forwarding times of the video clips with different time sequence values can be corrected according to the time sequence value of the video clips in the video to be processed, so that the wonderful degree of the video clips with different time sequence values can be corrected.

In another embodiment of the present invention, the step S102 can be implemented by steps S102A-S102C, which will not be described in detail herein.

In addition, the step of calculating the highlight of the video segment can also be realized by a highlight calculation model.

As can be seen from the above, the user may perform a variety of different operations during the playing of the video, including playing the video without fast forwarding or rewinding, fast forwarding, rewinding, and so on. Since a user often fast-forwards playing an unattractive video clip while viewing the video clip, the video clip fast-forwarded by the user is often less wonderful to the user. When a user views an attractive video clip, the user often does not fast forward or rewind the video clip for watching again, so that the video clip which is not fast forward or rewind or played by the user is often more wonderful for the user. In view of the above situation, when determining the highlight degree of a video segment by applying the scheme provided by the embodiment of the present invention, the highlight degree of each video segment can be determined according to the number of times of playing, the number of times of fast forwarding, and the number of times of rewinding of each video segment in the video to be processed. And because the playing times, the fast forward times and the backspacing times are generated in the process of playing the videos to be processed by a large number of users, compared with a mode determined by a worker subjectively, the scheme provided by the embodiment of the invention can reflect the common opinions of the large number of users on each video clip, thereby improving the accuracy of the wonderful degree of each video clip determined.

Referring to fig. 2, a flow chart of a second method for determining the highlight level of a video segment is provided, and compared with the aforementioned embodiment shown in fig. 1, the above step S102 can be implemented by the following steps S102A to S102C.

S102A: and calculating a first adjusting parameter for adjusting the order of the times to be adjusted of each video clip to the same order of magnitude as the preset reference times.

Wherein, the number of times to be adjusted is as follows: the number of times except the preset reference number of times among the number of times of playing, the number of times of fast forwarding and the number of times of rewinding.

The preset reference times are as follows: any one of the play count, the fast-forward count, and the rewind count.

Because the user does not fast forward or rewind during the process of playing the video, and the difference between the fast forward and rewind playing times is large, the playing times, fast forward times and rewind times of the video segments may be in different orders of magnitude, for example, the order of magnitude of the playing times is hundreds of thousands, the fast forward times is thousands, and the rewind times is thousands, so that if the playing times, the fast forward times and the rewind times are directly used for calculation, the influence degree of the different times on the calculated wonderful degree is large. Therefore, the order of the times to be adjusted in the playing times, the fast forwarding times and the rewinding times can be adjusted through the first adjusting parameter, so that the playing times, the fast forwarding times and the rewinding times are in the same order of magnitude.

When the preset reference times are the playing times, and the times to be adjusted are the fast forward times and the backspacing times, after the first adjustment parameter is obtained through calculation, the fast forward times and the backspacing times can be adjusted according to the first adjustment parameter, so that the magnitude order of the fast forward times and the backspacing times is the same as the magnitude order of the playing times.

Under the condition that the preset reference times are fast forward times, the times to be adjusted are playing times and backspacing times, and after the first adjusting parameter is obtained through calculation, the playing times and the backspacing times can be adjusted according to the first adjusting parameter, so that the order of the playing times and the backspacing times is the same as the order of the fast forward times.

Under the condition that the preset reference times are the backspacing times, the times to be adjusted are the fast forwarding times and the playing times, after the first adjusting parameter is obtained through calculation, the fast forwarding times and the playing times can be adjusted according to the first adjusting parameter, so that the magnitude order of the fast forwarding times and the playing times is the same as that of the backspacing times.

The inventor analyzes a large amount of videos to find that the backspacing times of all video clips are often less for most of videos to be processed, so that compared with the fast forward times or the playing times serving as preset reference times, under the condition that the backspacing times serve as the preset reference times, the backspacing times are less different from the adjusted fast forward times and playing times, the calculated wonderful degree is less influenced by the difference between the orders of magnitude of all times, and the calculated wonderful degree is higher in accuracy.

In an embodiment of the present invention, a total average value of the playing times, a total average value of the fast forwarding times, and a total average value of the backoff times of all video segments in each to-be-processed video may be determined, and a ratio of the total average value of the preset reference times to the total average value of the to-be-adjusted times is used as the first adjustment parameter, so that the first adjustment parameters of each to-be-processed video are the same.

In another embodiment of the present invention, the first adjustment parameter may also be calculated through steps a-B, which will not be described in detail herein.

S102B: and calculating a second adjusting parameter for adjusting the fast forward times of each video clip according to the time sequence value.

Wherein the second adjustment parameter: and the parameter is used for reducing the influence of fast forward playing of the video clips of the video to be processed on the wonderful degree by each user.

In an embodiment of the present invention, during the implementation of the scheme provided by the embodiment of the present invention, the inventor finds that most users practice inertial fast-forward playing of video clips at some time sequence value positions when watching videos, and therefore, statistical analysis may be performed on different videos to be processed, and a target time sequence value of a video clip of which the fast-forward playing frequency of the user is greater than a first time threshold is determined.

For example, in the case that the target sequence value obtained by performing statistical analysis on each to-be-processed video is the middle sequence of the to-be-processed video, the closer the sequence value of the video segment is to the middle sequence, the more the user tends to fast forward the video segment, so the fast forward times can be adjusted by the second adjustment parameter, so that the less the influence of each user on the highlight degree due to fast forward playing of the video segment which tends to the middle of the to-be-processed video is, the less the influence of the subjective factor on the calculated highlight degree is reduced, and the accuracy of the calculated highlight degree is improved.

In an embodiment of the present invention, the number of fast forwarding times of each video segment may be subtracted by a preset number of fast forwarding times, where the higher the time sequence value of the video segment is closer to the target time sequence value, the higher the preset number of fast forwarding times is, the second adjustment parameter may be represented by the preset number of fast forwarding times, for example, if the time sequence value of the video segment 1 is 5, the time sequence value of the video segment 2 is 11, and the target time sequence value is 10, the time sequence value of the video segment 2 is closer to the target time sequence value than the time sequence value of the video segment 1, the second adjustment parameter of the video segment 2 may be 500, the second adjustment parameter of the video segment 1 may be 100, and the like.

In another embodiment of the present invention, the first adjustment parameter for adjusting the number of times to be adjusted of each video segment may be calculated by the following formula:

wherein, A is as described above _i A first adjustment parameter for adjusting the ith number of times to be adjusted, i represents the type of the number of times to be adjusted, j is the time sequence value of the video clip, n is the total number of the video clips in the video to be processed, b _i,j Represents the jth viewThe ith number of times to be adjusted of a frequency segment, c _j Indicating a preset reference number of times of the jth video segment.

Specifically, if the preset reference number is the number of rollback times, and the parameters to be adjusted are the number of fast forwarding times and the number of playing times, then if i =1, the above b is used as the parameter to be adjusted _1, The number of fast-forwards of the jth video segment can be represented, in case of i =2, b above _2, Can indicate the playing times of the jth video clip, c _j Indicating the number of backspacing of the jth video segment.

If the preset reference number is the fast forward number, and the parameters to be adjusted are the back-off number and the play number, if i =1, the above-mentioned b _1, May represent the number of back-offs of the jth video clip, in case of i =2, b above _2, Can indicate the playing times of the jth video clip, c _j Indicating the number of fast-forwards for the jth video segment.

If the preset reference times are the playing times, and the parameters to be adjusted are the fast forward times and the rewind times, then if i =1, the above b _1, The number of fast-forwards of the jth video segment can be represented, in case of i =2, b above _2, May represent the number of backspacing of the jth video segment, c _j Indicating the number of plays of the jth video clip.

And the first adjusting parameters are the same for all video clips in the same video to be processed. Since the number of times to be adjusted includes 2 times, the calculated first adjustment parameter is 2.

As can be seen from the above, in the embodiment of the present invention, after the number of times to be adjusted is multiplied by the first adjustment parameter, the order of magnitude of the number of times to be adjusted is the same as the order of magnitude of the preset reference number of times, so that the influence on the wonderful degree, which is caused by the difference in the order of magnitude between the numbers of times, and further the calculated wonderful degree is eliminated, and therefore, the accuracy of the calculated wonderful degree is higher.

S102C: and determining the wonderful degree of each video clip according to the playing times, the fast forward times, the backspacing times, the first adjusting parameter and the second adjusting parameter of each video clip in a mode that the playing times and the backspacing times generate positive influence on the wonderful degree of the video clips and the fast forward times generate negative influence on the wonderful degree of the video clips.

Specifically, the order of magnitude of the number of times to be adjusted is adjusted through the first adjustment parameter, the number of times of fast forward is adjusted through the second adjustment parameter, and the wonderful degree of each video clip is determined through the adjusted play number, fast forward number and rewind number.

The first adjustment parameter may be multiplied by the number of times to be adjusted, the order of the number of times to be adjusted may be adjusted, the second adjustment parameter may be multiplied by the number of fast forwarding times, the number of fast forwarding times may be adjusted, and in the case where the second adjustment parameter is the preset number of fast forwarding times, the number of fast forwarding times may be subtracted from the second adjustment parameter, and the number of fast forwarding times may be adjusted.

In an embodiment of the present invention, under the condition that the number of backspacing is a preset reference number, and the number of playing times and the number of fast forwarding are to-be-adjusted times, the wonderful degree of each video segment may be determined according to the following second expression:

wherein, quality is the wonderful degree of each video clip, back is the above-mentioned backspace number, skip is the above-mentioned fast forward number, play is the above-mentioned Play number, w ₁ A first preset weight, w, corresponding to the rollback number ₂ A second predetermined weight, w, corresponding to the fast forward times ₃ And a and β are the first adjustment parameters, and f (t) is a second adjustment parameter corresponding to the video segment with the time sequence value t.

Specifically, since the higher the number of backspacing and playing of a video segment is, the more wonderful the video segment is for the user, the higher the wonderful degree of the video segment is, and since the wonderful degree is expressed by a positive value, the first preset weight w ₁ And a third predetermined weight w ₃ Is a positive value. And since the rewind play is better than the play without fast forward play or rewind playIndicating that the video clip is highly exciting to the user, w ₁ Is a value ratio w ₃ Large, e.g. w ₁ Is 2,w ₃ Is 1. Conversely, the higher the number of fast-forwards, the less wonderful the video segment is for the user, and the less wonderful the video segment is, so the second preset weight w ₂ Has a negative value, e.g. w ₂ Is-1. In addition, when the first adjustment parameters α and β in the above formula are calculated, the number of backspacing as a preset reference number, and the number of playing times and the number of fast forwarding as a number to be adjusted, so that the number of fast forwarding can be adjusted by multiplying the first adjustment parameter α by the number of fast forwarding, and the number of playing times can be adjusted by multiplying the first adjustment parameter β by the number of playing times.

In another embodiment of the present invention, when the number of times of playing is a preset reference number, and the number of times of rewinding and the number of times of fast forwarding are to-be-adjusted times, the second expression is:

compared with the situation that the backspacing times are used as the preset reference times, and the playing times and the fast forwarding times are used as the times to be adjusted, the backspacing times are adjusted by multiplying the first adjusting parameter alpha, and the fast forwarding times are adjusted by multiplying the first adjusting parameter beta and the fast forwarding times.

In another embodiment of the present invention, when the number of fast forwarding times is a preset reference number, and the number of backoff times and the number of play times are to-be-adjusted times, the second expression is:

compared with the situation that the backspacing times are used as the preset reference times, and the playing times and the fast forwarding times are used as the times to be adjusted, the backspacing times are adjusted by multiplying the first adjusting parameter alpha by the backspacing times, and the playing times are adjusted by multiplying the first adjusting parameter beta by the playing times.

As can be seen from the above, each number of times is adjusted by the first adjustment parameter and the second adjustment parameter obtained by calculation, and the wonderful degree of the video is calculated according to the adjusted play number, fast forward number and rewind number. The order of magnitude of the times to be adjusted is adjusted according to the first adjustment parameter, and the order of magnitude of the times to be adjusted is the same as the preset reference times, so that the influence on the wonderful degree caused by different orders of magnitude and further on the wonderful degree obtained by calculation is eliminated. And because the fast forward times are adjusted according to the second adjustment parameter, the fast forward times at the time sequence value of the habitual fast forward playing of the user are reduced, the influence of the habitual fast forward playing of the user on the fast forward times is eliminated, and the influence of the habitual fast forward playing of the user on the calculated wonderful degree is further eliminated, so that the accuracy of the calculated wonderful degree is higher.

In one embodiment of the present invention, the step S102B can be realized by the following steps a to B.

Step A: and determining the median of the time sequence values of the video clips of the video to be processed as a middle sequence.

Specifically, the number of the median values is different according to the number of video segments in the video to be processed, where the number of the video segments is an odd number, the median value is 1, and where the number of the video segments is an even number, the median value is 2. For example, if the timing values of the respective video segments are 1 to 9, the median value is 5, and the median order is 5, and if the timing values of the respective video segments are 1 to 10, the median values are 5 and 6, and the median order is 5 and 6.

In another embodiment of the present invention, when a video to be processed includes a video segment belonging to a slice header, if a middle order is determined according to timing sequence values of all video segments in the video to be processed, where the middle order is a middle order of the video segment belonging to the slice header, but a user tends to fast forward and play the video segment belonging to the slice header in the video to be processed, the fast forward times of the video segment belonging to the slice header are higher, and the fast forward times of the video segment belonging to the slice header do not have an objective meaning, each second adjustment parameter is calculated according to the middle order in the case of including the video segment belonging to the slice header, so that the calculated second adjustment parameter is affected by the video segment belonging to the slice header, and the calculated second adjustment parameter is inaccurate. The above-mentioned intermediate order can be determined with the timing values of the video segments belonging to the parts other than the slice header in the video to be processed.

Under the condition that the video to be processed contains the video clips belonging to the trailer, if the middle sequence is determined according to the time sequence values of all the video clips in the video to be processed, the middle sequence is the middle sequence containing the video clips belonging to the trailer, but because a user often fast-forwards plays the video clips belonging to the trailer in the video to be processed, the fast-forwarding times of the video clips belonging to the trailer are higher, and the fast-forwarding times of the video clips belonging to the trailer have no objective significance, each second adjustment parameter is calculated according to the middle sequence under the condition that the video clips belonging to the trailer are contained, the calculated second adjustment parameter is influenced by the video clips belonging to the trailer, and the calculated second adjustment parameter is inaccurate. Therefore, the above-described intermediate order may be determined by the timing values of video segments belonging to portions other than the end of the slice in the video to be processed.

And B, step B: and calculating a second adjusting parameter for adjusting the fast forward times of each video clip according to the intermediate sequence and the time sequence value of each video clip in the video to be processed.

Specifically, according to the statistical results of the inventor, the closer the video segment is to the middle of the video to be processed in time sequence, the higher the probability that the user habitually fast forwards the video, and therefore the second adjustment parameter is needed to adjust the fast forwarding times. The smaller the absolute value of the difference between the time sequence value and the middle sequence of the video segment is, the closer the video segment is to the middle of the video to be processed, so that the distance relationship between the video segment and the video to be processed can be determined according to the time sequence value and the middle sequence of the video segment, and the second adjustment parameter for adjusting the fast forward times can be calculated according to the distance relationship, so that the second adjustment parameter can adjust the fast forward times of the video segments with different distance relationships with the middle of the video to be processed to different degrees, and the influence of fast forward play of the video segments which tend to be in the middle of the video to be processed by each user on the wonderful degree is weaker.

Specifically, the second adjustment parameter for adjusting the fast forward times of the video segments may be calculated according to the following first expression:

wherein t is a time sequence value of each video clip in the video to be processed, f (t) is a second adjustment parameter corresponding to the video clip with the time sequence value of t, μ is the intermediate order, and a, b, and σ are preset parameters.

Specifically, a may be 200, b may be 200

Then this time

For a normal distribution function, σ may be 300.

As can be seen from the above, in the embodiment of the present invention, the second adjustment parameter of each video segment is calculated according to the time sequence value of each video segment in the video to be processed, so that the fast forward times of each video segment can be adjusted according to the second adjustment parameter, the fast forward times at the time sequence value position of the habitual fast forward play of the user are reduced, the influence on the fast forward times due to the habitual fast forward play of the user is eliminated, and the influence on the calculated wonderful degree due to the habitual fast forward play of the user is further eliminated, so that the accuracy of the calculated wonderful degree is higher.

In the process of obtaining the playing times, the fast forward times and the rewind times of each video clip, there may be a case that the playing times, the fast forward times or the rewind times are failed to be obtained, so before determining the wonderful degree of each video clip according to the playing times, the fast forward times and the rewind times, the times of failed obtaining need to be corrected.

Referring to fig. 3, a flowchart of a third method for determining the highlight level of a video segment according to an embodiment of the present invention is provided, and compared with the foregoing fig. 1, the method further includes the following steps S103 to S105 before step S102.

S103: and determining a first target video segment with the number of the incomplete times in each video segment.

Wherein the defect times are as follows: one or more of the number of plays, the number of fast-forwards, and the number of backs that have not been successfully obtained.

S104: and calculating the correction value of the incomplete times according to the times corresponding to the incomplete times in the second target video clip.

Wherein the second target video clip is: and a video clip having the smallest absolute value of the difference between the time sequence values of the first target video clip and the number of times corresponding to the number of the incomplete times is successfully obtained.

Specifically, the average value of the number of times corresponding to the number of the deletions in the second target video segment located at the front end of the first target video segment in the video to be processed and the number of times corresponding to the number of the deletions in the second target video segment located at the rear end of the first target video segment may be calculated as the correction value in time sequence order.

In addition, the number of times corresponding to the number of times of the cut-outs in the second target video segment located at the front end of the first target video segment or the number of times corresponding to the number of times of the cut-outs in the second target video segment located at the rear end of the first target video segment in time series order may be used as the correction value.

The defect number is the same as the number corresponding to the defect number, for example, if the defect number is the playing number, the number corresponding to the defect number is also the playing number.

S105: and correcting the incomplete times into the correction value.

Therefore, the method can correct the incomplete times in the first target video clip, and the highlight degree of the first target video clip can be calculated by using the corrected incomplete times, so that the accuracy of the highlight degree of the first target video clip with the incomplete times is improved.

Referring to fig. 4A, an embodiment of the present invention provides a flowchart of a fourth method for determining a highlight level of a video clip, where in the case that the to-be-processed video includes a video clip belonging to a slice header and/or a video clip belonging to a slice trailer, compared with the foregoing fig. 1, after the foregoing step S102, the method further includes the following step S106:

s106: and correcting the highlight degree of each video clip belonging to the leader according to the highlight degrees of a second preset number of video clips which are closest to the leader in time sequence order and do not belong to the leader in the video to be processed, and/or correcting the highlight degree of each video clip belonging to the leader according to the highlight degrees of a third preset number of video clips which are closest to the leader in time sequence order and do not belong to the leader in the video to be processed.

Specifically, an average value of the degrees of highlights of a second preset number of video segments not belonging to the leader, which are closest to the leader in time sequence order in the video to be processed, may be calculated, and the degrees of highlights of the video segments belonging to the leader may be corrected to the average value.

The highlight of each video segment belonging to the title can be corrected by gradually increasing the highlight of each video segment belonging to the title from 0 to the average value in the time sequence from front to back.

Other ways of correcting the highlights of the individual video segments belonging to the slice header can also be used.

Similarly, an average value of the highlights of a third preset number of video segments not belonging to the trailer, which are closest to the leader in time sequence order, in the video to be processed may be calculated, and the highlights of each video segment belonging to the trailer may be corrected to the average value.

The highlight of each video segment belonging to the trailer can also be corrected by sequentially narrowing down the highlight of each video segment belonging to the trailer from the average value to 0 from the front to the back in the time sequence.

Other ways of correcting the highlights of the individual video segments belonging to the trailer can also be used.

As can be seen from the above description, in the case that the video includes a video segment belonging to the head and/or a video segment belonging to the trailer, the video may start playing from the video segment not belonging to the head, which is closest to the head in time sequence, to the video segment not belonging to the trailer, which is closest to the head in time sequence, so that the data of the playing times, fast forwarding times, and rewinding times in the video segment belonging to the head and the video segment belonging to the trailer are less, and the wonderness of the video segments belonging to the head and the trailer cannot be accurately calculated, and thus the wonderness of the video segments belonging to the head and the trailer can be corrected by the above method.

Referring to fig. 4B, a flowchart of a method for determining the highlight level of a fifth video segment is provided.

As can be seen from fig. 4B, the video clip of each second in the video to be processed is taken as a unit for determining the highlight degree, the playing times, the fast forward times, and the rewind times of the video clip of each second are obtained, the incomplete times are obtained without success in correction, the highlight degree of each video clip is calculated according to the playing times, the fast forward times, and the rewind times of the video clip of each second, whether the video to be processed includes the title and the trailer is judged, if so, the highlight degree of the video clip in the title and the trailer is corrected, the highlight degree of each video clip in the video to be processed after correction is determined as the highlight degree of the video clip, otherwise, the highlight degree of the video clip obtained by calculation is directly determined as the highlight degree of the video clip.

Referring to fig. 5, an embodiment of the present invention provides a flowchart of a method for determining the accuracy of a highlight, and specifically, the method includes the following steps S501 to S508.

S501: and obtaining the sample playing times, the sample fast forwarding times and the sample backspacing times of each sample video clip in the sample video.

Specifically, the sample video includes a sample segment, and the sample segment is: video segments of the sample video that are pre-marked as highlight video segments.

The sample fragment may be staff marked.

S502: and calculating a first sample adjustment parameter for adjusting the times to be adjusted of each sample video clip.

S503: and calculating a second sample adjusting parameter for adjusting the sample fast forwarding times of each sample video segment.

S504: and determining the wonderful degree of each sample video clip as the wonderful degree of the sample according to the sample playing times, the sample fast forward times, the sample backspacing times, the first sample adjusting parameter and the second sample adjusting parameter.

Specifically, the embodiments of the steps S501 to S504 are similar to the steps S101 to S102C in the embodiment of fig. 2, and are not repeated herein.

S505: and determining whether the sample highlight degree of the sample video clip is greater than or equal to the sample highlight degree of each sample video clip in the first video clip and greater than or equal to the highlight degree of each sample video clip in the second video clip aiming at each sample video clip in the sample video, wherein if so, the sample video clip is the peak video clip.

Wherein, the first video clip is: and according to the time sequence, the video clip which is positioned before the sample video clip and consists of a first preset number of continuous sample video clips.

The second video segment is: and according to the time sequence order, the video clip consisting of a second preset number of continuous sample video clips positioned after the sample video clip.

The first video segment is adjacent to the sample video segment, and the second video segment is adjacent to the sample video segment.

The sum of the first preset number and the second preset number may be the number of video clips included in the sample segment with the largest number of sample video clips included in the sample video, so that at most one peak video clip may be included in the same sample segment.

Specifically, the first preset number may be the same as the second preset number, and the sample video clip is located between the first video clip and the second video clip, and the first preset number and the second preset number may also be different.

If a plurality of consecutive peak video clips exist, it is determined that the samples of the peak video clips have the same highlight level, and the top peak video clip of the plurality of consecutive peak video clips represents the plurality of consecutive peak video clips.

S506: for each sample video, a sample ratio of the number of peak video segments belonging to a sample segment to the number of sample segments is calculated.

Specifically, the sample video segment is the sample video segment which is marked as the highlight video segment in the sample video in advance, so that the sample highlight degree of the sample video segment in the sample video segment is higher, and therefore if the peak video segment belongs to the sample segment, it is described that the calculated sample highlight degree conforms to the rule that the highlight degree of the sample video segment in the sample segment is higher, and it is described that the accuracy of the calculated sample highlight degree of each sample video segment in the sample video is higher.

In an embodiment of the present invention, the sample segment may be a video segment with high wonderful degree and marked manually. The number of peak video segments belonging to a sample segment may be counted, and a ratio of the counted number to the number of sample segments may be calculated as a sample ratio.

For example, if the number of peak video segments belonging to a sample segment is 3 and the number of sample segments is 10, the sample ratio is 0.3.

In another embodiment of the present invention, although the peak video segment is the sample video segment with the highest sample highlight degree among the first preset number of consecutive sample video segments, if the sample highlight degrees of the first preset number of consecutive sample video segments are all lower, the sample highlight degree of the peak video segment is lower, and therefore even if the peak video segment belongs to the sample segments, the peak video segment does not conform to the characteristic that the sample highlight degree of the sample video segment in the sample segments is higher.

Moreover, the sample highlights of the first preset number of consecutive sample video clips may all be the same value, and the first preset number of consecutive sample video clips may all be regarded as peak video clips, and the result obtained by the above calculation does not meet the actual situation.

Thus, a sample ratio of the number of peak video segments belonging to the target sample segment to the number of sample segments can be calculated. Wherein, the target sample fragment is: and the difference value between the highest value and the lowest value of the sample highlights of the sample video segment is larger than the preset difference value.

Because the difference between the highest value and the lowest value of the sample highlights of the sample video clips included in the target sample clip is greater than the preset difference, it is indicated that the sample highlights of the peak video clip belonging to the target sample clip are greater than the preset difference, and the sample highlights of each sample video clip in the target sample clip are different, so that the result of judging the accuracy of the highlights determined by the second expression is more accurate.

S507: and judging the accuracy of the calculated wonderful degree according to the sample proportion of each sample video.

Specifically, an average value of the respective sample ratios may be calculated, and a higher average value indicates a higher accuracy of the degree of wonderness determined according to the above-described second expression. Therefore, a preset average value may be set, and if the average value is higher than the preset average value, the wonderness determined according to the second expression is considered to be accurate.

In addition, in another embodiment of the present invention, in the case of calculating the highlight degree of the video clip by using the highlight degree calculation model, the sample play times, the sample fast-forward times and the sample rollback times of the sample video clip may be input into the highlight degree calculation model, and the loss value of the highlight degree calculation model may be calculated, so as to obtain the accuracy of determining the calculated highlight degree.

As can be seen from the above, the accuracy of the wonderful degree obtained by the second expression is determined by using the sample video through the accuracy determination method, and further, the preset parameters in the first expression and the preset weights in the second expression can be adjusted according to the determined accuracy, so that the accuracy of the wonderful degree calculated by the second expression is improved.

Corresponding to the method for determining the highlight degree of the video clip, the embodiment of the invention also provides a device for determining the highlight degree of the video clip.

Referring to fig. 6, an embodiment of the present invention provides a schematic structural diagram of a first apparatus for determining a highlight level of a video clip, and specifically, the apparatus includes:

the number obtaining module 601 is configured to count the playing number, the fast forwarding number, and the backspacing number of each video segment in the video to be processed, where the playing number of each video segment is: the total times of playing the video clip without fast forward playing or rewinding in the process of playing the video clip by each user who plays the video to be processed are as follows: the total times of fast-forwarding and playing the video clips by each user, and the backspacing times of each video clip are as follows: the total times of playing the video clip by each user back;

the degree determining module 602 is configured to determine the wonderful degree of each video segment according to the playing times, the fast forwarding times, the backspacing times of each video segment, and the time sequence value of each video segment in the video to be processed, in a manner that the playing times and the backspacing times have a positive influence on the wonderful degree of the video segment and the fast forwarding times have a negative influence on the wonderful degree of the video segment.

As can be seen from the above, the user may perform a variety of different operations during the playing of the video, including playing the video without fast forwarding or rewinding, fast forwarding, rewinding, and so on. Since a user often fast-forwards playing an unattractive video clip while viewing the video clip, the video clip fast-forwarded by the user is often less wonderful to the user. When a user views an attractive video clip, the user often does not fast forward or rewind the video clip for watching again, so that the video clip which is not fast forward or rewind or played by the user is often more wonderful for the user. In view of the above situation, when determining the highlight degree of a video segment by applying the scheme provided by the embodiment of the present invention, the highlight degree of each video segment can be determined according to the number of times of playing, the number of times of fast forwarding, and the number of times of rewinding of each video segment in the video to be processed. And because the playing times, the fast forward times and the rollback times are generated in the process of playing the video to be processed by a large number of users, compared with a mode of subjective determination by a worker, the scheme provided by the embodiment of the invention can reflect the common opinions of the large number of users on each video clip, thereby improving the accuracy of the wonderful degree of each determined video clip.

Referring to fig. 7, an embodiment of the present invention provides a schematic structural diagram of a second apparatus for determining a highlight level of a video segment, where, compared with the embodiment shown in fig. 6, the above-mentioned level determining module 602 includes:

a first parameter calculating sub-module 602A, configured to calculate a first adjusting parameter for adjusting the order of the to-be-adjusted times of each video segment to be the same as a preset reference time, where the to-be-adjusted times are: the number of times except the preset reference number of times in the playing number of times, the fast forward number of times and the backoff number of times is as follows: any one of the play times, fast forward times and rollback times;

a second parameter calculating sub-module 602B, configured to calculate, according to the time sequence value, a second adjustment parameter for adjusting the fast forwarding times of each video segment, where the second adjustment parameter is: the parameter is used for reducing the influence of fast forward playing of the video clips of the video to be processed on the wonderful degree by each user;

the degree determining submodule 602C is configured to determine the degree of wonderness of each video segment according to the number of plays, the number of fast-forwarding times, the number of rewinding times, the first adjustment parameter, and the second adjustment parameter of each video segment in a manner that the number of plays and the number of rewinding times have a positive influence on the degree of wonderness of the video segment and the number of fast-forwarding times has a negative influence on the degree of wonderness of the video segment.

As can be seen from the above, each number of times is adjusted by the first adjustment parameter and the second adjustment parameter obtained by calculation, and the wonderful degree of the video is calculated according to the adjusted play number, fast forward number, and rewind number. The order of magnitude of the times to be adjusted is adjusted according to the first adjustment parameter, and the order of magnitude of the times to be adjusted is the same as the preset reference times, so that the influence on the wonderful degree caused by different orders of magnitude and further on the wonderful degree obtained by calculation is eliminated. And because the fast forward times are adjusted according to the second adjustment parameter, the fast forward times at the time sequence value of the habitual fast forward playing of the user are reduced, the influence of the habitual fast forward playing of the user on the fast forward times is eliminated, and the influence of the habitual fast forward playing of the user on the calculated wonderful degree is further eliminated, so that the accuracy of the calculated wonderful degree is higher.

In an embodiment of the present invention, the first parameter calculating sub-module 602A is specifically configured to:

wherein, A is _i A first adjustment parameter for adjusting the ith times to be adjusted, i represents the type of the times to be adjusted, j is the time sequence value of the video clip, n is the total number of the video clips in the video to be processed, b _i,j Represents the ith number of times to be adjusted of the jth video segment, c _j Indicating a preset reference number of times of the jth video clip.

Referring to fig. 8, an embodiment of the present invention provides a schematic structural diagram of a third apparatus for determining a highlight level of a video segment, and compared with the foregoing embodiment shown in fig. 7, the second parameter calculating sub-module 602B includes:

an order determining unit 602B1, configured to determine a median of the timing values of the video segments of the video to be processed as an intermediate order;

a second parameter calculating unit 602B2, configured to calculate a second adjustment parameter for adjusting the fast forwarding times of each video segment according to the intermediate order and a time sequence value of each video segment in the video to be processed.

In an embodiment of the invention, the second parameter calculating unit 602B2 is specifically configured to:

As can be seen from the above, in the embodiment of the present invention, the second adjustment parameter of each video segment is calculated according to the time sequence value of each video segment in the video to be processed, so that the fast forward times of each video segment can be adjusted according to the second adjustment parameter, the fast forward times at the time sequence value position of the habitual fast forward play of the user are reduced, the influence on the fast forward times due to the habitual fast forward play of the user is eliminated, the influence on the calculated wonderful degree due to the habitual fast forward play of the user is further eliminated, and therefore, the accuracy of the calculated wonderful degree is higher.

In an embodiment of the present invention, the degree determining sub-module 602C is specifically configured to:

wherein, quality is the wonderful degree of each video clip, back is the backspace time, skip is the fast forward time, play is the Play time, w ₁ A first preset weight, w, corresponding to the rollback number ₂ A second preset weight, w, corresponding to the fast forward times ₃ And setting alpha and beta as the first adjusting parameters and f (t) as the second adjusting parameters corresponding to the video clip with the time sequence value t as the third preset weight corresponding to the playing times.

Referring to fig. 9, a schematic structural diagram of a fourth apparatus for determining the highlight level of a video segment according to an embodiment of the present invention is provided, and compared with the foregoing embodiment shown in fig. 6, the apparatus further includes:

a target segment determining module 603, configured to determine, before the degree determining module 602 determines the highlight degree of each video segment, a first target video segment with a number of stutters in each video segment, where the number of stutters is: one or more items which are not successfully obtained in the playing times, the fast forwarding times and the backspacing times;

a correction value estimation module 604, configured to calculate a correction value of the number of the incomplete times according to the number of times corresponding to the number of the incomplete times in a second target video segment, where the second target video segment is: the absolute value of the difference value of the time sequence values between the first target video clip and the first target video clip is minimum, and the video clip of the times corresponding to the incomplete times is successfully obtained;

a number correction module 605, configured to correct the incomplete number into the correction value.

Referring to fig. 10, an embodiment of the present invention provides a schematic structural diagram of a fifth apparatus for determining a highlight level of a video segment, and compared with the foregoing embodiment shown in fig. 6, the apparatus further includes:

a head of film correcting module 606, configured to, after the degree determining module 602 determines the highlight degree of each video segment, correct the highlight degree of each video segment belonging to the head of film according to the highlight degrees of a second preset number of video segments not belonging to the head of film, in the video to be processed, where the second preset number is closest to the head of film in time sequence;

and/or

A trailer correcting module 607, configured to, after the degree determining module 602 determines the highlight degree of each video segment, correct the highlight degree of each video segment belonging to the head according to the highlight degrees of a third preset number of video segments not belonging to the trailer, in the video to be processed, where the third preset number is closest to the trailer in time sequence.

As can be seen from the above description, in the case that the video includes a video segment belonging to the head and/or a video segment belonging to the trailer, the video may start playing from the video segment not belonging to the head, which is closest to the head in time sequence, to the video segment not belonging to the trailer, which is closest to the trailer in time sequence, so that the data of the playing times, fast forwarding times, and rewinding times in the video segment belonging to the head and the video segment belonging to the trailer are less, and the wonderness of the video segments belonging to the head and the trailer cannot be accurately calculated, and thus the wonderness of the video segments belonging to the head and the trailer can be corrected by the above method.

Corresponding to the embodiment of the method shown in fig. 5, referring to fig. 11, an embodiment of the present invention provides a schematic structural diagram of an accuracy determining module, configured to determine the accuracy of the wonderful degree calculated according to the second expression, specifically, the accuracy determining module 1101 includes:

the sample number obtaining sub-module 1101A is configured to obtain a sample playing number, a sample fast forwarding number, and a sample returning number of each sample video segment in the sample video;

the first sample parameter calculation sub-module 1101B is configured to calculate a first sample adjustment parameter for adjusting the number of times to be adjusted of each sample video clip;

the second sample parameter calculation sub-module 1101C is configured to calculate a second sample adjustment parameter for adjusting the sample fast forwarding frequency of each sample video segment;

the sample degree determining submodule 1101D is configured to determine, according to the sample playing times, the sample fast forwarding times, the sample backspacing times, the first sample adjustment parameter, and the second sample adjustment parameter, a highlight degree of each sample video clip as a sample highlight degree;

the peak fragment determining sub-module 1101E is configured to determine, for each sample video fragment in the sample video, whether a sample highlight degree of the sample video fragment is greater than or equal to a sample highlight degree of each sample video fragment in a first video fragment and greater than or equal to a highlight degree of each sample video fragment in a second video fragment, and if so, the sample video fragment is the peak video fragment, where the first video fragment is: according to the time sequence, a video clip composed of a first preset number of consecutive sample video clips located before the sample video clip, wherein the second video clip is: according to the time sequence order, a video clip consisting of a second preset number of continuous sample video clips located behind the sample video clip, wherein the first video clip is adjacent to the sample video clip, and the second video clip is adjacent to the sample video clip;

the sample ratio calculating submodule 1101F is configured to calculate, for each sample video, a sample ratio between the number of peak video segments belonging to a sample segment and the number of sample segments, where the sample segment is: a sample video segment which is marked as a highlight video segment in the sample video in advance;

the accuracy determining sub-module 1101G is configured to determine the accuracy of the calculated highlight degree according to the sample ratio of each sample video.

An embodiment of the present invention further provides an electronic device, as shown in fig. 12, including a processor 1201, a communication interface 1202, a memory 1203, and a communication bus 1204, where the processor 1201, the communication interface 1202, and the memory 1203 complete mutual communication through the communication bus 1204,

a memory 1203 for storing a computer program;

the processor 1201 is configured to implement the method steps described in the embodiment of the method for determining the highlight level of any video clip when executing the program stored in the memory 1203.

When the embodiment provided by the embodiment of the invention is applied to determining the wonderful degree of the video clip, a user can generate a plurality of different operations in the process of playing the video, including fast forward playing and non-rewind playing, fast forward playing, rewind playing and the like in the process of playing the video clip. Since a user often fast-forwards playing an unattractive video clip while viewing the video clip, the video clip fast-forwarded by the user is often less wonderful to the user. When a user views an attractive video clip, the user often plays the attractive video clip without fast forwarding or rewinding, or plays the attractive video clip with rewinding, so that the user often has a high wonderful degree for the user to play the attractive video clip without fast forwarding or rewinding. In view of the above situation, when determining the highlight degree of a video segment by applying the scheme provided by the embodiment of the present invention, the highlight degree of each video segment can be determined according to the number of times of playing, the number of times of fast forwarding, and the number of times of rewinding of each video segment in the video to be processed. Moreover, because the playing times, the fast forward times and the backspacing times are generated in the process of playing the videos to be processed by a large number of users, compared with a mode determined by a staff subjectively, the scheme provided by the embodiment can reflect the general opinions of the large number of users on each video clip, so that the accuracy of determining the wonderful degree of each obtained video clip is improved.

The communication bus mentioned in the above terminal may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The communication bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is shown, but this is not intended to represent only one bus or type of bus.

The communication interface is used for communication between the terminal and other devices.

The Memory may include a Random Access Memory (RAM) or a non-volatile Memory (non-volatile Memory), such as at least one disk Memory. Optionally, the memory may also be at least one memory device located remotely from the processor.

The Processor may be a general-purpose Processor, and includes a Central Processing Unit (CPU), a Network Processor (NP), and the like; the Integrated Circuit may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other Programmable logic device, a discrete Gate or transistor logic device, or a discrete hardware component.

In a further embodiment of the present invention, a computer-readable storage medium is further provided, in which a computer program is stored, which, when being executed by a processor, implements the method steps of any of the above-mentioned method embodiments for determining the highlight level of a video segment.

When determining the wonderful degree of the video clip by using the computer program stored in the computer-readable storage medium provided by the embodiment, a user may generate a plurality of different operations during playing a video, including not fast-forwarding or not rewinding playing, fast-forwarding or rewinding playing, etc. during playing the video clip. Since a user often fast-forwards playing an unattractive video clip while viewing the video clip, the video clip fast-forwarded by the user is often less wonderful to the user. When a user views an attractive video clip, the user often does not fast forward or rewind the video clip for watching again, so that the video clip which is not fast forward or rewind or played by the user is often more wonderful for the user. In view of the above situation, when the scheme provided by the embodiment of the present invention is applied to determine the highlight degree of a video clip, the highlight degree of each video clip can be determined according to the number of times of playing, the number of times of fast forwarding, and the number of times of rewinding of each video clip in the video to be processed. Moreover, because the playing times, the fast forward times and the backspacing times are generated in the process of playing the videos to be processed by a large number of users, compared with a mode determined by a staff subjectively, the scheme provided by the embodiment can reflect the general opinions of the large number of users on each video clip, so that the accuracy of determining the wonderful degree of each obtained video clip is improved.

In a further embodiment of the present invention, there is also provided a computer program product comprising instructions which, when run on a computer, cause the computer to perform the method steps of any of the above-described method embodiments of the method of highlight determination of a video segment.

When the computer program product provided by the embodiment is executed to determine the wonderful degree of the video segment, a user may generate a plurality of different operations in the process of playing the video, including not fast-forwarding or not rewinding to play, fast-forwarding to play, rewinding to play, and the like in the process of playing the video segment. Since a user often fast-forwards playing an unattractive video clip while viewing it, a video clip that is fast-forwarded by a user is often less exciting to the user. When a user views an attractive video clip, the user often does not fast forward or rewind the video clip for watching again, so that the video clip which is not fast forward or rewind or played by the user is often more wonderful for the user. In view of the above situation, when the scheme provided by the embodiment of the present invention is applied to determine the highlight degree of a video clip, the highlight degree of each video clip can be determined according to the number of times of playing, the number of times of fast forwarding, and the number of times of rewinding of each video clip in the video to be processed. Moreover, because the playing times, the fast forward times and the backspacing times are generated in the process of playing the videos to be processed by a large number of users, compared with a mode determined by a staff subjectively, the scheme provided by the embodiment can reflect the general opinions of the large number of users on each video clip, so that the accuracy of determining the wonderful degree of each obtained video clip is improved.

In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When implemented in software, it may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a computer, cause the processes or functions described in accordance with the embodiments of the invention to be performed in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that incorporates one or more of the available media. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid State Disk (SSD)), among others.

It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrases "comprising one of 8230; \8230;" 8230; "does not exclude the presence of additional like elements in a process, method, article, or apparatus that comprises the element.

All the embodiments in the present specification are described in a related manner, and the same and similar parts among the embodiments may be referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the apparatus, the electronic device, the computer-readable storage medium and the computer program product, since they are substantially similar to the method embodiments, the description is relatively simple, and in relation to what is described in the partial description of the method embodiments.

The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention shall fall within the protection scope of the present invention.

Claims

1. A method for determining the highlight of a video segment, the method comprising:

counting the playing times, fast forward times and backspacing times of each video clip in the video to be processed, wherein the playing times of each video clip are as follows: the total times of playing the video clip by each user playing the video to be processed without fast forward playing or rewinding in the process of playing the video clip, wherein the fast forward times of each video clip are as follows: the total number of fast-forward playing of the video clips by each user is as follows: the total times of playing the video clip by each user back;

determining the wonderful degree of each video clip according to the playing times, the fast forwarding times, the backspacing times and the time sequence value of each video clip in the video to be processed in a mode that the playing times and the backspacing times have positive influence on the wonderful degree of the video clips and the fast forwarding times have negative influence on the wonderful degree of the video clips; wherein, the time sequence value of each video clip in the video to be processed is used for correcting the fast forward times of each video clip;

the positive influence of the playing times and the backspacing times on the wonderful degree of the video clip is as follows: the more the playing times are, the higher the wonderful degree of the video clip is, and the more the backspacing times are, the higher the wonderful degree of the video clip is; the negative influence of the fast forward times on the wonderful degree of the video clip is as follows: the more fast forward times, the lower the wonderful degree of the video clip;

the step of correcting the fast forward times of each video clip by using the time sequence value of each video clip in the video to be processed specifically comprises the following steps: for each video clip, determining a preset fast forward frequency for the video clip according to a time sequence value of the video clip in the video to be processed, wherein the closer the time sequence value is to a target time sequence value, the larger the numerical value of the preset fast forward frequency is, and the target time sequence value is the time sequence value of the video clip of which the fast forward frequency is greater than a first time threshold value; the predetermined fast forward times are subtracted from the fast forward times of the video segment.

2. The method according to claim 1, wherein determining the highlight of each video segment according to the play count, the fast forward count, the rewind count and the time sequence value of each video segment in the video to be processed in such a manner that the play count and the rewind count have a positive influence on the highlight of the video segment and the fast forward count has a negative influence on the highlight of the video segment comprises:

calculating a first adjustment parameter for adjusting the order of the times to be adjusted of each video clip to the same order of magnitude as a preset reference time, wherein the times to be adjusted are as follows: the number of times except the preset reference number of times in the playing number of times, the fast forward number of times and the backspacing number of times is as follows: any one of the play times, fast forward times and rollback times;

calculating a second adjustment parameter for adjusting the fast forward times of each video clip according to the time sequence value, wherein the second adjustment parameter is: the parameter is used for reducing the influence of fast forward playing of the video clips of the video to be processed on the wonderful degree by each user;

3. The method according to claim 2, wherein the calculating a first adjustment parameter for adjusting the order of the number of times to be adjusted of each video segment to the same order of magnitude as the preset reference number of times comprises:

wherein, A is _i A first adjustment parameter for adjusting the ith times to be adjusted, i represents the type of the times to be adjusted, j is the time sequence value of the video clip, n is the total number of the video clips in the video to be processed, b _i,j Represents the ith number of times to be adjusted of the jth video segment, c _j Representing the jth videoA preset reference number of segments.

4. The method according to claim 2, wherein said calculating a second adjustment parameter for adjusting the fast forward times of each video segment according to the time sequence value comprises:

determining a median value of the time sequence values of all video segments of the video to be processed as a middle order;

5. The method according to claim 4, wherein calculating a second adjustment parameter for adjusting the fast forward times of each video segment according to the intermediate order and the time sequence value of each video segment in the video to be processed comprises:

6. The method according to any one of claims 2 to 5, wherein determining the highlight of each video segment according to the play count, the fast forward count, the rewind count, the first adjustment parameter and the second adjustment parameter of each video segment in such a manner that the play count and the rewind count have a positive influence on the highlight of the video segment and the fast forward count has a negative influence on the highlight of the video segment comprises:

7. The method according to any one of claims 1 to 5, wherein before determining the highlight of each video segment according to the play count, the fast forward count, the rewind count and the time sequence value of each video segment in the video to be processed, in such a manner that the play count and the rewind count have a positive influence on the highlight of the video segment and the fast forward count has a negative influence on the highlight of the video segment, the method further comprises:

determining a first target video clip with the incomplete times in each video clip, wherein the incomplete times are as follows: one or more items which are not successfully obtained in the playing times, the fast forwarding times and the backspacing times;

and correcting the incomplete times into the correction value.

8. The method according to any one of claims 1 to 5, wherein, in a case that the video to be processed includes a video segment belonging to a slice head and/or a video segment belonging to a slice tail, after determining the highlight degree of each video segment according to the play count, the fast forward count, the rewind count and the time sequence value of each video segment in the video to be processed, in a manner that the play count and the rewind count have a positive influence on the highlight degree of the video segment and the fast forward count has a negative influence on the highlight degree of the video segment, the method further comprises:

and/or

And correcting the wonderful degree of each video clip belonging to the head according to the wonderful degree of a third preset number of video clips which are closest to the head and do not belong to the head in the video to be processed according to the time sequence order.

9. The method according to any one of claims 1 to 5, wherein the accuracy of the calculated wonderful level is judged in the following manner:

determining the wonderful degree of each sample video clip as the sample wonderful degree according to the sample playing times, the sample fast forward times, the sample backspacing times, the first sample adjusting parameter and the second sample adjusting parameter;

10. An apparatus for determining a highlight level of a video clip, the apparatus comprising:

the number obtaining module is used for counting the playing number, the fast forward number and the backspacing number of each video clip in the video to be processed, wherein the playing number of each video clip is as follows: the total times of playing the video clip without fast forward playing or rewinding in the process of playing the video clip by each user who plays the video to be processed are as follows: the total times of fast-forwarding and playing the video clips by each user, and the backspacing times of each video clip are as follows: the total times of playing the video clip by each user back;

the degree determining module is used for determining the wonderful degree of each video clip according to the playing times, the fast forward times and the backspacing times of each video clip and the time sequence value of each video clip in the video to be processed in a mode that the playing times and the backspacing times have positive influence on the wonderful degree of the video clip and the fast forward times have negative influence on the wonderful degree of the video clip; wherein, the time sequence value of each video clip in the video to be processed is used for correcting the fast forward times of each video clip;

the positive influence of the playing times and the backspacing times on the wonderful degree of the video clip is as follows: the more the playing times are, the higher the wonderful degree of the video clip is, and the more the backspacing times are, the higher the wonderful degree of the video clip is; the negative influence of the fast forward times on the wonderful degree of the video clip is as follows: the more the fast forward times are, the lower the wonderful degree of the video clip is;

the step of correcting the fast forward times of each video clip by using the time sequence value of each video clip in the video to be processed specifically comprises the following steps: for each video clip, determining a preset fast forward frequency for the video clip according to a time sequence value of the video clip in the video to be processed, wherein the closer the time sequence value is to a target time sequence value, the larger the numerical value of the preset fast forward frequency is, and the target time sequence value is the time sequence value of the video clip of which the fast forward frequency is greater than a first time threshold value; subtracting the preset fast forward times from the fast forward times of the video clip.

11. An electronic device is characterized by comprising a processor, a communication interface, a memory and a communication bus, wherein the processor and the communication interface are used for realizing mutual communication by the memory through the communication bus;

a memory for storing a computer program;

a processor for implementing the method steps of any one of claims 1 to 9 when executing a program stored in a memory.

12. A computer-readable storage medium, characterized in that a computer program is stored in the computer-readable storage medium, which computer program, when being executed by a processor, carries out the method steps of any one of the claims 1-9.