CN115866347B

CN115866347B - Video processing method and device and electronic equipment

Info

Publication number: CN115866347B
Application number: CN202310188852.7A
Authority: CN
Inventors: 刘林; 南秀; 张宝玉
Original assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Current assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Priority date: 2023-02-22
Filing date: 2023-02-22
Publication date: 2023-08-01
Anticipated expiration: 2043-02-22
Also published as: CN115866347A

Abstract

The disclosure provides a video processing method, a video processing device and electronic equipment, relates to the technical field of artificial intelligence, and particularly relates to the fields of knowledge graph, natural language processing and the like. The specific implementation scheme is as follows: acquiring a long video to be processed; under the condition that the long video to be processed is segmented to obtain one or more candidate short videos, obtaining one or more quality feedback information of each candidate short video in the one or more candidate short videos; one or more target short videos are screened from the one or more candidate short videos based on the one or more quality feedback information of each candidate short video. The method improves the video processing efficiency and ensures the quality of the target short video.

Description

Video processing method and device and electronic equipment

Technical Field

The present disclosure relates to the field of artificial intelligence, and in particular, to the field of knowledge graph and natural language processing.

Background

With the development of computer technology and internet technology, AIGC (artificial intelligence creating content, artificial Intelligence Generated Content) technology is widely used in content creation, and the AIGC technology can be widely used in various scenes, for example, the AIGC technology is introduced to intelligently generate video based on characters; however, the video obtained in the above manner still needs a lot of man-machine interaction if it is processed into a short video and finally provided to the user, so how to efficiently process the video into the short video and simultaneously ensure the quality of the short video becomes a problem to be solved.

Disclosure of Invention

The disclosure provides a video processing method, a video processing device, electronic equipment and a storage medium.

According to a first aspect of the present disclosure, there is provided a video processing method, including:

acquiring a long video to be processed;

under the condition that the long video to be processed is segmented to obtain one or more candidate short videos, obtaining one or more quality feedback information of each candidate short video in the one or more candidate short videos;

one or more target short videos are screened from the one or more candidate short videos based on the one or more quality feedback information of each candidate short video.

According to a second aspect of the present disclosure, there is provided a video processing apparatus comprising:

the acquisition module is used for acquiring the long video to be processed;

the analysis module is used for acquiring one or more quality feedback information of each candidate short video in the one or more candidate short videos under the condition that the long video to be processed is segmented to obtain the one or more candidate short videos;

and the filtering module is used for filtering one or more target short videos from the one or more candidate short videos based on the one or more quality feedback information of each candidate short video.

According to a third aspect of the present disclosure, there is provided an electronic device comprising:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein,,

the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the video processing method of the first aspect described above.

According to a fourth aspect of the present disclosure, there is provided a non-transitory computer readable storage medium storing computer instructions for causing the computer to perform the aforementioned method.

According to a fifth aspect of the present disclosure, there is provided a computer program product comprising a computer program which, when executed by a processor, implements the aforementioned method.

It should be understood that the description in this section is not intended to identify key or critical features of the embodiments of the disclosure, nor is it intended to be used to limit the scope of the disclosure. Other features of the present disclosure will become apparent from the following specification.

According to the scheme provided by the embodiment, the long video to be processed can be segmented, the candidate short video obtained after segmentation is analyzed to obtain quality feedback information, and then the target short video is obtained based on the quality feedback information of the candidate short video through filtering. Therefore, the long video can be processed automatically in the whole flow, and the video processing efficiency is improved; and, because the target short video is screened from one or more candidate short videos according to the quality feedback information of the candidate short videos, the quality of the finally obtained target short video can be ensured.

Drawings

The drawings are for a better understanding of the present solution and are not to be construed as limiting the present disclosure. Wherein:

FIG. 1 is a flow diagram of a video processing method according to an embodiment of the present disclosure;

fig. 2 is a schematic view of a composition structure of a video processing apparatus according to another embodiment of the present disclosure;

fig. 3 is a schematic view of still another constituent structure of a video processing apparatus according to another embodiment of the present disclosure;

fig. 4 is a schematic view of a processing scenario of a video processing apparatus according to another embodiment of the present disclosure;

FIGS. 5 a-5 c are schematic diagrams of a plurality of process flows of a video processing method based on a video processing apparatus according to an embodiment of the disclosure;

fig. 6 is a block diagram of an electronic device for implementing a video processing method of an embodiment of the present disclosure.

Detailed Description

Exemplary embodiments of the present disclosure are described below in conjunction with the accompanying drawings, which include various details of the embodiments of the present disclosure to facilitate understanding, and should be considered as merely exemplary. Accordingly, one of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the present disclosure. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.

An embodiment of a first aspect of the present disclosure provides a video processing method, as shown in fig. 1, including:

s101: acquiring a long video to be processed;

s102: under the condition that the long video to be processed is segmented to obtain one or more candidate short videos, obtaining one or more quality feedback information of each candidate short video in the one or more candidate short videos;

s103: and screening one or more target short videos from the one or more candidate short videos based on the one or more quality feedback information of each candidate short video.

The video processing method can be implemented by an electronic device. The electronic device may be, for example, a first server.

By adopting the scheme, the long video to be processed can be segmented, the candidate short video obtained after segmentation is analyzed to obtain quality feedback information, and then the target short video is obtained based on the quality feedback information of the candidate short video through filtering. Therefore, the long video can be processed automatically in the whole flow, and the video processing efficiency is improved; and, because the target short video is screened from one or more candidate short videos according to the quality feedback information of the candidate short videos, the quality of the finally obtained target short video can be ensured.

In some possible embodiments, the acquiring the long video to be processed may include: and acquiring the related information of the long video to be processed from the related information of all the long videos stored in the first message queue, and acquiring the long video to be processed based on the related information of the long video to be processed.

Here, the number of the long videos to be processed may be one or more, and since the processing for each long video to be processed is the same, a detailed description is omitted.

The obtaining the relevant information of the long video to be processed from the relevant information of all the long videos stored in the first message queue may be: based on the related information of all long videos stored in a first message queue and the arrangement sequence of each long video in all long videos, acquiring the related information of the long videos to be processed with the arrangement sequence of the first N bits from the first message queue; n is a positive integer. The foregoing N may be a positive integer, and the specific value of N may be set according to practical situations, for example, may be 4, 6, or more or less, which is not limited herein.

The first message queue may be stored in a first storage space, where the first storage space may be a storage space of the first server, or may be a storage space of another server (such as a second server), and this embodiment is not limited thereto.

The first message queue may store information related to each long video in the full long video. Furthermore, in addition to storing the related information of each long video in the first message queue, all the long videos in the first message queue can be arranged according to a preset rule, that is, each long video has a corresponding arrangement sequence; the preset rule may be set according to the actual situation, for example, may be in chronological order, that is, the earliest row.

The information related to the long video to be processed may specifically be URL (Uniform Resource Locator ) of the long video to be processed.

The method for acquiring the related information of any one long video stored in the first message queue may include: acquiring a target article, and determining keywords based on the title of the target article and the content of the target article; based on the keywords, acquiring related information of the long video; the information about the long video is added to the first message queue. Here, the target article may be any one of a plurality of articles, and the target article may be an article that does not contain video content.

The determining the keyword based on the title of the target article and the content of the target article may include: and performing word segmentation on each sentence in one or more sentences contained in the content of the target article to obtain one or more word segmentation results, and determining the keywords based on the title of the target article and the one or more word segmentation results.

Acquiring related information of the long video based on the keywords; adding the long video related information to the first message queue may refer to: searching long videos matched with the keywords, taking the URL of the long videos as related information of the long videos, and adding the related information of the long videos into a first message queue. Wherein the processing of adding the related information of the long video to the first message queue includes: judging whether the first message queue stores the same content as the related information of the long video, if so, not processing, and if not, adding the related information of the long video into the first message queue.

By adopting the scheme, the long videos to be processed can be obtained in batches, so that the problem that in the related art, the mapping relation cost is required to be maintained because the long videos to be processed are segmented and then sent to different servers to be processed by the different servers can be avoided.

In some possible embodiments, the method further comprises: calling a segmentation operator to process the long video to be processed, and obtaining segmentation processing information fed back by the segmentation operator; the segmentation operator is an operator positioned in a first calling sequence in a plurality of operators contained in an operator calling graph; and determining the one or more candidate short videos based on the segmentation processing result contained in the segmentation processing information under the condition that the segmentation processing state contained in the segmentation processing information is that the segmentation processing is successful.

The calling segmentation operator processes the long video to be processed to obtain segmentation processing information fed back by the segmentation operator, which specifically may be: and under the condition that operators in the first calling sequence are determined to be segmentation operators based on the operator calling graph, calling the segmentation operators to process the long video to be processed based on calling addresses of the segmentation operators, and obtaining segmentation processing information fed back by the segmentation operators.

The operator call graph can be configured and preset according to actual conditions. Specifically, the operator call graph may be preconfigured by a manager according to actual requirements, and preset in a first server executing the video processing method of the embodiment; the operator call graph may include: related information of each operator in the plurality of operators, and calling sequence of each operator. The related information of each operator may include: identification information of the operator and a calling address of the operator; wherein, the identification information of the operator can comprise at least one of the identification of the operator, the name of the operator, the number of the operator and the like.

The segmentation processing information may include: cutting the processing state; the segmentation processing state comprises: the segmentation process fails or the segmentation process is successful. Further, in the case that the segmentation processing state included in the segmentation processing information is that the segmentation processing fails, other contents may not be included in the segmentation processing information; in the case that the segmentation processing state included in the segmentation processing information is that the segmentation processing is successful, the segmentation processing information may also include a segmentation processing result, where the segmentation processing result may include a plurality of segmentation videos.

The segmentation operator can be an operator or a model with a segmentation function. The specific functions of the segmentation operator can be: determining a starting time point and an ending time point of each shot in one or more shots in the long video to be processed, and segmenting the long video to be processed based on the starting time point and the ending time point of each shot in the video to be processed to obtain segmented videos corresponding to each shot. Here, the slice video corresponding to any shot only includes video under one shot, that is, there is no shot switching.

In one possible example, the determining the one or more candidate short videos based on the segmentation processing result included in the segmentation processing information may include: and extracting one or more segmented videos from the segmentation processing results contained in the segmentation processing information, and directly taking the one or more segmented videos as the one or more candidate short videos. That is, all the segmented videos included in the segmentation processing result in the segmentation processing information fed back by the segmentation operator can be directly used as all the candidate short videos.

By adopting the scheme, the operator of the first calling sequence can be determined to be the segmentation operator based on the operator calling graph, and then the segmentation operator is called to process the long video to be processed so as to obtain one or more candidate short videos. The operator call graph can be generated and configured in advance through the graph engine technology, so that operators can be directly called in sequence through the operator call graph in the video processing process, and the problem of low efficiency caused by manual participation in adjusting the operators in the video processing process is avoided; in addition, as the operator call graph can be flexibly arranged, the operator call graph can be efficiently updated and iterated, and the flexible convenience of arranging and processing is also ensured.

In some possible embodiments, the determining the one or more candidate short videos based on the segmentation processing result included in the segmentation processing information includes: extracting one or more segmentation videos from the segmentation processing results contained in the segmentation processing information; from the one or more sliced videos, the one or more candidate short videos within a video length threshold are determined.

The details of the foregoing embodiments are described in detail, and the description thereof will not be repeated.

The determining the one or more candidate short videos within the video length threshold from the one or more segmented videos may specifically include: selecting a kth segmented video from the one or more segmented videos; judging whether the time length of the kth video segmentation is within a video length threshold value or not; if the video length is within the video length threshold, the kth segmented video is used as one of the one or more candidate short videos; and if the video length is not within the video length threshold, deleting the kth segmented video. The k is a positive integer, the kth segmented video is any one of the one or more segmented videos, and the processing of each segmented video in the one or more segmented videos is the same as that of the kth segmented video, so that a detailed description is omitted.

Wherein the video length threshold may include a video length minimum; the video length minimum value may be set according to the actual situation, for example, the video length minimum value may be 3 seconds, or 5 seconds, or longer or shorter. Alternatively, the video length threshold may include: video length minimum and video length maximum; wherein, the maximum value of the video length is larger than the minimum value of the video length; the maximum video length may be set according to the actual situation, and may be 10 seconds, or 15 seconds, or longer or shorter, for example. Alternatively, the video length threshold may comprise a video length maximum.

By adopting the scheme, the one or more segmented videos contained in the segmentation processing information can be subjected to coarse screening and filtering to obtain one or more candidate short videos meeting the requirement of the preset video length, so that the duration of the target short videos obtained later is not less than the preset video length, and the quality of the target short videos is ensured.

In some possible embodiments, the method further comprises: recording segmentation failure information of the long video to be processed under the condition that the processing state contained in the segmentation processing information is the segmentation processing failure; the segmentation failure information of the long video to be processed comprises: and the related information of the long video to be processed and the failure information of the segmentation operator.

The recording of the segmentation failure information of the long video to be processed may refer to: the first server generates segmentation failure information of the long video to be processed, and stores the segmentation failure information of the long video to be processed in a second storage space. The second storage space may be a storage space local to the first server, or may be a storage space of another server; the other server may refer to any one of servers other than the first server.

The information about the long video to be processed may include: the name of the long video to be processed, the URL of the long video to be processed, the number of the long video to be processed, the identification of the long video to be processed, and the like. The failure information of the segmentation operator may include: the method comprises the steps of determining at least one of names of segmentation operators, identification of the segmentation operators, number of the segmentation operators, segmentation processing states of the long video to be processed by the segmentation operators are segmentation processing failure and the like.

By adopting the scheme, the method and the device can be based on the related information of the long video to be processed and the failure information of the segmentation operator under the condition that the processing of the segmentation operator fails; therefore, accurate reference information can be provided when updating is carried out later, and the accuracy of operator iteration is ensured.

In some possible embodiments, the method further comprises: determining a current long video to be processed based on related information of the long video to be processed in the segmentation failure information of the long video to be processed; based on the failure information of the segmentation operator in the segmentation failure information of the long video to be processed, re-calling the segmentation operator to process the current long video to be processed, and obtaining updated segmentation processing information fed back by the segmentation operator; determining one or more updated candidate short videos based on a processing result contained in the updated segmentation processing information under the condition that the processing state contained in the updated segmentation processing information is successful in segmentation processing, and acquiring one or more quality feedback information of each updated candidate short video in the one or more updated candidate short videos; and screening one or more target short videos from the one or more updated candidate short videos based on the one or more quality feedback information of each updated candidate short video.

The determining the current long video to be processed based on the related information of the long video to be processed in the segmentation failure information of the long video to be processed may specifically include: and responding to the first instruction, and determining the current long video to be processed based on the related information of the long video to be processed in the segmentation failure information of the long video to be processed.

The first instruction is specifically configured to instruct re-splitting of the long video to be processed included in the splitting failure information. The first instruction may specifically be initiated periodically by the first server, for example, the second storage space is a storage space of the first server, and may be the first instruction that, when the current period is over, extracts the segmentation failure information of all the long videos to be processed stored in the second storage space, and generates the segmentation failure information carrying all the long videos to be processed. Or, the first instruction may be sent periodically by another server, for example, the second storage space is a storage space of the other server, and when the other server determines that the current period is over, the other server extracts the splitting failure information of the long video to be processed stored in the second storage space, generates a first instruction carrying the splitting failure information of the long video to be processed, and then sends the first instruction to the first server.

The determining the current long video to be processed based on the related information of the long video to be processed in the segmentation failure information of the long video to be processed may specifically refer to: and acquiring the long video to be processed based on the URL of the long video to be processed in the segmentation failure information of the long video to be processed, and taking the long video to be processed as the current long video to be processed.

The specific processing mode of the updated segmentation processing information fed back by the segmentation operator is the same as the processing mode of the segmentation processing information fed back by the segmentation operator, and therefore details are not repeated.

The determining, based on the processing result included in the updated segmentation processing information, one or more updated candidate short videos when the processing state included in the updated segmentation processing information is that the segmentation processing is successful, and the determining, based on the segmentation processing result included in the segmentation processing information, that the specific processing of the one or more candidate short videos is also the same when the segmentation processing state included in the segmentation processing information is that the segmentation processing is successful, so that repeated description is not made.

It should be noted that, in the case that the processing state included in the updated segmentation processing information is that the segmentation processing fails, the current short video to be processed may not be processed again, for example, the current short video to be processed may be deleted; or, the segmentation failure information of the long video to be processed can be recorded again, and the segmentation processing is carried out on the current short video to be processed again based on the segmentation failure information of the long video to be processed in the next period, so that the process is executed circularly until the segmentation ending condition is met; the segmentation ending condition may include: the current short video to be processed is successfully segmented, or the number of times of re-segmentation of the current short video to be processed reaches a first preset number of times; the first preset number of times may be set according to practical situations, for example, 3 times, or 4 times, or more or less, and is not limited.

Therefore, by adopting the scheme, under the condition that the processing of the long video to be processed by the segmentation operator fails, the segmentation operator is automatically re-called to re-segment the long video to be processed which is failed to be processed based on the segmentation failure information of the long video to be processed, so that the problem that the processing efficiency is affected caused by manually updating the call operator can be avoided, and the high efficiency and the accuracy of video processing are ensured.

In some possible embodiments, the obtaining the one or more quality feedback information of each candidate short video of the one or more candidate short videos includes:

invoking one or more analysis operators to process an ith candidate short video in the one or more candidate short videos to obtain one or more quality feedback information of the ith candidate short video; wherein, each analysis operator in the one or more analysis operators is an operator positioned in a second calling sequence in the operator calling graph, and the functions of different analysis operators in the one or more analysis operators are different; in one or more quality feedback information of the ith candidate short video, different quality feedback information is obtained by different analysis operators; i is a positive integer.

The i-th candidate short video is any one of the one or more candidate short videos, and the processing for each candidate short video in the one or more candidate short videos is the same as the i-th candidate short video, so that a detailed description is omitted.

The aforementioned one or more analysis operators refer to: one or more operators located in the second call order are juxtaposed in the operator call graph. The second calling order follows the first calling order, that is, the calling order of each of the one or more analysis operators is parallel and follows the calling order of the segmentation operator.

Any one of the one or more analysis operators may also be referred to as an analysis model, or a model with analysis functionality, or a neural network with analysis functionality, etc.

Among the aforementioned one or more analysis operators, different analysis operators differ in function. For example, the aforementioned one or more analysis operators may include at least one of: a definition operator, a black edge operator, a watermark operator, a face detection operator, a mosaic detection operator and a correlation operator; the definition operator is used for scoring the definition of the candidate short video; the black edge operator is used for judging whether black edges exist in the candidate short videos or not; the watermark operator is used for determining whether watermark information of the bid product exists in the candidate short video; the face detection operator is used for determining whether a face exists in the candidate short video, and labeling relevant information of the face when the face exists, wherein the relevant information of the face can comprise at least one of the following steps: the number of the face, the identification of the face, the name of the face, etc.; the mosaic detection operator is used for determining whether a mosaic exists in the candidate short video; the correlation operator is used for determining correlation between the candidate short video and a video tag, where the video tag may refer to a video tag of a long video to be processed corresponding to the candidate short video, and may include at least one of a video title, a classification tag, and the like. It should be appreciated that the above is merely an exemplary illustration of analysis operators, and that in actual processing, the one or more analysis operators described above may include, but are not limited to, the operators listed above, but are not exhaustive in this embodiment.

It should be further noted that, for the foregoing one or more analysis operators, the one or more analysis operators may be disposed in the first server, or may be disposed in another server (such as a third server, a fourth server, etc.), and different analysis operators may be disposed in different servers, which is not exhaustive about possible locations where each analysis operator may be disposed in this embodiment.

The invoking one or more analysis operators to process the ith candidate short video in the one or more candidate short videos to obtain one or more quality feedback information of the ith candidate short video may specifically include: and simultaneously calling one or more analysis operators to process the ith candidate short video in the one or more candidate short videos to obtain quality feedback information of the ith candidate short video fed back by each analysis operator in the one or more analysis operators.

By adopting the scheme, one or more operators in the second calling sequence in the operator calling graph can be directly used as one or more analysis operators, and one or more quality feedback information of the ith candidate short video can be obtained by calling one or more analysis operators to process the ith candidate short video in parallel. In this way, the operator call graph can be pre-generated and configured through the graph engine technology, so that operators can be directly called in sequence through the operator call graph in the video processing process, and the problem of lower efficiency caused by manual participation in adjusting operators in the video processing process is avoided; in addition, as the operator call graph can be flexibly arranged, the operator call graph can be efficiently updated and iterated, and the flexible convenience of arranging and processing is also ensured.

In some possible implementations, the screening one or more target short videos from the one or more candidate short videos based on the one or more quality feedback information of each candidate short video includes: determining one or more quality detection results of the ith candidate short video based on the one or more quality feedback information of the ith candidate short video under the condition that the processing state of each of the one or more analysis operators is successful in processing; and determining the ith candidate short video as one of the one or more target short videos in the case that the preset condition is met based on one or more quality detection results of the ith candidate short video.

Based on the one or more quality feedback information of the i-th candidate short video, determining a manner of processing state of each of the one or more analysis operators may include: acquiring a processing state of a j-th analysis operator from j-th quality feedback information of the i-th candidate short video; the processing state of the jth analysis operator may include whether the processing state of the jth analysis operator on the ith candidate short video is successful.

The j-th quality feedback information is any one of one or more quality feedback information, and is fed back by a j-th analysis operator; the j-th analysis operator is any one of the one or more analysis operators.

Specifically, the j-th quality feedback information may include: the j-th analysis operator's related information; whether the processing state of the j-th analysis operator to the i-th candidate short video is successful; and under the condition that the processing state of the j-th analysis operator on the i-th candidate short video is successful in processing, the j-th analysis operator detects the quality of the i-th candidate short video. The reason why the processing state of the jth analysis operator on the ith candidate short video is the processing failure may include: the analysis operator fails to call, and the quality detection result obtained by the analysis operator does not accord with the preset requirement. The preset requirements can be set according to actual conditions, and the preset requirements of different analysis operators can be different; for example, any analysis operator is a definition operator, where the definition operator is used to determine the definition of the ith candidate short video, and correspondingly, the preset requirement of the definition operator may be that the type of the quality detection result is a numerical value.

The determining, based on the one or more quality feedback information of the i-th candidate short video, one or more quality detection results of the i-th candidate short video includes: extracting a quality detection result of a j-th analysis operator on the i-th candidate short video from the j-th quality feedback information of the i-th candidate short video; and taking the quality detection result of the j-th analysis operator on the i-th candidate short video as the j-th quality detection result of the i-th candidate short video. The j-th quality feedback information is any one of the one or more quality feedback information, and the processing based on each quality feedback information is the same, so that a detailed description is omitted.

By adopting the scheme, whether the candidate short video can be used as the target short video can be determined based on preset conditions after all quality detection results of the candidate short video are obtained under the condition that the processing state of each analysis operator is determined to be successful for a certain candidate short video. Therefore, the final target short video is ensured to be obtained efficiently, and the quality of the target short video can be ensured because the final target short video needs to be the short video filtered based on the quality detection result of the candidate short video.

In some possible embodiments, the one or more quality measurements include at least one of: definition scoring, whether there is a mosaic, correlation scoring. The preset condition includes at least one of: the sharpness score is greater than a preset sharpness threshold, no mosaic is present, and the correlation score is greater than a preset correlation threshold.

The preset definition threshold can be configured according to actual conditions, and is not limited; the preset correlation threshold may also be configured according to practical situations, for example, may be greater than 90%, greater than 95%, or greater than or less than 90%, which is not limited herein.

It should be appreciated that the one or more quality measurements described above may include additional quality measurements in addition to those described above, such as at least one of: whether there is a black edge, and the position and size of the black edge in the case where there is a black edge, and the like; whether a bid watermark exists, and if so, the related information of the bid may include: at least one of the name of the bid, the identification of the bid, the number of the bid, etc.; whether a face exists or not, and related information of the face in case that the face exists, the related information of the face may include: at least one of identification of a face, number of a face, name of a face, and the like. It should be understood that this is also merely an exemplary illustration, the analysis operator may be adjusted according to the actual requirement in the actual processing, and the quality detection result of any corresponding candidate short video may also be different according to the analysis operator, which is not exhaustive in this embodiment.

In one example, the quality detection results of the sharpness score of the i-th candidate short video, whether a mosaic exists in the i-th candidate short video, and the correlation between the i-th candidate short video and the video tag may be scored as one or more quality detection results of the i-th candidate short video for comparison. Correspondingly, the preset conditions may also include: one or more sub-conditions corresponding to each of the quality detection results for comparison; for example, the quality detection result for comparison is that the definition score of the i candidate short video is greater than a preset definition threshold value; the quality detection result for comparison is whether the ith candidate short video has a mosaic or not, and the corresponding sub-condition is that the mosaic does not exist; and the quality detection result for comparison is used for scoring the correlation between the ith candidate short video and the video label, and the corresponding sub-condition is that the correlation score is larger than a preset correlation threshold.

In this example, in a case where it is determined that a preset condition is satisfied based on one or more quality detection results of the i-th candidate short video, the i-th candidate short video is determined to be one of the one or more target short videos, specifically: judging whether each quality detection result for comparison in one or more quality detection results of the ith candidate short video meets a corresponding sub-condition or not; and under the condition that each quality detection result for comparison meets the corresponding sub-condition, determining the ith candidate short video as one of the one or more target short videos. In addition, it may further include: and under the condition that any quality detection result for comparison does not meet the corresponding sub-condition, determining that the ith candidate short video is not one of the one or more target short videos.

Therefore, by adopting the scheme, whether the candidate short video can be used as the target short video can be judged based on the quality detection result of any one candidate short video and the corresponding preset condition. Therefore, candidate short videos can be screened in multiple dimensions, and the quality of the finally obtained target short videos is guaranteed.

In some possible embodiments, the method further comprises: when determining that the processing state of at least one analysis operator in the one or more analysis operators is processing failure based on one or more quality feedback information of the ith candidate short video, taking the ith candidate short video as an mth short video to be processed, and recording failure information of the mth short video to be processed; m is a positive integer;

wherein the failure information of the mth short video to be processed includes at least one of the following: the related information of the mth short video to be processed, the related information of each failure analysis operator in at least one failure analysis operator, and the quality feedback information of the mth short video to be processed corresponding to each success analysis operator in at least one success analysis operator; each failure analysis operator is an analysis operator with a processing state of processing failure; and each successful analysis operator is an analysis operator with a processing state of processing success.

The recording the failure information of the mth short video to be processed may refer to: and generating failure information of the mth short video to be processed, and storing the failure information of the mth short video to be processed in a second storage space.

The related information of the mth short video to be processed may include: the name of the mth short video to be processed, the URL of the mth short video to be processed, the number of the mth short video to be processed, the identification of the mth short video to be processed, and the like. The relevant information for each failure analysis operator may include: at least one of the name of the failure analysis operator, the identity of the failure analysis operator, the number of the failure analysis operator, etc.

By adopting the scheme, under the condition that any one candidate short video fails to be processed, the failure information of the candidate short video can be recorded, and the failure information of the candidate short video can comprise the related information of a failure analysis operator and the related information of the candidate short video. Therefore, effective information can be provided for subsequent updating of the failure analysis operator and reprocessing of the candidate short video, the short video is guaranteed to be fully explored, and the success rate of the analysis operator in subsequent processing is improved.

In some possible embodiments, the method may further include: acquiring an mth short video to be processed based on related information of the mth short video to be processed in failure information of the mth short video to be processed; determining current analysis operators corresponding to the at least one failure analysis operator respectively based on related information of the at least one failure analysis operator in failure information of the mth short video to be processed; invoking the current analysis operators respectively corresponding to the at least one failure analysis operator, and processing the mth short video to be processed to obtain at least one updated quality feedback information of the mth short video to be processed; obtaining one or more quality feedback information of the mth short video to be processed based on the quality feedback information of the mth short video to be processed, which is updated, and the quality feedback information of the mth short video to be processed corresponding to each successful analysis operator in the failure information of the mth short video to be processed; determining one or more quality detection results of the mth short video to be processed based on the one or more quality feedback information of the mth short video to be processed; and determining the mth short video to be processed as one of the one or more target short videos under the condition that the preset condition is met based on one or more quality detection results of the mth short video to be processed.

The method may further include, before obtaining the mth short video to be processed, obtaining related information of the mth short video to be processed in the failure information based on the mth short video to be processed, where the related information includes: obtaining failure information of each short video to be processed in one or more short videos to be processed in a current period; and extracting the failure information of the mth short video to be processed from the failure information of each short video to be processed. The length of the current period may be set according to practical situations, for example, may be 1 day, 2 days, or longer or shorter, and is not limited herein.

For example, the obtaining the failure information of each short video to be processed in the one or more short videos to be processed in the current period may be: and responding to the second instruction, and extracting failure information of each short video to be processed in one or more short videos to be processed in the current period contained in the second instruction.

The second instruction is specifically configured to instruct to recall a failure analysis operator to process each short video to be processed based on failure information of each short video to be processed. The second instruction may specifically be initiated periodically by the first server, for example, when the current period ends, the first server extracts failure information of all short videos to be processed in the current period stored in the second storage space of the first server, and generates a second instruction carrying the failure information of all short videos to be processed in the current period. Or, the second instruction may be sent periodically by another server (such as the second server), for example, in the case that the other server determines that the current period ends, extracts failure information of all short videos to be processed in the current period stored in the second storage space of the other server, generates a second instruction carrying the failure information of all short videos to be processed in the current period, and then sends the second instruction to the first server.

The specific processing manner of calling the current analysis operator corresponding to the at least one failure analysis operator to process the mth short video to be processed to obtain the at least one updated quality feedback information of the mth short video to be processed is the same as the processing manner of calling the one or more analysis operators to process the ith candidate short video in the one or more candidate short videos to obtain the one or more quality feedback information of the ith candidate short video, so that the detailed description is omitted.

The obtaining the one or more quality feedback information of the mth short video to be processed based on the quality feedback information of the mth short video to be processed corresponding to the each successful analysis operator in the at least one updated quality feedback information of the mth short video to be processed and the failure information of the mth short video to be processed may specifically include:

and merging the quality feedback information of at least one updated m short video to be processed and the quality feedback information of the m short video to be processed corresponding to each successful analysis operator contained in the failure information of the m short video to be processed under the condition that the processing states of the current analysis operators respectively corresponding to the at least one failure analysis operator are all successful in processing, so as to obtain one or more quality feedback information of the m short video to be processed.

The content that may be included in the one or more quality detection results and the content that may be included in the preset condition are the same as those in the foregoing embodiment, and therefore will not be described in detail.

It should be further noted that, in the case that it is determined, based on the at least one updated quality feedback information of the mth short video to be processed, that the processing state of at least one current analysis operator exists in the current analysis operators corresponding to the at least one failure analysis operator respectively is a processing failure, the processing of the mth short video to be processed may be ended, for example, the mth short video to be processed may be deleted; or, the failure information of the mth short video to be processed can be recorded again, and the mth short video to be processed is reprocessed based on the failure information of the mth short video to be processed in the next period, and the processing is circularly performed until the analysis ending condition is met; wherein the analysis end condition may include: the m-th short video to be processed is successfully processed, or the number of times of reprocessing the m-th short video to be processed reaches a second preset number of times; the second preset number of times may be set according to practical situations, such as 2 times, or 3 times, or more or less, and is not limited thereto.

By adopting the scheme, under the condition that the failure operator appears in the processing process of any one candidate short video, the candidate short video is used as the short video to be processed, and then the current analysis operator corresponding to the failure analysis operator is called again to analyze the short video to be processed based on the failure information of the short video to be processed. Therefore, the effectiveness of the quality detection result of the short video to be processed and the quality of the target short video can be ensured.

In some possible embodiments, after screening one or more target short videos from the one or more candidate short videos, the method further comprises: responding to a received search request, and selecting a short video to be sent from the one or more target short videos based on the search request; and generating and sending a reply message of the search request based on the related information of the short video to be sent.

The search request may be sent by the user equipment, and accordingly, a reply message of the search request may also be sent to the user equipment.

The selecting, based on the search request, a short video to be sent from the one or more target short videos may specifically include: and acquiring a search keyword from the search request, and selecting a short video to be sent, which is matched with the search keyword, from the one or more target short videos.

Each of the one or more target short videos may further have a tag of the target short video, and the tag of any one target short video may include at least one of the following: video tags, face tags, and the like. The video tag may be a video tag of a long video to be processed corresponding to the target short video, and the face tag may be related information of a face contained in a quality detection result obtained by a face detection operator. Correspondingly, the selecting the short video to be sent, which is matched with the search keyword, from the one or more target short videos includes: and selecting the short video to be sent, which is matched with the search keyword, based on the label of each target short video in the one or more target short videos.

It should be understood that the number of short videos to be sent may be one or more, and the number is not limited in this embodiment.

In addition, the one or more target short videos can be stored in the second message queue first, then the one or more target short videos are extracted from the second message queue, and the one or more target short videos are stored in an inverted storage mode; it should be noted that the one or more target short videos may also be stored in other storage manners, which is not meant to be exhaustive in this embodiment.

Further, after receiving the reply message of the search request, the user equipment may extract the short video to be sent from the reply message of the search request and store the short video as a local short video, and then may edit and merge the local short video to obtain a merged video.

By adopting the scheme, when the search request is received, the short video to be sent can be determined from one or more target short videos, and the related information of the short video to be sent is carried in the reply message of the search request and sent. Therefore, the high-quality short video provided for the user side can be ensured, and the quality of the combined video obtained by the end user is ensured.

An embodiment of a second aspect of the present disclosure provides a video processing apparatus, as shown in fig. 2, including:

an acquisition module 201, configured to acquire a long video to be processed;

an analysis module 202, configured to obtain one or more quality feedback information of each candidate short video in the one or more candidate short videos when the long video to be processed is segmented to obtain the one or more candidate short videos;

and a filtering module 203, configured to filter one or more target short videos from the one or more candidate short videos based on the one or more quality feedback information of each candidate short video.

On the basis of fig. 2, as shown in fig. 3, the apparatus further includes:

the segmentation processing module 301 is configured to invoke a segmentation operator to process the long video to be processed, so as to obtain segmentation processing information fed back by the segmentation operator; the segmentation operator is an operator positioned in a first calling sequence in a plurality of operators contained in an operator calling graph; and determining the one or more candidate short videos based on the segmentation processing result contained in the segmentation processing information under the condition that the segmentation processing state contained in the segmentation processing information is that the segmentation processing is successful.

The segmentation processing module 301 is configured to extract one or more segmentation videos from the segmentation processing result included in the segmentation processing information; from the one or more sliced videos, the one or more candidate short videos within a video length threshold are determined.

The apparatus further comprises: the failure recording module 302 is configured to record segmentation failure information of the long video to be processed when the segmentation processing state included in the segmentation processing information is that the segmentation processing fails; the segmentation failure information of the long video to be processed comprises: and the related information of the long video to be processed and the failure information of the segmentation operator.

The apparatus further comprises: an updating module 303, configured to determine a current long video to be processed based on related information of the long video to be processed in the segmentation failure information of the long video to be processed; based on the failure information of the segmentation operator in the segmentation failure information of the long video to be processed, re-calling the segmentation operator to process the current long video to be processed, and obtaining updated segmentation processing information fed back by the segmentation operator; determining one or more updated candidate short videos based on a processing result contained in the updated segmentation processing information under the condition that the processing state contained in the updated segmentation processing information is successful in segmentation processing, and acquiring one or more quality feedback information of each updated candidate short video in the one or more updated candidate short videos;

the filtering module 203 is configured to filter one or more target short videos from the one or more updated candidate short videos based on the one or more quality feedback information of each updated candidate short video.

The analysis module 202 is configured to invoke one or more analysis operators to process an ith candidate short video in the one or more candidate short videos, so as to obtain one or more quality feedback information of the ith candidate short video; wherein, each analysis operator in the one or more analysis operators is an operator positioned in a second calling sequence in the operator calling graph, and the functions of different analysis operators in the one or more analysis operators are different; in one or more quality feedback information of the ith candidate short video, different quality feedback information is obtained by different analysis operators; i is a positive integer.

The filtering module 203 is configured to determine, based on the one or more quality feedback information of the i-th candidate short video, one or more quality detection results of the i-th candidate short video if it is determined that the processing state of each of the one or more analysis operators is successful; and determining the ith candidate short video as one of the one or more target short videos in the case that the preset condition is met based on one or more quality detection results of the ith candidate short video.

The one or more quality measurements include at least one of: definition scoring, whether mosaic exists or not, and correlation scoring; the preset condition includes at least one of: the sharpness score is greater than a preset sharpness threshold, no mosaic is present, and the correlation score is greater than a preset correlation threshold.

A failure recording module 302, configured to, when it is determined that a processing state of at least one of the one or more analysis operators is a processing failure based on one or more quality feedback information of the ith candidate short video, take the ith candidate short video as an mth short video to be processed, and record failure information of the mth short video to be processed; m is a positive integer;

Wherein the failure information of the mth short video to be processed includes at least one of the following: the related information of the mth short video to be processed, the related information of at least one failure analysis operator and the quality feedback information of the mth short video to be processed corresponding to each success analysis operator in at least one success analysis operator; each failure analysis operator is an analysis operator with a processing state of processing failure; and each successful analysis operator is an analysis operator with a processing state of processing success.

An updating module 303, configured to obtain an mth short video to be processed based on related information of the mth short video to be processed in failure information of the mth short video to be processed; determining current analysis operators corresponding to the at least one failure analysis operator respectively based on related information of the at least one failure analysis operator in failure information of the mth short video to be processed; invoking the current analysis operators respectively corresponding to the at least one failure analysis operator, and processing the mth short video to be processed to obtain at least one updated quality feedback information of the mth short video to be processed; obtaining one or more quality feedback information of the mth short video to be processed based on the quality feedback information of the mth short video to be processed, which is updated, and the quality feedback information of the mth short video to be processed corresponding to each successful analysis operator in the failure information of the mth short video to be processed;

The filtering module is used for determining one or more quality detection results of the mth short video to be processed based on one or more quality feedback information of the mth short video to be processed; and determining the mth short video to be processed as one of the one or more target short videos under the condition that the preset condition is met based on one or more quality detection results of the mth short video to be processed.

The apparatus further comprises: a storage module 304, configured to store the one or more target short videos; a retrieving module 305, configured to select, in response to receiving a retrieving request, a short video to be sent from the one or more target short videos based on the retrieving request; and generating and sending a reply message of the search request based on the related information of the short video to be sent.

On the basis of the video processing apparatus illustrated in fig. 2 or fig. 3, the video processing apparatus provided in this embodiment may further include: the long video mining module is used for acquiring a target article and determining keywords based on the title of the target article and the content of the target article; based on the keywords, acquiring related information of the long video; the information about the long video is added to the first message queue.

The acquisition module is used for acquiring the relevant information of each candidate long video in one or more candidate long videos from the relevant information of all the long videos stored in the first message queue; acquiring each candidate long video based on the related information of each candidate long video; and taking one of the one or more candidate long videos as the long video to be processed.

The acquisition module is used for acquiring the related information of each candidate long video in the candidate long videos with the arrangement sequence of the first N bits from the first message queue based on the related information of all the long videos stored in the first message queue and the arrangement sequence of each long video in all the long videos; n is a positive integer. For example, the obtaining module is configured to use any one of the one or more candidate long videos as the long video to be processed; or, using the earliest one of the one or more candidate long videos as the long video to be processed.

The first message queue may be stored in the first storage space, where information about each long video in the full long video may be stored in the first message queue. Furthermore, in addition to storing the related information of each long video in the first message queue, all the long videos in the first message queue can be arranged according to a preset rule, that is, each long video has a corresponding arrangement sequence; the preset rule may be set according to the actual situation, for example, may be in chronological order, that is, the earliest row.

The foregoing N may be a positive integer, and the specific value of N may be set according to practical situations, for example, may be 4, 6 or more or less, which is not limited herein.

The foregoing information about each long video may be specifically a URL of the long video.

The aforementioned target article may be an article that does not contain video content.

The long video mining module is used for cutting each sentence in one or more sentences contained in the content of the target article to obtain one or more word cutting results, and determining the keywords based on the title of the target article and the one or more word cutting results; searching long videos matched with the keywords, taking the URL of the long videos as related information of the long videos, and adding the related information of the long videos into a first message queue.

It should be noted that the foregoing modules may be disposed on the same server or may be disposed on different servers. For example, the acquisition module, the analysis module, the filtering module, the segmentation processing module, the updating module, the failure recording module, and the retrieval module may be disposed on a first server, and the long video mining module may be disposed on a second server; wherein the first server and the second server are different. For another example, the long video mining module, the obtaining module, the analyzing module, the filtering module, the segmentation processing module, the updating module, the failure recording module, the sorting module, and the retrieving module may all be disposed on the first server. For another example, the acquisition module, the analysis module, the filtering module, and the segmentation processing module may be disposed in a first server, the long video mining module may be disposed in a second server, the update module and the failure recording module may be disposed in a third server, and the retrieval module may be disposed in a fourth server; the first server, the second server, the third server and the fourth server are respectively different servers. The above is merely an exemplary description that each module is disposed on the same or different servers, and there may be more manners of disposing each module in actual processing, and this embodiment is not exhaustive of the manner in which each module is disposed on a server.

An exemplary description is given of the video processing method provided in combination with the foregoing embodiment of the first aspect, and the video processing apparatus provided in the embodiment of the second aspect, with reference to fig. 4, 5a, 5b, and 5 c. First, description will be made with reference to fig. 4 and 5a, which specifically include:

s501: the acquisition module 411 acquires a long video to be processed.

Specifically, the obtaining module 411 may obtain relevant information of the long video to be processed from relevant information of all long videos stored in the first message queue 401, and obtain the long video to be processed based on the relevant information of the long video to be processed. The number of the long videos to be processed may be one or more, and this embodiment is not limited thereto.

The obtaining of the related information about any long video saved in the first message queue 401 may be implemented by the long video mining module 417 in fig. 4, specifically, the long video mining module 417 obtains a target article, and determines a keyword based on a title of the target article and a content of the target article; based on the keywords, acquiring related information of the long video; the relevant information of the long video is added to the first message queue 401.

S502: the segmentation processing module 412 invokes a segmentation operator 420 to process the long video to be processed, so as to obtain segmentation processing information fed back by the segmentation operator 420.

S503: the segmentation processing module 412 determines whether the segmentation processing status included in the segmentation processing information is that the segmentation processing is successful, if yes, S504 is executed; otherwise, S505 is performed.

S504: the segmentation processing module 412 extracts one or more segmented videos from the segmentation processing result included in the segmentation processing information; from the one or more sliced videos, the one or more candidate short videos within a video length threshold are determined.

S505: the failure recording module 413 records the segmentation failure information of the long video to be processed.

The specific content of the segmentation failure information of the long video to be processed is the same as that of the foregoing embodiment, and will not be repeated here.

Referring to fig. 4 and 5b, the process after completing the foregoing step S505 is described, and may specifically include:

s5061: the updating module 416 determines the current long video to be processed based on the related information of the long video to be processed in the segmentation failure information of the long video to be processed.

S5062: the updating module 416 recalls the segmentation operator to process the current long video to be processed based on the failure information of the segmentation operator in the segmentation failure information of the long video to be processed, and obtains updated segmentation processing information fed back by the segmentation operator.

S5063: the updating module 416 determines one or more updated candidate short videos based on a processing result included in the updated segmentation processing information when the processing state included in the updated segmentation processing information is that the segmentation processing is successful.

S5064: the update module 416 obtains one or more quality feedback information for each of the one or more updated candidate short videos.

S5065: the filtering module 415 filters one or more target short videos from the one or more updated candidate short videos based on the one or more quality feedback information for each updated candidate short video.

After the foregoing S504 is completed, one or more candidate short videos may be obtained, where any one of the one or more candidate short videos is hereinafter referred to as an i-th candidate short video, and exemplary steps are described with reference to fig. 4 and 5c, and include:

s5071: the analysis module 414 invokes one or more analysis operators to process an ith candidate short video in the one or more candidate short videos to obtain one or more quality feedback information of the ith candidate short video.

Illustratively, the one or more analysis operators may, as shown in fig. 4, include: sharpness operator 421, black edge operator 422, watermark operator 423, face detection operator 424, mosaic detection operator 425, correlation operator 426. It should be noted that the above analysis operators may be located on the same server or on different servers, which is not limited in this example.

S5072: the filtering module 415 determines, based on the one or more quality feedback information of the i-th candidate short video, whether the processing status of each of the one or more analysis operators is successful, if so, S5073 is executed, otherwise S5075 is executed.

S5073: the filtering module 415 determines one or more quality detection results of the ith candidate short video based on the one or more quality feedback information of the ith candidate short video.

The different analysis operators are used for obtaining different quality detection results. The quality detection results of the sharpness operator include, for example, in connection with the individual analysis operators contained in fig. 3: scoring the definition; the quality detection result of the black edge operator comprises: whether a black edge exists; the quality detection result of the watermark operator comprises: whether watermark information of the bid exists or not; the quality detection result of the face detection operator comprises: whether a face exists or not, and labeling relevant information of the face in the case of the face, wherein the relevant information of the face can comprise at least one of the following steps: the number of the face, the identification of the face, the name of the face, etc.; the quality detection result of the mosaic detection operator comprises: whether or not a mosaic is present; the quality detection result of the correlation operator comprises: correlation with video tags.

S5074: the filtering module 415 determines whether one or more quality detection results of the i-th candidate short video meet a preset condition, and if so, determines that the i-th candidate short video is one of the one or more target short videos; if not, deleting the ith candidate short video.

After S5074 is completed, the processing of the ith candidate short video may be ended, and then the same processing as that of the ith candidate short video may be performed for other candidate short videos, which is not described in detail.

S5075: the failure recording module 413 takes the ith candidate short video as an mth short video to be processed, and records failure information of the mth short video to be processed.

The specific content of the failure information of the mth short video to be processed is the same as that of the foregoing embodiment, and a repetitive description thereof will not be made here.

S5076: the updating module 416 obtains the mth short video to be processed based on the related information of the mth short video to be processed in the failure information of the mth short video to be processed; determining current analysis operators corresponding to the at least one failure analysis operator respectively based on related information of the at least one failure analysis operator in failure information of the mth short video to be processed; invoking the current analysis operators respectively corresponding to the at least one failure analysis operator, and processing the mth short video to be processed to obtain at least one updated quality feedback information of the mth short video to be processed; and obtaining one or more quality feedback information of the mth short video to be processed based on the quality feedback information of the mth short video to be processed, which is updated, and the quality feedback information of the mth short video to be processed corresponding to each successful analysis operator in the failure information of the mth short video to be processed.

S5077: the filtering module 415 determines one or more quality detection results of the mth short video to be processed based on the one or more quality feedback information of the mth short video to be processed; and determining the mth short video to be processed as one of the one or more target short videos under the condition that the preset condition is met based on one or more quality detection results of the mth short video to be processed.

After the aforementioned S5074 or S5077 is performed for each candidate short video, one or more target short videos may be obtained. After obtaining the one or more target short videos, in conjunction with fig. 4, the method may further include: saving the one or more target short videos in a second message queue 402; the one or more target short videos are extracted from the second message queue and stored in the storage module 418 in an inverted storage manner. Further, it may further include: the retrieval module 419, in response to receiving the retrieval request, selects a short video to be transmitted from the one or more target short videos based on the retrieval request; and generating and sending a reply message of the search request based on the related information of the short video to be sent.

The following describes in detail the beneficial effects of the video processing method and the video processing apparatus provided in this embodiment in combination with the related art:

firstly, short video is a main driving force for supporting the duration of internet users, and is used for supplementing video productivity, so that AIGC technology is gradually introduced to support intelligent production and further assist content creation. However, on the basis of model operator service, video processing is in an interactive state between people and machines, and full-flow automation of video processing is not supported. According to the scheme provided by the disclosure, the long video can be processed automatically in the whole flow, and the video processing efficiency is improved; and, because the target short video is screened from one or more candidate short videos according to the quality feedback information of the candidate short videos, the quality of the finally obtained target short video can be ensured.

In addition, in the related art, service scheduling control, error compensation and brushing, result quality checking and assembly are all required to be completed manually, so that the cost of manpower resources is high, the quality is affected, and the strategy iteration efficiency is improved. Specifically, in the related art, although the function of model operator service can be realized, the automatic scheduling of the whole process is lacked, and the process of the process needs to be manually interfered; model strategies and architecture engineering work are mutually kneaded, and special persons cannot do special work, so that the work efficiency is influenced to the maximum; and a flow scheduling scheme may be introduced in the related art, but the degree of plugin is low. According to the video processing method provided by the disclosure, the calling sequence of each operator can be arranged by pre-configuring the operator calling graph, so that the operator scheduling can be flexibly arranged or adjusted in parallel and serial modes, and in addition, when the processing of any operator fails, the failed operator can be automatically replaced, so that the improvement of the working efficiency is ensured.

Finally, when the source volume of the video material is large, other schemes in the related art can split the source into a plurality of fragments, and video processing is completed through the mapping relationship, so that the mapping relationship needs to be maintained. The method can enable one server to obtain the long videos to be processed in batches, so that the problem that the mapping relation cost is required to be maintained because the long videos to be processed are sent to different servers to be processed by different servers after being segmented can be avoided.

In the technical scheme of the disclosure, the acquisition, storage, application and the like of the related user personal information all conform to the regulations of related laws and regulations, and the public sequence is not violated.

According to embodiments of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.

Fig. 6 illustrates a schematic block diagram of an example electronic device 600 that may be used to implement embodiments of the present disclosure. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the disclosure described and/or claimed herein.

As shown in fig. 6, the electronic device 600 includes a computing unit 601 that can perform various appropriate actions and processes according to a computer program stored in a Read Only Memory (ROM) 602 or a computer program loaded from a storage unit 608 into a Random Access Memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the electronic device 600 can also be stored. The computing unit 601, ROM 602, and RAM 603 are connected to each other by a bus 604. An input/output (I/O) interface 605 is also connected to bus 604.

A number of components in the electronic device 600 are connected to the I/O interface 605, including: an input unit 606 such as a keyboard, mouse, etc.; an output unit 607 such as various types of displays, speakers, and the like; a storage unit 608, such as a magnetic disk, optical disk, or the like; and a communication unit 609 such as a network card, modem, wireless communication transceiver, etc. The communication unit 609 allows the electronic device 600 to exchange information/data with other devices through a computer network, such as the internet, and/or various telecommunication networks.

The computing unit 601 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of computing unit 601 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various computing units running machine learning model algorithms, a Digital Signal Processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 601 performs the various methods and processes described above. For example, in some embodiments, the various methods described above may be implemented as a computer software program tangibly embodied on a machine-readable medium, such as storage unit 608. In some embodiments, part or all of the computer program may be loaded and/or installed onto the electronic device 600 via the ROM 602 and/or the communication unit 609. When the computer program is loaded into RAM 603 and executed by the computing unit 601, one or more steps of the various methods described above may be performed. Alternatively, in other embodiments, the computing unit 601 may be configured to perform the various methods described above in any other suitable manner (e.g., by means of firmware).

Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuit systems, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), systems On Chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.

Program code for carrying out methods of the present disclosure may be written in any combination of one or more programming languages. These program code may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus such that the program code, when executed by the processor or controller, causes the functions/operations specified in the flowchart and/or block diagram to be implemented. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.

In the context of this disclosure, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. The machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and pointing device (e.g., a mouse or trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.

The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), and the internet.

The computer system may include a client and a server. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server may be a cloud server, a server of a distributed system, or a server incorporating a blockchain.

It should be appreciated that various forms of the flows shown above may be used to reorder, add, or delete steps. For example, the steps recited in the present disclosure may be performed in parallel, sequentially, or in a different order, provided that the desired results of the disclosed aspects are achieved, and are not limited herein.

The above detailed description should not be taken as limiting the scope of the present disclosure. It will be apparent to those skilled in the art that various modifications, combinations, sub-combinations and alternatives are possible, depending on design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present disclosure are intended to be included within the scope of the present disclosure.

Claims

1. A video processing method, comprising:

acquiring a long video to be processed;

under the condition that the long video to be processed is segmented to obtain one or more candidate short videos, acquiring a plurality of quality feedback information of each candidate short video in the one or more candidate short videos; screening one or more target short videos from the one or more candidate short videos based on the multiple quality feedback information of each candidate short video;

The obtaining the multiple quality feedback information of each candidate short video in the one or more candidate short videos includes: invoking a plurality of analysis operators to process an ith candidate short video in the one or more candidate short videos to obtain a plurality of quality feedback information of the ith candidate short video; i is a positive integer;

further comprises: determining a current long video to be processed based on related information of the long video to be processed in the segmentation failure information of the long video to be processed; based on failure information of a segmentation operator in the segmentation failure information of the long video to be processed, re-calling the segmentation operator to process the current long video to be processed, and obtaining updated segmentation processing information fed back by the segmentation operator; determining one or more updated candidate short videos based on a processing result contained in the updated segmentation processing information under the condition that the processing state contained in the updated segmentation processing information is that the segmentation processing is successful; acquiring a plurality of quality feedback information of each updated candidate short video in the one or more updated candidate short videos; and screening one or more target short videos from the one or more updated candidate short videos based on the multiple quality feedback information of each updated candidate short video.

2. The method of claim 1, further comprising:

calling a segmentation operator to process the long video to be processed, and obtaining segmentation processing information fed back by the segmentation operator; the segmentation operator is an operator positioned in a first calling sequence in a plurality of operators contained in an operator calling graph;

and determining the one or more candidate short videos based on the segmentation processing result contained in the segmentation processing information under the condition that the segmentation processing state contained in the segmentation processing information is that the segmentation processing is successful.

3. The method of claim 2, wherein the determining the one or more candidate short videos based on the segmentation process results contained in the segmentation process information comprises:

extracting one or more segmentation videos from the segmentation processing results contained in the segmentation processing information;

from the one or more sliced videos, the one or more candidate short videos within a video length threshold are determined.

4. The method of claim 2, further comprising:

recording segmentation failure information of the long video to be processed under the condition that the segmentation processing state contained in the segmentation processing information is segmentation processing failure; the segmentation failure information of the long video to be processed comprises: and the related information of the long video to be processed and the failure information of the segmentation operator.

5. The method of claim 2, wherein each analysis operator of the plurality of analysis operators is an operator of the operator call graph that is in a second call order, and different analysis operators of the plurality of analysis operators are different in function; and different quality feedback information is obtained by different analysis operators in the multiple quality feedback information of the ith candidate short video.

6. The method of claim 5, wherein the screening one or more target short videos from the one or more candidate short videos based on the plurality of quality feedback information for each candidate short video comprises:

determining a plurality of quality detection results of the ith candidate short video based on the plurality of quality feedback information of the ith candidate short video under the condition that the processing state of each analysis operator in the plurality of analysis operators is successfully processed;

and determining the ith candidate short video as one of the one or more target short videos under the condition that the preset condition is met based on a plurality of quality detection results of the ith candidate short video.

7. The method of claim 6, wherein the plurality of quality detection results comprises at least one of: definition scoring, whether mosaic exists or not, and correlation scoring;

the preset condition includes at least one of: the sharpness score is greater than a preset sharpness threshold, no mosaic is present, and the correlation score is greater than a preset correlation threshold.

8. The method of claim 7, further comprising:

under the condition that the processing state of at least one analysis operator in the plurality of analysis operators is determined to be processing failure based on the multiple quality feedback information of the ith candidate short video, taking the ith candidate short video as an mth short video to be processed, and recording failure information of the mth short video to be processed; m is a positive integer;

9. The method of claim 8, further comprising:

acquiring an mth short video to be processed based on related information of the mth short video to be processed in failure information of the mth short video to be processed;

determining current analysis operators corresponding to the at least one failure analysis operator respectively based on related information of the at least one failure analysis operator in failure information of the mth short video to be processed;

invoking the current analysis operators respectively corresponding to the at least one failure analysis operator, and processing the mth short video to be processed to obtain at least one updated quality feedback information of the mth short video to be processed;

obtaining a plurality of quality feedback information of the mth short video to be processed based on the quality feedback information of the mth short video to be processed, which is updated, and the quality feedback information of the mth short video to be processed corresponding to each successful analysis operator in the failure information of the mth short video to be processed;

determining a plurality of quality detection results of the mth short video to be processed based on the quality feedback information of the mth short video to be processed;

And determining the mth short video to be processed as one of the one or more target short videos under the condition that the preset condition is met based on a plurality of quality detection results of the mth short video to be processed.

10. The method of any of claims 1-9, wherein after said screening one or more target short videos from the one or more candidate short videos, further comprising:

responding to a received search request, and selecting a short video to be sent from the one or more target short videos based on the search request;

and generating and sending a reply message of the search request based on the related information of the short video to be sent.

11. A video processing apparatus comprising:

the acquisition module is used for acquiring the long video to be processed;

the analysis module is used for acquiring a plurality of quality feedback information of each candidate short video in the one or more candidate short videos under the condition that the long video to be processed is segmented to obtain the one or more candidate short videos;

the filtering module is used for filtering one or more target short videos from the one or more candidate short videos based on the multiple quality feedback information of each candidate short video;

The analysis module is used for calling a plurality of analysis operators to process an ith candidate short video in the one or more candidate short videos so as to obtain a plurality of quality feedback information of the ith candidate short video; i is a positive integer;

further comprises: the updating module is used for determining the current long video to be processed based on the related information of the long video to be processed in the segmentation failure information of the long video to be processed; based on failure information of a segmentation operator in the segmentation failure information of the long video to be processed, re-calling the segmentation operator to process the current long video to be processed, and obtaining updated segmentation processing information fed back by the segmentation operator; determining one or more updated candidate short videos based on a processing result contained in the updated segmentation processing information under the condition that the processing state contained in the updated segmentation processing information is successful in segmentation processing, and acquiring a plurality of quality feedback information of each updated candidate short video in the one or more updated candidate short videos; the filtering module is configured to filter one or more target short videos from the one or more updated candidate short videos based on the multiple quality feedback information of each updated candidate short video.

12. The apparatus of claim 11, further comprising:

the segmentation processing module is used for calling a segmentation operator to process the long video to be processed to obtain segmentation processing information fed back by the segmentation operator; the segmentation operator is an operator positioned in a first calling sequence in a plurality of operators contained in an operator calling graph; and determining the one or more candidate short videos based on the segmentation processing result contained in the segmentation processing information under the condition that the segmentation processing state contained in the segmentation processing information is that the segmentation processing is successful.

13. The apparatus of claim 12, wherein the segmentation processing module is configured to extract one or more segmentation videos from the segmentation processing result included in the segmentation processing information; from the one or more sliced videos, the one or more candidate short videos within a video length threshold are determined.

14. The apparatus of claim 12, further comprising:

the failure recording module is used for recording the segmentation failure information of the long video to be processed under the condition that the segmentation processing state contained in the segmentation processing information is the segmentation processing failure; the segmentation failure information of the long video to be processed comprises: and the related information of the long video to be processed and the failure information of the segmentation operator.

15. The apparatus of claim 12, wherein each analysis operator of the plurality of analysis operators is an operator of the operator call graph that is in a second call order, and different analysis operators of the plurality of analysis operators are different in function; and different quality feedback information is obtained by different analysis operators in the multiple quality feedback information of the ith candidate short video.

16. The apparatus of claim 15, wherein the filtering module is configured to determine, based on the multiple quality feedback information of the i-th candidate short video, multiple quality detection results of the i-th candidate short video if it is determined that the processing state of each of the multiple analysis operators is successful; and determining the ith candidate short video as one of the one or more target short videos under the condition that the preset condition is met based on a plurality of quality detection results of the ith candidate short video.

17. The apparatus of claim 16, wherein the plurality of quality detection results comprises at least one of: definition scoring, whether mosaic exists or not, and correlation scoring;

18. The apparatus of claim 17, further comprising:

the failure recording module is used for taking the ith candidate short video as an mth short video to be processed and recording failure information of the mth short video to be processed under the condition that the processing state of at least one analysis operator in the plurality of analysis operators is determined to be processing failure based on the quality feedback information of the ith candidate short video; m is a positive integer;

19. The apparatus of claim 18, further comprising:

The updating module is used for acquiring the mth short video to be processed based on the related information of the mth short video to be processed in the failure information of the mth short video to be processed; determining current analysis operators corresponding to the at least one failure analysis operator respectively based on related information of the at least one failure analysis operator in failure information of the mth short video to be processed; invoking the current analysis operators respectively corresponding to the at least one failure analysis operator, and processing the mth short video to be processed to obtain at least one updated quality feedback information of the mth short video to be processed; obtaining a plurality of quality feedback information of the mth short video to be processed based on the quality feedback information of the mth short video to be processed, which is updated, and the quality feedback information of the mth short video to be processed corresponding to each successful analysis operator in the failure information of the mth short video to be processed;

the filtering module is used for determining a plurality of quality detection results of the mth short video to be processed based on the quality feedback information of the mth short video to be processed; and determining the mth short video to be processed as one of the one or more target short videos under the condition that the preset condition is met based on a plurality of quality detection results of the mth short video to be processed.

20. The apparatus of any of claims 11-19, further comprising:

the storage module is used for storing the one or more target short videos;

the searching module is used for responding to a received searching request, and selecting a short video to be sent from the one or more target short videos based on the searching request; and generating and sending a reply message of the search request based on the related information of the short video to be sent.

21. An electronic device, comprising:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein,,

the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-10.

22. A non-transitory computer readable storage medium storing computer instructions for causing the computer to perform the method of any one of claims 1-10.