WO2020052084A1 - 视频封面选择方法、装置和计算机可读存储介质 - Google Patents

视频封面选择方法、装置和计算机可读存储介质 Download PDF

Info

Publication number
WO2020052084A1
WO2020052084A1 PCT/CN2018/117713 CN2018117713W WO2020052084A1 WO 2020052084 A1 WO2020052084 A1 WO 2020052084A1 CN 2018117713 W CN2018117713 W CN 2018117713W WO 2020052084 A1 WO2020052084 A1 WO 2020052084A1
Authority
WO
WIPO (PCT)
Prior art keywords
image frame
video
image
evaluated
cover
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2018/117713
Other languages
English (en)
French (fr)
Inventor
黄凯
王长虎
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing ByteDance Network Technology Co Ltd
Original Assignee
Beijing ByteDance Network Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing ByteDance Network Technology Co Ltd filed Critical Beijing ByteDance Network Technology Co Ltd
Publication of WO2020052084A1 publication Critical patent/WO2020052084A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10016—Video; Image sequence
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30168—Image quality inspection

Definitions

  • the present disclosure relates to the technical field of information processing, and in particular, to a video cover selection method, device, and computer-readable storage medium.
  • the cover of a video is generally automatically generated randomly or specified by the author.
  • the selected cover of the video is not targeted and is not conducive to the promotion of the video.
  • the technical problem solved by the present disclosure is to provide a video cover selection method to at least partially solve the technical problem that the selected video cover is not targeted and is not conducive to the promotion of the video.
  • a video cover selection device, a video cover selection hardware device, a computer-readable storage medium, and a video cover selection terminal are also provided.
  • a video cover selection method includes:
  • an image frame that satisfies a preset cover condition is selected from the at least one image frame to be evaluated as a cover of the video to be processed.
  • the step of evaluating at least one image frame to be evaluated extracted from the video to be processed includes:
  • the step of selecting an image frame satisfying a preset cover condition from the at least one image frame to be evaluated as a cover of the video to be processed according to the evaluation result includes:
  • an image frame that satisfies a preset cover condition is selected from the image frames closest to the cluster center as the cover of the video to be processed.
  • the step of evaluating at least one image frame to be evaluated extracted from the video to be processed includes:
  • the click rate prediction is performed on the at least one image frame to be evaluated, and the evaluation result is the click rate.
  • the step of evaluating at least one image frame to be evaluated extracted from the video to be processed includes:
  • the evaluation result is a picture quality level and a click rate
  • the picture quality level and the click rate are combined, and the evaluation result is recalculated.
  • the step of evaluating at least one image frame to be evaluated extracted from the video to be processed includes:
  • the at least one image frame to be evaluated is input into an image evaluation model trained in advance for evaluation, and an output result of the image evaluation model is used as the evaluation result.
  • the image evaluation model includes: an image quality evaluation model and / or a click rate prediction model, the image quality evaluation model is used to output a picture quality level, and the click rate prediction model is used to output a click rate.
  • the method further includes:
  • Pre-processing at least one image frame extracted from the video to be processed, and selecting an image frame that meets a preset standard as the image frame to be evaluated.
  • a video cover selection device includes:
  • An image evaluation module configured to evaluate at least one image frame to be evaluated extracted from the video to be processed
  • a cover selection module is configured to select an image frame that satisfies a preset cover condition from the at least one image frame to be evaluated as a cover of the video to be processed according to the evaluation result.
  • the image evaluation module is specifically configured to perform cluster analysis on the at least one image frame to be evaluated to obtain at least one cluster center; and for each cluster center, perform an image closest to the cluster center. Frame for evaluation;
  • the cover selection module is specifically configured to: according to the evaluation result, select an image frame that meets a preset cover condition from the image frames closest to the cluster center as the cover of the video to be processed.
  • the image evaluation module is specifically configured to perform picture quality evaluation on the at least one image frame to be evaluated, and the evaluation result is a picture quality level; and / or, the at least one frame to be evaluated image Frames are click-through rate predicted, and the evaluation result is click-through rate.
  • the image evaluation module is specifically configured to: if the evaluation result is a picture quality level and a click rate, merge the picture quality level and the click rate, and recalculate the evaluation result.
  • the image evaluation module is specifically configured to: input the at least one image frame to be evaluated into an image evaluation model trained in advance for evaluation, and use an output result of the image evaluation model as the evaluation result.
  • the image evaluation model includes: an image quality evaluation model and / or a click rate prediction model, the image quality evaluation model is used to output a picture quality level, and the click rate prediction model is used to output a click rate.
  • the device further includes:
  • the pre-processing module is configured to pre-process at least one image frame extracted from the video to be processed, and select an image frame that meets a preset standard as the image frame to be evaluated.
  • a video cover selection hardware device includes:
  • Memory for storing non-transitory computer-readable instructions
  • a processor configured to run the computer-readable instructions, so that the processor, when executed, implements the steps described in any one of the foregoing technical solutions of the video cover selection method.
  • a computer-readable storage medium is configured to store non-transitory computer-readable instructions, and when the non-transitory computer-readable instructions are executed by a computer, cause the computer to execute any of the technical solutions of the video cover selection method described above. The steps described.
  • a video cover selection terminal includes any of the above video cover selection devices.
  • Embodiments of the present disclosure provide a video cover selection method, a video cover selection device, a video cover selection hardware device, a computer-readable storage medium, and a video cover selection terminal.
  • the video cover selection method includes evaluating at least one image frame to be evaluated extracted from the video to be processed, and selecting an image frame that satisfies a preset cover condition from the at least one image frame to be evaluated according to the evaluation result.
  • the embodiment of the present disclosure first evaluates at least one image frame to be evaluated extracted from a video to be processed, and selects, from the at least one image frame to be evaluated, an image frame that satisfies a preset cover condition from the at least one image frame to be evaluated as the image frame.
  • the cover of the video to be processed The cover selected in this way is targeted, which is conducive to the promotion of the video.
  • FIG. 1a is a schematic flowchart of a video cover selection method according to an embodiment of the present disclosure
  • FIG. 1b is a schematic flowchart of a video cover selection method according to another embodiment of the present disclosure.
  • 1c is a schematic flowchart of a video cover selection method according to another embodiment of the present disclosure.
  • FIG. 1d is a schematic flowchart of a video cover selection method according to another embodiment of the present disclosure.
  • FIG. 2a is a schematic structural diagram of a device for selecting a video cover according to an embodiment of the present disclosure
  • FIG. 3 is a schematic structural diagram of a video cover selection hardware device according to an embodiment of the present disclosure.
  • FIG. 4 is a schematic structural diagram of a computer-readable storage medium according to an embodiment of the present disclosure.
  • FIG. 5 is a schematic structural diagram of a video cover selection terminal according to an embodiment of the present disclosure.
  • the embodiment of the present disclosure provides a video cover selection method.
  • the video cover selection method mainly includes the following steps S1 to S2. among them:
  • Step S1 Evaluate at least one image frame to be evaluated extracted from the video to be processed.
  • several frames of images from the video to be processed may be extracted as the image frames to be evaluated, or all image frames in the video to be processed may be taken as the image frames to be evaluated.
  • Step S2 According to the evaluation result, an image frame satisfying a preset cover condition is selected from at least one image frame to be evaluated as a cover of the video to be processed.
  • the evaluation results include, but are not limited to, image quality levels and / or click-through rates.
  • the preset cover condition may be, but is not limited to, an image quality level meeting a preset requirement and / or an image click rate exceeding a preset click rate.
  • this step when the evaluation result is a picture quality level, through this step, an image frame with an image quality level that meets a preset requirement, that is, a relatively high quality level, can be selected as the cover of the video to be processed, thereby helping to attract users to watch.
  • a preset requirement that is, a relatively high quality level
  • this step can select the image frame with the click-through rate exceeding the preset click-through rate, that is, the image frame with the higher click-through rate, as the cover of the video to be processed. Users are interested in this image frame, so it is also conducive to attract users to watch and facilitate the promotion of the video.
  • At least one image frame to be evaluated extracted from the video to be processed is evaluated, and an image frame that satisfies a preset cover condition is selected from the at least one image frame to be evaluated as the cover of the video to be processed according to the evaluation result.
  • the cover selected in this way is targeted, which is conducive to the promotion of the video.
  • step S1 includes:
  • S11 Perform cluster analysis on at least one image frame to be evaluated extracted from the video to be processed to obtain at least one cluster center.
  • the applicable clustering algorithms include, but are not limited to, any of the following: K-Means clustering, mean-shift clustering, density-based clustering methods, clustering clustering, graph group detection clustering, etc. .
  • the nearest several samples of the class center are used as a class to obtain at least one type of image set.
  • the Euclidean distance and the cosine distance can be used to represent the distance between the image frame and the cluster center.
  • step S2 specifically includes:
  • an image frame that satisfies a preset cover condition is selected from the image frames closest to the cluster center as the cover of the video to be processed.
  • the evaluation results include, but are not limited to, image quality levels and / or click-through rates.
  • At least one image frame to be evaluated is subjected to cluster analysis to obtain at least one cluster center.
  • cluster analysis to obtain at least one cluster center.
  • an image frame closest to the cluster center is evaluated, and according to the evaluation result, the image is clustered from the distance.
  • the image frame that satisfies the preset cover condition is selected as the cover of the video to be processed from the nearest image frame of the center, so that the selected cover is targeted, which is conducive to the promotion of the video.
  • step S1 specifically includes:
  • the click rate prediction is performed on at least one image frame to be evaluated extracted from the video to be processed, and the evaluation result is the click rate.
  • the picture quality level is divided into four levels, for example, A, B, C, and D, or determined based on the image scoring results, with 0-30 as a level, 30-60 as a level, and 60-80 as a level. Level, with 80-100 as a level.
  • the standards for different level divisions have different corresponding cover conditions. For example, when divided into four levels: A, B, C, and D, the image frame can meet the level A. Set a preset cover condition. When determining based on the scoring result, you can set the image frame score to 80-100 as the preset cover condition.
  • the click rate of a frame image is the percentage of clicks on the frame image to the total clicks on all frames in the video.
  • step S1 specifically includes:
  • the evaluation result is an image quality level and a click rate
  • the image quality level and the click rate are combined, and the evaluation result is recalculated.
  • the weighted sum of the picture quality level and the click rate can be calculated according to the weight calculation method, and the weighted sum can be used as the evaluation result.
  • At least one image frame to be evaluated is subjected to picture quality evaluation and / or click-through rate prediction, and according to the image quality level and / or click-through rate, a picture that meets a preset cover condition is selected from the at least one image-to-be evaluated frame.
  • the image frame is used as the cover of the video to be processed.
  • the cover selected in this way is targeted to the promotion of the video.
  • step S1 specifically includes:
  • At least one frame of the image to be evaluated extracted from the video to be processed is input to an image evaluation model trained in advance for evaluation, and the output result of the image evaluation model is used as the evaluation result.
  • the image evaluation model includes: an image quality evaluation model and / or a click rate prediction model, the image quality evaluation model is used to output a picture quality level, and the click rate prediction model is used to output a click rate.
  • the training of the image quality assessment model includes the following steps: pre-statistically calculate the quality levels of the images in each frame of the video, use the images of known image quality levels as training samples, and label the training samples according to different levels, and then use
  • the deep learning classification algorithm trains and learns the labeled training samples to obtain an image quality evaluation model.
  • the click-through rate prediction model the click-through rate of each frame of images in the video is counted in advance, and the images with known click-through rates are used as training samples.
  • the training samples are labeled according to different click-through rates, and then the deep learning classification algorithm is used to train the labeled training.
  • the samples are trained and learned to obtain a click rate prediction model.
  • the deep learning classification algorithms include, but are not limited to, any of the following: Naive Bayes algorithm, artificial neural network algorithm, genetic algorithm, K-Nearest Neighbor (KNN) classification algorithm, clustering algorithm, and the like.
  • KNN K-Nearest Neighbor
  • At least one image frame to be evaluated is input into an image evaluation model trained in advance for evaluation, and an output result of the image evaluation model is used as an evaluation result, and then based on the output result, at least one image frame to be evaluated is selected.
  • the image frames that meet the preset cover conditions are used as the cover of the video to be processed. The cover selected in this way is targeted, which is conducive to the promotion of the video.
  • the method in this embodiment further includes:
  • Preprocess at least one image frame extracted from the video to be processed, and filter out image frames that meet the preset criteria as the image frames to be evaluated.
  • the screening criterion may be image sharpness.
  • low-quality images such as images with black edges or blurring, can be further filtered out.
  • the following is a device embodiment of the present disclosure.
  • the device embodiment of the present disclosure can be used to perform the steps implemented by the method embodiments of the present disclosure.
  • Only parts related to the embodiments of the present disclosure are shown. Specific technical details are not disclosed. Reference is made to the method embodiments of the present disclosure.
  • an embodiment of the present disclosure provides a video cover selection device.
  • the device may perform the steps in the foregoing embodiment of the method for selecting a video cover.
  • the device mainly includes: an image evaluation module 21 and a cover selection module 22; wherein, the image evaluation module 21 is configured to evaluate at least one image frame to be evaluated extracted from a video to be processed; a cover selection module 22 is used to select an image frame that satisfies a preset cover condition from at least one image frame to be evaluated as a cover of the video to be processed according to the evaluation result.
  • the image evaluation module 21 may extract several frames from the video to be processed as the image frames to be evaluated, or may use all the image frames in the video to be processed as the image frames to be evaluated.
  • the evaluation results include, but are not limited to, image quality levels and / or click-through rates.
  • the preset cover condition may be, but is not limited to, an image quality level meeting a preset requirement and / or an image click rate exceeding a preset click rate.
  • the image evaluation module 21 evaluates at least one image frame to be evaluated extracted from the video to be processed
  • the cover selection module 22 selects at least one image frame to be evaluated from the at least one image frame to be evaluated according to the evaluation result.
  • the image frame is used as the cover of the video to be processed, so the selected cover is targeted, which is conducive to the promotion of the video.
  • the image evaluation module 21 is specifically configured to: perform cluster analysis on at least one image frame to be evaluated to obtain at least one cluster center; and for each cluster center, The image frame closest to the cluster center is evaluated;
  • the cover selection module 22 is specifically configured to: according to the evaluation result, select an image frame that satisfies a preset cover condition from the image frames closest to the cluster center as the cover of the video to be processed.
  • the clustering algorithm that can be used by the image evaluation module 21 includes, but is not limited to, any of the following: K-Means clustering, mean-shift clustering, density-based clustering methods, agglomeration hierarchical clustering, and graph community Detect clusters, etc.
  • the nearest several samples of the class center are used as a class to obtain at least one type of image set.
  • the Euclidean distance and the cosine distance can be used to represent the distance between the image frame and the cluster center.
  • the evaluation results include, but are not limited to, image quality levels and / or click-through rates.
  • the image evaluation module 21 performs cluster analysis on at least one image frame to be evaluated to obtain at least one cluster center. For each cluster center, the image frame closest to the cluster center is evaluated, and a cover selection module is used. 22 According to the evaluation result, an image frame that satisfies the preset cover condition is selected from the image frames closest to the clustering center as the cover of the video to be processed. The selected cover is targeted and thus facilitates the promotion of the video.
  • the image evaluation module 21 is specifically configured to perform picture quality evaluation on at least one image frame to be evaluated, and the evaluation result is a picture quality level; and / or, for at least one A click-through rate prediction is performed for each image frame to be evaluated, and the evaluation result is the click-through rate.
  • the picture quality level is divided into four levels, for example, A, B, C, and D, or determined based on the image scoring results, with 0-30 as a level, 30-60 as a level, and 60-80 as a level. Level, with 80-100 as a level.
  • the standards for different level divisions have different corresponding cover conditions. For example, when divided into four levels: A, B, C, and D, the image frame can meet the level A. Set a preset cover condition. When determining based on the scoring result, you can set the image frame score to 80-100 as the preset cover condition.
  • the click rate of a frame image is the percentage of clicks on the frame image to the total clicks on all frames in the video.
  • the image evaluation module 21 is specifically configured to: if the evaluation result is an image quality level and a click rate, merge the image quality level and the click rate, and recalculate the evaluation result.
  • the image evaluation module 21 may calculate a weighted sum of a picture quality level and a click rate according to a weight calculation method, and use the weighted sum as an evaluation result.
  • the image evaluation module 21 performs picture quality evaluation and / or click-through rate prediction on at least one image frame to be evaluated
  • the cover selection module 22 uses the picture quality level and / or click-through rate to select at least one frame from the image to be evaluated.
  • An image frame that satisfies a preset cover condition is selected as the cover of the video to be processed, and the selected cover is targeted, which is beneficial to the promotion of the video.
  • the image evaluation module 21 is specifically configured to: input at least one image frame to be evaluated into an image evaluation model trained in advance for evaluation, and output the output of the image evaluation model As a result of the evaluation.
  • the image evaluation model includes: an image quality evaluation model and / or a click rate prediction model, the image quality evaluation model is used to output a picture quality level, and the click rate prediction model is used to output a click rate.
  • the training of the image quality assessment model includes the following steps: pre-statistically calculate the quality levels of the images in each frame of the video, use the images of known image quality levels as training samples, and label the training samples according to different levels, and then use
  • the deep learning classification algorithm trains and learns the labeled training samples to obtain an image quality evaluation model.
  • the click-through rate prediction model the click-through rate of each frame of images in the video is counted in advance, and the images with known click-through rates are used as training samples.
  • the training samples are labeled according to different click-through rates, and then the deep learning classification algorithm is used to train the labeled training.
  • the samples are trained and learned to obtain a click rate prediction model.
  • the deep learning classification algorithms include, but are not limited to, any of the following: Naive Bayes algorithm, artificial neural network algorithm, genetic algorithm, K-Nearest Neighbor (KNN) classification algorithm, clustering algorithm, and the like.
  • KNN K-Nearest Neighbor
  • the image evaluation module 21 inputs at least one image frame to be evaluated into a pre-trained image evaluation model for evaluation, and uses the output result of the image evaluation model as the evaluation result.
  • the cover selection module 22 uses the output result to An image frame that satisfies a preset cover condition is selected as the cover of the video to be processed from at least one image frame to be evaluated, so that the selected cover is targeted, which is beneficial to the promotion of the video.
  • the device further includes: a pre-processing module 23; wherein the pre-processing module 23 is configured to pre-process at least one image frame extracted from the video to be processed, and filter out those that meet a preset standard.
  • the image frame is used as the image frame to be evaluated.
  • the screening criterion may be image sharpness.
  • low-quality images such as images with black edges or blurring, can be further filtered out.
  • FIG. 3 is a hardware block diagram illustrating a video cover selection hardware device according to an embodiment of the present disclosure.
  • the video cover selection hardware device 30 according to an embodiment of the present disclosure includes a memory 31 and a processor 32.
  • the memory 31 is configured to store non-transitory computer-readable instructions.
  • the memory 31 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory.
  • the volatile memory may include, for example, a random access memory (RAM) and / or a cache memory.
  • the non-volatile memory may include, for example, a read-only memory (ROM), a hard disk, a flash memory, and the like.
  • the processor 32 may be a central processing unit (CPU) or other form of processing unit having data processing capabilities and / or instruction execution capabilities, and may control other components in the video cover selection hardware device 30 to perform a desired function.
  • the processor 32 is configured to run the computer-readable instructions stored in the memory 31, so that the video cover selection hardware device 30 executes the foregoing video cover selection method of the embodiments of the present disclosure. All or part of the steps.
  • this embodiment may also include well-known structures such as a communication bus and an interface. These well-known structures should also be included in the protection scope of the present disclosure. within.
  • FIG. 4 is a schematic diagram illustrating a computer-readable storage medium according to an embodiment of the present disclosure.
  • a computer-readable storage medium 40 according to an embodiment of the present disclosure stores non-transitory computer-readable instructions 41 thereon.
  • the non-transitory computer-readable instruction 41 is executed by a processor, all or part of the steps of the method for comparing video features of the foregoing embodiments of the present disclosure are performed.
  • the computer-readable storage medium 40 includes, but is not limited to, optical storage media (for example, CD-ROM and DVD), magneto-optical storage media (for example, MO), magnetic storage media (for example, magnetic tape or mobile hard disk), Non-volatile memory rewritable media (for example: memory card) and media with built-in ROM (for example: ROM box).
  • optical storage media for example, CD-ROM and DVD
  • magneto-optical storage media for example, MO
  • magnetic storage media for example, magnetic tape or mobile hard disk
  • Non-volatile memory rewritable media for example: memory card
  • media with built-in ROM for example: ROM box
  • FIG. 5 is a schematic diagram illustrating a hardware structure of a terminal according to an embodiment of the present disclosure. As shown in FIG. 5, the video cover selection terminal 50 includes the foregoing video cover selection device embodiment.
  • the terminal may be implemented in various forms, and the terminal in the present disclosure may include, but is not limited to, such as a mobile phone, a smart phone, a notebook computer, a digital broadcast receiver, a PDA (personal digital assistant), a PAD (tablet computer), a PMP ( Portable multimedia players), navigation devices, on-board terminals, on-board display terminals, on-board electronic rear-view mirrors, and other mobile terminals, and fixed terminals such as digital TVs, desktop computers, and the like.
  • a mobile phone such as a mobile phone, a smart phone, a notebook computer, a digital broadcast receiver, a PDA (personal digital assistant), a PAD (tablet computer), a PMP ( Portable multimedia players), navigation devices, on-board terminals, on-board display terminals, on-board electronic rear-view mirrors, and other mobile terminals, and fixed terminals such as digital TVs, desktop computers, and the like.
  • PDA personal digital assistant
  • PAD tablet computer
  • PMP Portable multimedia players
  • navigation devices
  • the terminal may further include other components.
  • the video cover selection terminal 50 may include a power supply unit 51, a wireless communication unit 52, an A / V (audio / video) input unit 53, a user input unit 54, a sensing unit 55, an interface unit 56, and a control unit.
  • FIG. 5 illustrates a terminal having various components, but it should be understood that it is not required to implement all the illustrated components, and more or fewer components may be implemented instead.
  • the wireless communication unit 52 allows radio communication between the terminal 50 and a wireless communication system or network.
  • the A / V input unit 53 is used to receive audio or video signals.
  • the user input unit 54 may generate key input data according to a command input by the user to control various operations of the terminal.
  • the sensing unit 55 detects the current state of the terminal 50, the position of the terminal 50, the presence or absence of a user's touch input to the terminal 50, the orientation of the terminal 50, the acceleration or deceleration movement and direction of the terminal 50, and the like, and generates a signal for controlling the terminal 50 commands or signals for operation.
  • the interface unit 56 functions as an interface through which at least one external device can be connected to the terminal 50.
  • the output unit 58 is configured to provide an output signal in a visual, audio, and / or tactile manner.
  • the memory 59 may store software programs and the like for processing and control operations performed by the controller 55, or may temporarily store data that has been output or is to be output.
  • the memory 59 may include at least one type of storage medium.
  • the terminal 50 may cooperate with a network storage device that performs a storage function of the memory 59 through a network connection.
  • the controller 57 generally controls the overall operation of the terminal.
  • the controller 57 may include a multimedia module for reproducing or playing back multimedia data.
  • the controller 57 may perform a pattern recognition process to recognize a handwriting input or a picture drawing input performed on the touch screen as characters or images.
  • the power supply unit 51 receives external power or internal power under the control of the controller 57 and provides appropriate power required to operate each element and component.
  • Various embodiments of the video feature comparison method proposed by the present disclosure may be implemented in a computer-readable medium using, for example, computer software, hardware, or any combination thereof.
  • various embodiments of the video feature comparison method proposed in the present disclosure can be implemented by using an application-specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), and a programmable logic device. (PLD), field programmable gate array (FPGA), processor, controller, microcontroller, microprocessor, electronic unit designed to perform the functions described herein, and in some cases implemented
  • ASIC application-specific integrated circuit
  • DSP digital signal processor
  • DSPD digital signal processing device
  • PLD programmable logic device
  • FPGA field programmable gate array
  • processor controller
  • microcontroller microprocessor
  • electronic unit designed to perform the functions described herein and in some cases implemented
  • Various embodiments of the video feature comparison method proposed in the present disclosure may be implemented in the controller 57.
  • various embodiments of the video feature comparison method proposed by the present disclosure can be implemented with a separate software module that allows at least one function or operation to be performed.
  • the software codes may be implemented by a software application (or program) written in any suitable programming language, and the software codes may be stored in the memory 59 and executed by the controller 57.
  • an "or” used in an enumeration of items beginning with “at least one” indicates a separate enumeration such that, for example, an "at least one of A, B, or C” enumeration means A or B or C, or AB or AC or BC, or ABC (ie A and B and C).
  • the word "exemplary” does not mean that the described example is preferred or better than other examples.
  • each component or each step can be disassembled and / or recombined.

Landscapes

  • Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Quality & Reliability (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Television Signal Processing For Recording (AREA)
  • Testing, Inspecting, Measuring Of Stereoscopic Televisions And Televisions (AREA)
  • Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)

Abstract

本公开公开一种视频封面选择方法、视频封面选择装置、视频封面选择硬件装置和计算机可读存储介质。其中，该视频封面选择方法包括对从待处理视频中抽取的至少一帧待评估图像帧进行评估；根据评估结果，从所述至少一帧待评估图像帧中选取满足预设封面条件的图像帧作为所述待处理视频的封面。本公开实施例首先对从待处理视频中抽取的至少一帧待评估图像帧进行评估，根据评估结果，从所述至少一帧待评估图像帧中选取满足预设封面条件的图像帧作为所述待处理视频的封面，这样选出的封面具有针对性，从而有利于视频的推广。

Description

视频封面选择方法、装置和计算机可读存储介质
交叉引用
本公开引用于2018年09月13日递交的名称为“视频封面选择方法、装置和计算机可读存储介质”的、申请号为201811069772.5的中国专利申请，其通过引用被全部并入本申请。
技术领域
本公开涉及一种信息处理技术领域，特别是涉及一种视频封面选择方法、装置和计算机可读存储介质。
背景技术
近年来，随着多媒体技术和计算机网络的飞速发展，数字视频的容量正以惊人的速度增长。怎么从众多的视频中脱颖而出，引起人们的兴趣，成为视频制作者研究的热点。
现有技术中，视频的封面一般自动随机生成或者由作者自己指定，这样选出来的视频封面并没有针对性，不利于视频的推广。
发明内容
本公开解决的技术问题是提供一种视频封面选择方法，以至少部分地解决选出来的视频封面没有针对性，不利于视频的推广的技术问题。此外，还提供一种视频封面选择装置、视频封面选择硬件装置、计算机可读存储介质和视频封面选择终端。
为了实现上述目的，根据本公开的一个方面，提供以下技术方案：
一种视频封面选择方法，包括：
对从待处理视频中抽取的至少一帧待评估图像帧进行评估；
根据评估结果，从所述至少一帧待评估图像帧中选取满足预设封面条件的图像帧作为所述待处理视频的封面。
进一步的，所述对从待处理视频中抽取的至少一帧待评估图像帧进行评估的步骤，包括：
对所述至少一帧待评估图像帧进行聚类分析，得到至少一个聚类中心；
针对各聚类中心，对距离所述聚类中心最近的图像帧进行评估；
所述根据评估结果，从所述至少一帧待评估图像帧中选取满足预设封面条件的图像帧作为所述待处理视频的封面的步骤，包括：
根据评估结果，从所述距离所述聚类中心最近的图像帧中选取满足预设封面条件的图像帧作为所述待处理视频的封面。
进一步的，所述对从待处理视频中抽取的至少一帧待评估图像帧进行评估的步骤，包括：
对所述至少一帧待评估图像帧分别进行图片质量评估，所述评估结果为图片质量等级；和/或，
对所述至少一帧待评估图像帧分别进行点击率预测，所述评估结果为点击率。
进一步的，所述对从待处理视频中抽取的至少一帧待评估图像帧进行评估的步骤，包括：
若所述评估结果为图片质量等级和点击率，则融合所述图片质量等级和点击率，重新计算评估结果。
进一步的，所述对从待处理视频中抽取的至少一帧待评估图像帧进行评估的步骤，包括：
将所述至少一帧待评估图像帧输入预先训练得到的图像评估模型中进行评估，将所述图像评估模型的输出结果作为所述评估结果。
进一步的，所述图像评估模型包括：图像质量评估模型和/或点击率预测模型，所述图像质量评估模型用于输出图片质量等级，所述点击率预测模型用于输出点击率。
进一步的，所述方法还包括：
对从所述待处理视频中抽取的至少一帧图像帧进行预处理，筛选出符合预设标准的图像帧作为所述待评估图像帧。
为了实现上述目的，根据本公开的又一个方面，还提供以下技术方案：
一种视频封面选择装置，包括：
图像评估模块，用于对从待处理视频中抽取的至少一帧待评估图像帧进行评估；
封面选取模块，用于根据评估结果，从所述至少一帧待评估图像帧中选取满足预设封面条件的图像帧作为所述待处理视频的封面。
进一步的，所述图像评估模块具体用于：对所述至少一帧待评估图像帧进行聚类分析，得到至少一个聚类中心；针对各聚类中心，对距离所述聚类中心最近的图像帧进行评估；
所述封面选取模块具体用于：根据评估结果，从所述距离所述聚类中心最近的图像帧中选取满足预设封面条件的图像帧作为所述待处理视频的封面。
进一步的，所述图像评估模块具体用于：对所述至少一帧待评估图像帧分别进行图片质量评估，所述评估结果为图片质量等级；和/或，对所述至少一帧待评估图像帧分别进行点击率预测，所述评估结果为点击率。
进一步的，所述图像评估模块具体用于：若所述评估结果为图片质量等级和点击率，则融合所述图片质量等级和点击率，重新计算评估结果。
进一步的，所述图像评估模块具体用于：将所述至少一帧待评估图像帧输入预先训练得到的图像评估模型中进行评估，将所述图像评估模型的输出结果作为所述评估结果。
进一步的，所述图像评估模型包括：图像质量评估模型和/或点击率预测模型，所述图像质量评估模型用于输出图片质量等级，所述点击率预测模型用于输出点击率。
进一步的，所述装置还包括：
预处理模块，用于对从所述待处理视频中抽取的至少一帧图像帧进行预处理，筛选出符合预设标准的图像帧作为所述待评估图像帧。
为了实现上述目的，根据本公开的又一个方面，还提供以下技术方案：
一种视频封面选择硬件装置，包括：
存储器，用于存储非暂时性计算机可读指令；以及
处理器，用于运行所述计算机可读指令，使得所述处理器执行时实现上述任一视频封面选择方法技术方案中所述的步骤。
为了实现上述目的，根据本公开的又一个方面，还提供以下技术方案：
一种计算机可读存储介质，用于存储非暂时性计算机可读指令，当所述非暂时性计算机可读指令由计算机执行时，使得所述计算机执行上述任一视频封面选择方法技术方案中所述的步骤。
为了实现上述目的，根据本公开的又一个方面，还提供以下技术方案：
一种视频封面选择终端，包括上述任一视频封面选择装置。
本公开实施例提供一种视频封面选择方法、视频封面选择装置、视频封面选择硬件装置、计算机可读存储介质和视频封面选择终端。其中，该视频封面选择方法包括对从待处理视频中抽取的至少一帧待评估图像帧进行评估；根据评估结果，从所述至少一帧待评估图像帧中选取满足预设封面条件的图像帧作为所述待处理视频的封面。本公开实施例首先对从待处理视频中抽取的至少一帧待评估图像帧进行评估，根据评估结果，从所述至少一帧待评估图像帧中选取满足预设封面条件的图像帧作为所述待处理视频的封面，这样选出的封面具有针对性，从而有利于视频的推广。
上述说明仅是本公开技术方案的概述，为了能更清楚了解本公开的技术手段，而可依照说明书的内容予以实施，并且为让本公开的上述和其他目的、特征和优点能够更明显易懂，以下特举较佳实施例,并配合附图，详细说明如下。
附图说明
图1a为根据本公开一个实施例的视频封面选择方法的流程示意图；
图1b为根据本公开另一个实施例的视频封面选择方法的流程示意图；
图1c为根据本公开另一个实施例的视频封面选择方法的流程示意图；
图1d为根据本公开另一个实施例的视频封面选择方法的流程示意图；
图2a为根据本公开一个实施例的视频封面选择的装置的结构示意图；
图3为根据本公开一个实施例的视频封面选择硬件装置的结构示意图；
图4为根据本公开一个实施例的计算机可读存储介质的结构示意图；
图5为根据本公开一个实施例的视频封面选择终端的结构示意图。
具体实施方式
以下通过特定的具体实例说明本公开的实施方式，本领域技术人员可由本说明书所揭露的内容轻易地了解本公开的其他优点与功效。显然，所描述的实施例仅仅是本公开一部分实施例，而不是全部的实施例。本公开还可以通过另外不同的具体实施方式加以实施或应用，本说明书中的各项细节也可以基于不同观点与应用，在没有背离本公开的精神下进行各种修饰或改变。需说明的是，在不冲突的情况下，以下实施例及实施例中的特征可以相互组合。基于本公开中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本公开保护的范围。
需要说明的是，下文描述在所附权利要求书的范围内的实施例的各种方面。应显而易见，本文中所描述的方面可体现于广泛多种形式中，且本文中所描述的任何特定结构及/或功能仅为说明性的。基于本公开，所属领域的技术人员应了解，本文中所描述的一个方面可与任何其它方面独立地实施，且可以各种方式组合这些方面中的两者或两者以上。举例来说，可使用本文中所阐述的任何数目个方面来实施设备及/或实践方法。另外，可使用除了本文中所阐述的方面中的一或多者之外的其它结构及/或功能性实施此设备及/或实践此方法。
还需要说明的是，以下实施例中所提供的图示仅以示意方式说明本公开的基本构想，图式中仅显示与本公开中有关的组件而非按照实际实施时的组件数目、形状及尺寸绘制，其实际实施时各组件的型态、数量及比例可为一种随意的改变，且其组件布局型态也可能更为复杂。
另外，在以下描述中，提供具体细节是为了便于透彻理解实例。然而，所属领域的技术人员将理解，可在没有这些特定细节的情况下实践所述方面。
为了解决选出来的视频封面没有针对性，不利于视频的推广的技术问 题，本公开实施例提供一种视频封面选择方法。如图1a所示，该视频封面选择方法主要包括如下步骤S1至步骤S2。其中：
步骤S1：对从待处理视频中抽取的至少一帧待评估图像帧进行评估。
具体的，可以从待处理视频中抽取若干帧图像作为待评估图像帧，也可以将待处理视频中的全部图像帧作为待评估图像帧。
步骤S2：根据评估结果，从至少一帧待评估图像帧中选取满足预设封面条件的图像帧作为待处理视频的封面。
其中，评估结果包括但不限于图片质量等级和/或点击率。
其中，预设封面条件可以为但不限于为图像质量等级达到预设要求和/或图像点击率超过预设点击率等。
具体的，当评估结果为图片质量等级时，通过本步骤可以选择出图像质量等级达到预设要求即质量等级比较高的图像帧作为待处理视频的封面，从而利于吸引用户观看。当评估结果为点击率时，通过本步骤可以选择出图像点击率超过预设点击率的图像帧即点击率比较高的图像帧作为待处理视频的封面，由于该图像帧点击率比较高，说明用户对该图像帧感兴趣，因此也利于吸引用户观看，便于视频的推广。
本实施例通过对从待处理视频中抽取的至少一帧待评估图像帧进行评估，根据评估结果，从至少一帧待评估图像帧中选取满足预设封面条件的图像帧作为待处理视频的封面，这样选出的封面具有针对性，从而有利于视频的推广。
在一个可选的实施例中，如图1b所示，步骤S1包括：
S11：对从待处理视频中抽取的至少一帧待评估图像帧进行聚类分析，得到至少一个聚类中心。
其中，可采用的聚类算法包括但不限于以下任意一种：K均值(K-Means)聚类、均值漂移聚类、基于密度的聚类方法、凝聚层次聚类、图团体检测聚类等。
具体的，首先从至少一帧待评估图像帧中选定若干个图像帧作为初始聚类中心，然后将剩余的图像帧作为样本，计算每个样本与初始聚类中心的距离，选择距离初始聚类中心最近的若干个样本作为一类，这样得到至少一类图像集合，然后针对各类图像集合，重新计算聚类中心不断迭代，直到满 足预设条件，这样就得到至少一个聚类中心。
S12：针对各聚类中心，对距离聚类中心最近的图像帧进行评估。
其中，可采用欧式距离和余弦距离表示图像帧与聚类中心之间的距离。
相应的，步骤S2具体包括：
根据评估结果，从距离聚类中心最近的图像帧中选取满足预设封面条件的图像帧作为待处理视频的封面。
其中，评估结果包括但不限于图片质量等级和/或点击率。
本实施例通过对至少一帧待评估图像帧进行聚类分析，得到至少一个聚类中心，针对各聚类中心，对距离聚类中心最近的图像帧进行评估，根据评估结果，从距离聚类中心最近的图像帧中选取满足预设封面条件的图像帧作为待处理视频的封面，这样选出的封面具有针对性，从而有利于视频的推广。
在一个可选的实施例中，如图1c所示，步骤S1具体包括：
对从待处理视频中抽取的至少一帧待评估图像帧分别进行图片质量评估，评估结果为图片质量等级；和/或，
对从待处理视频中抽取的至少一帧待评估图像帧分别进行点击率预测，评估结果为点击率。
其中，图片质量等级例如，分为A、B、C、D四个等级，或者根据图像打分结果进行确定，将0-30作为一个等级，将30-60作为一个等级，将60-80作为一个等级，将80-100作为一个等级，等级划分的标准不同，其对应的预设封面条件也不同，例如，当分为A、B、C、D四个等级的时候，可将图像帧满足A等级设为预设封面条件，当根据打分结果进行确定时，可将图像帧的分值满足80-100设为预设封面条件。
其中，某帧图像的点击率为对该帧图像的点击次数占对该视频中所有帧总点击次数的百分比。
进一步的，步骤S1具体包括：
若评估结果为图片质量等级和点击率，则融合图片质量等级和点击率，重新计算评估结果。
具体的，可根据权重计算法，计算图片质量等级和点击率的加权和，将 加权和作为评估结果。
本实施例通过对至少一帧待评估图像帧分别进行图片质量评估和/或点击率预测，根据图片质量等级和/或点击率，从至少一帧待评估图像帧中选取满足预设封面条件的图像帧作为待处理视频的封面，这样选出的封面具有针对性，从而有利于视频的推广。
在一个可选的实施例中，如图1d所示,步骤S1具体包括：
将从待处理视频中抽取的至少一帧待评估图像帧输入预先训练得到的图像评估模型中进行评估，将图像评估模型的输出结果作为评估结果。
进一步的，图像评估模型包括：图像质量评估模型和/或点击率预测模型，图像质量评估模型用于输出图片质量等级，点击率预测模型用于输出点击率。
具体的，针对图像质量评估模型的训练，包括以下步骤：预先统计视频中各帧图像的质量等级，将已知图像质量等级的图像作为训练样本，并根据不同等级对训练样本进行标注，然后采用深度学习分类算法对标注后的训练样本进行训练学习，得到图像质量评估模型。针对点击率预测模型，预先统计视频中各帧图像的点击率，将已知点击率的图像作为训练样本，并根据不同点击率对训练样本进行标注，然后采用深度学习分类算法对标注后的训练样本进行训练学习，得到点击率预测模型。
其中，可采用的深度学习分类算法包括但不限于以下任意一种：朴素贝叶斯算法、人工神经网络算法、遗传算法、K最近邻(K-NearestNeighbor，KNN)分类算法、聚类算法等。
本实施例通过将至少一帧待评估图像帧输入预先训练得到的图像评估模型中进行评估，将图像评估模型的输出结果作为评估结果，进而根据输出结果，从至少一帧待评估图像帧中选取满足预设封面条件的图像帧作为待处理视频的封面，这样选出的封面具有针对性，从而有利于视频的推广。
在一个可选的实施例中，本实施例的方法还包括：
对从待处理视频中抽取的至少一帧图像帧进行预处理，筛选出符合预设标准的图像帧作为待评估图像帧。
其中，筛选标准可以为图像清晰度。
具体的，通过对图像帧进行预处理，可以进一步过滤掉低质量的图像， 例如带黑边或模糊的图像。
本领域技术人员应能理解，在上述各个实施例的基础上，还可以进行明显变型(例如，对所列举的模式进行组合)或等同替换。
在上文中，虽然按照上述的顺序描述了视频封面选择方法实施例中的各个步骤，本领域技术人员应清楚，本公开实施例中的步骤并不必然按照上述顺序执行，其也可以倒序、并行、交叉等其他顺序执行，而且，在上述步骤的基础上，本领域技术人员也可以再加入其他步骤，这些明显变型或等同替换的方式也应包含在本公开的保护范围之内，在此不再赘述。
下面为本公开装置实施例，本公开装置实施例可用于执行本公开方法实施例实现的步骤，为了便于说明，仅示出了与本公开实施例相关的部分，具体技术细节未揭示的，请参照本公开方法实施例。
为了解决选出来的视频封面没有针对性，不利于视频的推广的技术问题，本公开实施例提供一种视频封面选择装置。该装置可以执行上述视频封面选择方法实施例中的步骤。如图2a所示，该装置主要包括：图像评估模块21和封面选取模块22；其中，图像评估模块21用于对从待处理视频中抽取的至少一帧待评估图像帧进行评估；封面选取模块22用于根据评估结果，从至少一帧待评估图像帧中选取满足预设封面条件的图像帧作为待处理视频的封面。
具体的，图像评估模块21可以从待处理视频中抽取若干帧图像作为待评估图像帧，也可以将待处理视频中的全部图像帧作为待评估图像帧。
其中，评估结果包括但不限于图片质量等级和/或点击率。
其中，预设封面条件可以为但不限于为图像质量等级达到预设要求和/或图像点击率超过预设点击率等。
本实施例通过图像评估模块21对从待处理视频中抽取的至少一帧待评估图像帧进行评估，通过封面选取模块22根据评估结果，从至少一帧待评估图像帧中选取满足预设封面条件的图像帧作为待处理视频的封面，这样选出的封面具有针对性，从而有利于视频的推广。
在一个可选的实施例中，基于图2a所示，图像评估模块21具体用于：对至少一帧待评估图像帧进行聚类分析，得到至少一个聚类中心；针对各聚类中心，对距离聚类中心最近的图像帧进行评估；
封面选取模块22具体用于：根据评估结果，从距离聚类中心最近的图像帧中选取满足预设封面条件的图像帧作为待处理视频的封面。
其中，图像评估模块21可采用的聚类算法包括但不限于以下任意一种：K均值(K-Means)聚类、均值漂移聚类、基于密度的聚类方法、凝聚层次聚类、图团体检测聚类等。
具体的，首先从至少一帧待评估图像帧中选定若干个图像帧作为初始聚类中心，然后将剩余的图像帧作为样本，计算每个样本与初始聚类中心的距离，选择距离初始聚类中心最近的若干个样本作为一类，这样得到至少一类图像集合，然后针对各类图像集合，重新计算聚类中心不断迭代，直到满足预设条件，这样就得到至少一个聚类中心。
其中，可采用欧式距离和余弦距离表示图像帧与聚类中心之间的距离。
其中，评估结果包括但不限于图片质量等级和/或点击率。
本实施例通过图像评估模块21对至少一帧待评估图像帧进行聚类分析，得到至少一个聚类中心，针对各聚类中心，对距离聚类中心最近的图像帧进行评估，通过封面选取模块22根据评估结果，从距离聚类中心最近的图像帧中选取满足预设封面条件的图像帧作为待处理视频的封面，这样选出的封面具有针对性，从而有利于视频的推广。
在一个可选的实施例中，基于图2a所示，图像评估模块21具体用于：对至少一帧待评估图像帧分别进行图片质量评估，评估结果为图片质量等级；和/或，对至少一帧待评估图像帧分别进行点击率预测，评估结果为点击率。
其中，图片质量等级例如，分为A、B、C、D四个等级，或者根据图像打分结果进行确定，将0-30作为一个等级，将30-60作为一个等级，将60-80作为一个等级，将80-100作为一个等级，等级划分的标准不同，其对应的预设封面条件也不同，例如，当分为A、B、C、D四个等级的时候，可将图像帧满足A等级设为预设封面条件，当根据打分结果进行确定时，可将图像帧的分值满足80-100设为预设封面条件。
其中，某帧图像的点击率为对该帧图像的点击次数占对该视频中所有帧总点击次数的百分比。
进一步的，图像评估模块21具体用于：若评估结果为图片质量等级和 点击率，则融合图片质量等级和点击率，重新计算评估结果。
具体的，图像评估模块21可根据权重计算法，计算图片质量等级和点击率的加权和，将加权和作为评估结果。
本实施例通过图像评估模块21对至少一帧待评估图像帧分别进行图片质量评估和/或点击率预测，通过封面选取模块22根据图片质量等级和/或点击率，从至少一帧待评估图像帧中选取满足预设封面条件的图像帧作为待处理视频的封面，这样选出的封面具有针对性，从而有利于视频的推广。
在一个可选的实施例中，基于图2a所示，图像评估模块21具体用于：将至少一帧待评估图像帧输入预先训练得到的图像评估模型中进行评估，将图像评估模型的输出结果作为评估结果。
进一步的，图像评估模型包括：图像质量评估模型和/或点击率预测模型，图像质量评估模型用于输出图片质量等级，点击率预测模型用于输出点击率。
具体的，针对图像质量评估模型的训练，包括以下步骤：预先统计视频中各帧图像的质量等级，将已知图像质量等级的图像作为训练样本，并根据不同等级对训练样本进行标注，然后采用深度学习分类算法对标注后的训练样本进行训练学习，得到图像质量评估模型。针对点击率预测模型，预先统计视频中各帧图像的点击率，将已知点击率的图像作为训练样本，并根据不同点击率对训练样本进行标注，然后采用深度学习分类算法对标注后的训练样本进行训练学习，得到点击率预测模型。
其中，可采用的深度学习分类算法包括但不限于以下任意一种：朴素贝叶斯算法、人工神经网络算法、遗传算法、K最近邻(K-NearestNeighbor，KNN)分类算法、聚类算法等。
本实施例通过图像评估模块21将至少一帧待评估图像帧输入预先训练得到的图像评估模型中进行评估，将图像评估模型的输出结果作为评估结果，进而通过封面选取模块22根据输出结果，从至少一帧待评估图像帧中选取满足预设封面条件的图像帧作为待处理视频的封面，这样选出的封面具有针对性，从而有利于视频的推广。
在一个可选的实施例中，装置还包括：预处理模块23；其中，预处理模块23用于对从待处理视频中抽取的至少一帧图像帧进行预处理，筛选出 符合预设标准的图像帧作为待评估图像帧。
其中，筛选标准可以为图像清晰度。
具体的，通过对图像帧进行预处理，可以进一步过滤掉低质量的图像，例如带黑边或模糊的图像。
有关视频封面选择装置实施例的工作原理、实现的技术效果等详细说明可以参考前述视频封面选择方法实施例中的相关说明，在此不再赘述。
图3是图示根据本公开的实施例的视频封面选择硬件装置的硬件框图。如图3所示，根据本公开实施例的视频封面选择硬件装置30包括存储器31和处理器32。
该存储器31用于存储非暂时性计算机可读指令。具体地，存储器31可以包括一个或多个计算机程序产品，该计算机程序产品可以包括各种形式的计算机可读存储介质，例如易失性存储器和/或非易失性存储器。该易失性存储器例如可以包括随机存取存储器(RAM)和/或高速缓冲存储器(cache)等。该非易失性存储器例如可以包括只读存储器(ROM)、硬盘、闪存等。
该处理器32可以是中央处理单元(CPU)或者具有数据处理能力和/或指令执行能力的其它形式的处理单元，并且可以控制视频封面选择硬件装置30中的其它组件以执行期望的功能。在本公开的一个实施例中，该处理器32用于运行该存储器31中存储的该计算机可读指令，使得该视频封面选择硬件装置30执行前述的本公开各实施例的视频封面选择方法的全部或部分步骤。
本领域技术人员应能理解，为了解决如何获得良好用户体验效果的技术问题，本实施例中也可以包括诸如通信总线、接口等公知的结构，这些公知的结构也应包含在本公开的保护范围之内。
有关本实施例的详细说明可以参考前述各实施例中的相应说明，在此不再赘述。
图4是图示根据本公开的实施例的计算机可读存储介质的示意图。如图4所示，根据本公开实施例的计算机可读存储介质40，其上存储有非暂时性计算机可读指令41。当该非暂时性计算机可读指令41由处理器运行时，执行前述的本公开各实施例的视频特征的比对方法的全部或部分步骤。
上述计算机可读存储介质40包括但不限于：光存储介质(例如：CD－ROM和DVD)、磁光存储介质(例如：MO)、磁存储介质(例如：磁带或移动硬盘)、具有内置的可重写非易失性存储器的媒体(例如：存储卡)和具有内置ROM的媒体(例如：ROM盒)。
有关本实施例的详细说明可以参考前述各实施例中的相应说明，在此不再赘述。
图5是图示根据本公开实施例的终端的硬件结构示意图。如图5所示，该视频封面选择终端50包括上述视频封面选择装置实施例。
该终端可以以各种形式来实施，本公开中的终端可以包括但不限于诸如移动电话、智能电话、笔记本电脑、数字广播接收器、PDA(个人数字助理)、PAD(平板电脑)、PMP(便携式多媒体播放器)、导航装置、车载终端、车载显示终端、车载电子后视镜等等的移动终端以及诸如数字TV、台式计算机等等的固定终端。
作为等同替换的实施方式，该终端还可以包括其他组件。如图5所示，该视频封面选择终端50可以包括电源单元51、无线通信单元52、A/V(音频/视频)输入单元53、用户输入单元54、感测单元55、接口单元56、控制器57、输出单元58和存储器59等等。图5示出了具有各种组件的终端，但是应理解的是，并不要求实施所有示出的组件，也可以替代地实施更多或更少的组件。
其中，无线通信单元52允许终端50与无线通信系统或网络之间的无线电通信。A/V输入单元53用于接收音频或视频信号。用户输入单元54可以根据用户输入的命令生成键输入数据以控制终端的各种操作。感测单元55检测终端50的当前状态、终端50的位置、用户对于终端50的触摸输入的有无、终端50的取向、终端50的加速或减速移动和方向等等，并且生成用于控制终端50的操作的命令或信号。接口单元56用作至少一个外部装置与终端50连接可以通过的接口。输出单元58被构造为以视觉、音频和/或触觉方式提供输出信号。存储器59可以存储由控制器55执行的处理和控制操作的软件程序等等，或者可以暂时地存储己经输出或将要输出的数据。存储器59可以包括至少一种类型的存储介质。而且，终端50可以与通过网络连接执行存储器59的存储功能的网络存储装置协作。控制器57通常控制终端的总体操作。另外，控制器57可以包括用于再现或回放多媒 体数据的多媒体模块。控制器57可以执行模式识别处理，以将在触摸屏上执行的手写输入或者图片绘制输入识别为字符或图像。电源单元51在控制器57的控制下接收外部电力或内部电力并且提供操作各元件和组件所需的适当的电力。
本公开提出的视频特征的比对方法的各种实施方式可以以使用例如计算机软件、硬件或其任何组合的计算机可读介质来实施。对于硬件实施，本公开提出的视频特征的比对方法的各种实施方式可以通过使用特定用途集成电路(ASIC)、数字信号处理器(DSP)、数字信号处理装置(DSPD)、可编程逻辑装置(PLD)、现场可编程门阵列(FPGA)、处理器、控制器、微控制器、微处理器、被设计为执行这里描述的功能的电子单元中的至少一种来实施，在一些情况下，本公开提出的视频特征的比对方法的各种实施方式可以在控制器57中实施。对于软件实施，本公开提出的视频特征的比对方法的各种实施方式可以与允许执行至少一种功能或操作的单独的软件模块来实施。软件代码可以由以任何适当的编程语言编写的软件应用程序(或程序)来实施，软件代码可以存储在存储器59中并且由控制器57执行。
有关本实施例的详细说明可以参考前述各实施例中的相应说明，在此不再赘述。
以上结合具体实施例描述了本公开的基本原理，但是，需要指出的是，在本公开中提及的优点、优势、效果等仅是示例而非限制，不能认为这些优点、优势、效果等是本公开的各个实施例必须具备的。另外，上述公开的具体细节仅是为了示例的作用和便于理解的作用，而非限制，上述细节并不限制本公开为必须采用上述具体的细节来实现。
本公开中涉及的器件、装置、设备、系统的方框图仅作为例示性的例子并且不意图要求或暗示必须按照方框图示出的方式进行连接、布置、配置。如本领域技术人员将认识到的，可以按任意方式连接、布置、配置这些器件、装置、设备、系统。诸如“包括”、“包含”、“具有”等等的词语是开放性词汇，指“包括但不限于”，且可与其互换使用。这里所使用的词汇“或”和“和”指词汇“和/或”，且可与其互换使用，除非上下文明确指示不是如此。这里所使用的词汇“诸如”指词组“诸如但不限于”，且可与其互换使用。
另外，如在此使用的，在以“至少一个”开始的项的列举中使用的 “或”指示分离的列举，以便例如“A、B或C的至少一个”的列举意味着A或B或C，或AB或AC或BC，或ABC(即A和B和C)。此外，措辞“示例的”不意味着描述的例子是优选的或者比其他例子更好。
还需要指出的是，在本公开的系统和方法中，各部件或各步骤是可以分解和/或重新组合的。这些分解和/或重新组合应视为本公开的等效方案。
可以不脱离由所附权利要求定义的教导的技术而进行对在此所述的技术的各种改变、替换和更改。此外，本公开的权利要求的范围不限于以上所述的处理、机器、制造、事件的组成、手段、方法和动作的具体方面。可以利用与在此所述的相应方面进行基本相同的功能或者实现基本相同的结果的当前存在的或者稍后要开发的处理、机器、制造、事件的组成、手段、方法或动作。因而，所附权利要求包括在其范围内的这样的处理、机器、制造、事件的组成、手段、方法或动作。
提供所公开的方面的以上描述以使本领域的任何技术人员能够做出或者使用本公开。对这些方面的各种修改对于本领域技术人员而言是非常显而易见的，并且在此定义的一般原理可以应用于其他方面而不脱离本公开的范围。因此，本公开不意图被限制到在此示出的方面，而是按照与在此公开的原理和新颖的特征一致的最宽范围。
为了例示和描述的目的已经给出了以上描述。此外，此描述不意图将本公开的实施例限制到在此公开的形式。尽管以上已经讨论了多个示例方面和实施例，但是本领域技术人员将认识到其某些变型、修改、改变、添加和子组合。

Claims (16)

  1. 一种视频封面选择方法，其特征在于，包括：
    对从待处理视频中抽取的至少一帧待评估图像帧进行评估；
    根据评估结果，从所述至少一帧待评估图像帧中选取满足预设封面条件的图像帧作为所述待处理视频的封面。
  2. 根据权利要求1所述的方法，其特征在于，所述对从待处理视频中抽取的至少一帧待评估图像帧进行评估的步骤，包括：
    对所述至少一帧待评估图像帧进行聚类分析，得到至少一个聚类中心；
    针对各聚类中心，对距离所述聚类中心最近的图像帧进行评估；
    所述根据评估结果，从所述至少一帧待评估图像帧中选取满足预设封面条件的图像帧作为所述待处理视频的封面的步骤，包括：
    根据评估结果，从所述距离所述聚类中心最近的图像帧中选取满足预设封面条件的图像帧作为所述待处理视频的封面。
  3. 根据权利要求1所述的方法，其特征在于，所述对从待处理视频中抽取的至少一帧待评估图像帧进行评估的步骤，包括：
    对所述至少一帧待评估图像帧分别进行图片质量评估，所述评估结果为图片质量等级；和/或，
    对所述至少一帧待评估图像帧分别进行点击率预测，所述评估结果为点击率。
  4. 根据权利要求3所述的方法，其特征在于，所述对从待处理视频中抽取的至少一帧待评估图像帧进行评估的步骤，包括：
    若所述评估结果为图片质量等级和点击率，则融合所述图片质量等级和点击率，重新计算评估结果。
  5. 根据权利要求1所述的方法，其特征在于，所述对从待处理视频中抽取的至少一帧待评估图像帧进行评估的步骤，包括：
    将所述至少一帧待评估图像帧输入预先训练得到的图像评估模型中进行评估，将所述图像评估模型的输出结果作为所述评估结果。
  6. 根据权利要求5所述的方法，其特征在于，所述图像评估模型包括： 图像质量评估模型和/或点击率预测模型，所述图像质量评估模型用于输出图片质量等级，所述点击率预测模型用于输出点击率。
  7. 根据权利要求1-6任一项所述的方法，其特征在于，所述方法还包括：
    对从所述待处理视频中抽取的至少一帧图像帧进行预处理，筛选出符合预设标准的图像帧作为所述待评估图像帧。
  8. 一种视频封面选择装置，其特征在于，包括：
    图像评估模块，用于对从待处理视频中抽取的至少一帧待评估图像帧进行评估；
    封面选取模块，用于根据评估结果，从所述至少一帧待评估图像帧中选取满足预设封面条件的图像帧作为所述待处理视频的封面。
  9. 根据权利要求8所述的装置，其特征在于，所述图像评估模块具体用于：对所述至少一帧待评估图像帧进行聚类分析，得到至少一个聚类中心；针对各聚类中心，对距离所述聚类中心最近的图像帧进行评估；
    所述封面选取模块具体用于：根据评估结果，从所述距离所述聚类中心最近的图像帧中选取满足预设封面条件的图像帧作为所述待处理视频的封面。
  10. 根据权利要求8所述的装置，其特征在于，所述图像评估模块具体用于：对所述至少一帧待评估图像帧分别进行图片质量评估，所述评估结果为图片质量等级；和/或，对所述至少一帧待评估图像帧分别进行点击率预测，所述评估结果为点击率。
  11. 根据权利要求10所述的装置，其特征在于，所述图像评估模块具体用于：若所述评估结果为图片质量等级和点击率，则融合所述图片质量等级和点击率，重新计算评估结果。
  12. 根据权利要求8所述的装置，其特征在于，所述图像评估模块具体用于：将所述至少一帧待评估图像帧输入预先训练得到的图像评估模型中进行评估，将所述图像评估模型的输出结果作为所述评估结果。
  13. 根据权利要求12所述的装置，其特征在于，所述图像评估模型包括：图像质量评估模型和/或点击率预测模型，所述图像质量评估模型用于输出图片质量等级，所述点击率预测模型用于输出点击率。
  14. 根据权利要求8-13任一项所述的装置，其特征在于，所述装置还包括：
    预处理模块，用于对从所述待处理视频中抽取的至少一帧图像帧进行预处理，筛选出符合预设标准的图像帧作为所述待评估图像帧。
  15. 一种视频封面选择硬件装置，包括：
    存储器，用于存储非暂时性计算机可读指令；以及
    处理器，用于运行所述计算机可读指令，使得所述处理器执行时实现根据权利要求1-7中任意一项所述的视频封面选择方法。
  16. 一种计算机可读存储介质，用于存储非暂时性计算机可读指令，当所述非暂时性计算机可读指令由计算机执行时，使得所述计算机执行权利要求1-7中任意一项所述的视频封面选择方法。
PCT/CN2018/117713 2018-09-13 2018-11-27 视频封面选择方法、装置和计算机可读存储介质 Ceased WO2020052084A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811069772.5 2018-09-13
CN201811069772.5A CN109165301B (zh) 2018-09-13 2018-09-13 视频封面选择方法、装置和计算机可读存储介质

Publications (1)

Publication Number Publication Date
WO2020052084A1 true WO2020052084A1 (zh) 2020-03-19

Family

ID=64879171

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/117713 Ceased WO2020052084A1 (zh) 2018-09-13 2018-11-27 视频封面选择方法、装置和计算机可读存储介质

Country Status (2)

Country Link
CN (1) CN109165301B (zh)
WO (1) WO2020052084A1 (zh)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111859022A (zh) * 2020-07-07 2020-10-30 咪咕文化科技有限公司 封面生成方法、电子设备及计算机可读存储介质
CN112233171A (zh) * 2020-09-03 2021-01-15 上海眼控科技股份有限公司 目标标注质量检验方法、装置、计算机设备和存储介质
CN112651918A (zh) * 2021-01-15 2021-04-13 北京小米松果电子有限公司 图像处理方法及装置、电子设备、存储介质
CN112749298A (zh) * 2020-04-08 2021-05-04 腾讯科技（深圳）有限公司 视频封面确定方法、装置、电子设备及计算机存储介质
CN113761275A (zh) * 2020-11-18 2021-12-07 北京沃东天骏信息技术有限公司 视频预览动图生成方法、装置、设备及可读存储介质
WO2022206729A1 (zh) * 2021-04-01 2022-10-06 影石创新科技股份有限公司 视频封面选择方法、装置、计算机设备和存储介质

Families Citing this family (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111491202B (zh) * 2019-01-29 2021-06-15 广州市百果园信息技术有限公司 一种视频发布方法、装置、设备和存储介质
CN109996091A (zh) * 2019-03-28 2019-07-09 苏州八叉树智能科技有限公司 生成视频封面的方法、装置、电子设备和计算机可读存储介质
CN110287343A (zh) * 2019-06-10 2019-09-27 北京深演智能科技股份有限公司 图片生成方法及装置
CN110149532B (zh) * 2019-06-24 2021-08-17 北京奇艺世纪科技有限公司 一种封面选取方法及相关设备
CN110381368A (zh) * 2019-07-11 2019-10-25 北京字节跳动网络技术有限公司 视频封面生成方法、装置及电子设备
CN110532422B (zh) * 2019-08-07 2022-03-08 北京三快在线科技有限公司 封面生成装置和方法、电子设备和计算机可读存储介质
CN111182295B (zh) * 2020-01-06 2023-08-25 腾讯科技（深圳）有限公司 一种视频数据处理方法、装置、设备以及可读存储介质
CN111464833B (zh) * 2020-03-23 2023-08-04 腾讯科技（深圳）有限公司 目标图像生成方法、目标图像生成装置、介质及电子设备
CN111984821B (zh) * 2020-06-22 2024-11-05 汉海信息技术(上海)有限公司 确定视频动态封面的方法、装置、存储介质及电子设备
CN111918130A (zh) * 2020-08-11 2020-11-10 北京达佳互联信息技术有限公司 视频封面确定方法、装置、电子设备及存储介质
CN112287965A (zh) * 2020-09-21 2021-01-29 卓尔智联(武汉)研究院有限公司 图像质量检测模型训练方法、装置和计算机设备
CN112004033B (zh) * 2020-09-27 2023-05-26 北京小米松果电子有限公司 视频封面确定方法及装置、存储介质
CN112383830A (zh) 2020-11-06 2021-02-19 北京小米移动软件有限公司 视频封面确定方法及装置、存储介质
CN112689187A (zh) * 2020-12-17 2021-04-20 北京达佳互联信息技术有限公司 视频处理方法及装置、电子设备及存储介质
CN112800276B (zh) * 2021-01-20 2023-06-20 北京有竹居网络技术有限公司 视频封面确定方法、装置、介质及设备
CN113343832B (zh) * 2021-06-01 2024-02-02 北京奇艺世纪科技有限公司 视频封面判别方法、装置、设备及计算机可读介质
CN116309236A (zh) * 2021-12-21 2023-06-23 北京奇虎科技有限公司 视频封面图选取方法、装置、设备及存储介质
CN116932788B (zh) * 2022-03-31 2025-08-22 腾讯科技（深圳）有限公司 封面图像提取方法、装置、设备及计算机存储介质
CN115914775B (zh) * 2022-11-21 2024-08-30 北京奇艺世纪科技有限公司 一种封面确定方法、装置、电子设备及存储介质
CN118797099A (zh) * 2024-06-19 2024-10-18 北京优酷科技有限公司 视频封面的选择方法、装置、电子设备及存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104506968A (zh) * 2014-12-31 2015-04-08 北京奇艺世纪科技有限公司 一种视频摘要图确定方法及装置
CN106503693A (zh) * 2016-11-28 2017-03-15 北京字节跳动科技有限公司 视频封面的提供方法及装置
CN107832725A (zh) * 2017-11-17 2018-03-23 北京奇虎科技有限公司 基于评价指标的视频封面提取方法及装置
CN107918656A (zh) * 2017-11-17 2018-04-17 北京奇虎科技有限公司 基于视频标题的视频封面提取方法及装置
CN107958030A (zh) * 2017-11-17 2018-04-24 北京奇虎科技有限公司 视频封面推荐模型优化方法及装置

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2688029A1 (en) * 2012-04-11 2014-01-22 Taboola.com Ltd. Click through rate estimation in varying display situations
CN107330049B (zh) * 2017-06-28 2020-05-22 北京搜狐新媒体信息技术有限公司 一种新闻热度预估方法及系统

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104506968A (zh) * 2014-12-31 2015-04-08 北京奇艺世纪科技有限公司 一种视频摘要图确定方法及装置
CN106503693A (zh) * 2016-11-28 2017-03-15 北京字节跳动科技有限公司 视频封面的提供方法及装置
CN107832725A (zh) * 2017-11-17 2018-03-23 北京奇虎科技有限公司 基于评价指标的视频封面提取方法及装置
CN107918656A (zh) * 2017-11-17 2018-04-17 北京奇虎科技有限公司 基于视频标题的视频封面提取方法及装置
CN107958030A (zh) * 2017-11-17 2018-04-24 北京奇虎科技有限公司 视频封面推荐模型优化方法及装置

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112749298A (zh) * 2020-04-08 2021-05-04 腾讯科技（深圳）有限公司 视频封面确定方法、装置、电子设备及计算机存储介质
CN112749298B (zh) * 2020-04-08 2024-02-09 腾讯科技（深圳）有限公司 视频封面确定方法、装置、电子设备及计算机存储介质
CN111859022A (zh) * 2020-07-07 2020-10-30 咪咕文化科技有限公司 封面生成方法、电子设备及计算机可读存储介质
CN112233171A (zh) * 2020-09-03 2021-01-15 上海眼控科技股份有限公司 目标标注质量检验方法、装置、计算机设备和存储介质
CN113761275A (zh) * 2020-11-18 2021-12-07 北京沃东天骏信息技术有限公司 视频预览动图生成方法、装置、设备及可读存储介质
CN112651918A (zh) * 2021-01-15 2021-04-13 北京小米松果电子有限公司 图像处理方法及装置、电子设备、存储介质
WO2022206729A1 (zh) * 2021-04-01 2022-10-06 影石创新科技股份有限公司 视频封面选择方法、装置、计算机设备和存储介质

Also Published As

Publication number Publication date
CN109165301B (zh) 2021-04-20
CN109165301A (zh) 2019-01-08

Similar Documents

Publication Publication Date Title
WO2020052084A1 (zh) 视频封面选择方法、装置和计算机可读存储介质
CN107678845B (zh) 应用程序管控方法、装置、存储介质及电子设备
CN109598231B (zh) 一种视频水印的识别方法、装置、设备及存储介质
CN105320957B (zh) 分类器训练方法和装置
CN110942011B (zh) 一种视频事件识别方法、系统、电子设备及介质
CN103747346B (zh) 一种多媒体视频播放的控制方法及多媒体视频播放器
CN113766330A (zh) 基于视频生成推荐信息的方法和装置
JP2011215963A (ja) 電子機器、画像処理方法及びプログラム
WO2020052083A1 (zh) 侵权图片的识别方法、装置和计算机可读存储介质
CN113111197B (zh) 多媒体内容的推荐方法、装置、设备及存储介质
CN102622399A (zh) 搜索装置、搜索方法和程序
CN110334753B (zh) 视频分类方法、装置、电子设备及存储介质
CN110019777B (zh) 一种信息分类的方法及设备
US10701434B1 (en) Extracting session information from video content to facilitate seeking
CN109891405B (zh) 基于用户装置的消费模式来修改视频内容在用户装置上的呈现的方法、系统和介质
CN103970808A (zh) 媒体内容推荐方法及媒体播放系统
CN111598012B (zh) 一种图片聚类管理方法、系统、设备及介质
CN105760488A (zh) 基于多层次特征融合的图像表达方法和装置
CN109376764B (zh) 基于聚类的数据收集方法、装置和计算机可读存储介质
WO2022194084A1 (zh) 视频播放方法、终端设备、装置、系统及存储介质
Mironică et al. Fisher kernel temporal variation-based relevance feedback for video retrieval
CN113221451B (zh) 跌倒预测方法、系统、电子设备及存储介质
US8498978B2 (en) Slideshow video file detection
US12517950B2 (en) Methods, systems, and media for modifying search results based on search query risk
CN117692692A (zh) 一种展示方法、装置、电子设备、以及计算机可读介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18933007

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205 DATED 24/06/2021)

122 Ep: pct application non-entry in european phase

Ref document number: 18933007

Country of ref document: EP

Kind code of ref document: A1