Similarity-based video automatic segmentation mirror method, apparatus and device, and medium

Abstract

The invention provides a similarity-based video automatic splitting method, device and equipment and a medium, and relates to the technical field of computer vision. The method comprises the following steps: extracting color histogram features of a video to be segmented, and calculating the similarity of adjacent video frames based on an earth transporter distance to obtain a similarity set; then calculating a difference value between a front similarity and a rear similarity in the similarity set to obtain a similarity difference value set; finding out a maximum value in the similarity difference value set to obtain a maximum difference value; and dynamically setting a minimum change threshold value according to the maximum difference value, storing the video frames exceeding the minimum change threshold value in the similarity set into a sub-shot point list, and segmenting the video by taking the sub-shot points as segmentation points to realize automatic segmentation of the video. According to the method, based on the distance of the earth porters, the influence of the distance and time between the video frames on the similarity is considered at the same time, the similarity between the video frames can be better captured, and therefore the accuracy of video segmentation is improved as much as possible.

Description

Technical Field

[0001] The present invention relates to the field of computer vision technology, and in particular to a method and device for automatically splitting videos based on similarity. Background Art

[0002] Computer vision technology plays a core role in video processing. Through computer vision technology, the system can identify key frames, motion changes, color distribution and other features in the video, thereby determining the start and end positions of the shot, making it easier to understand the video and imitate shooting.

[0003] The existing technology generally extracts the features of video frames by calculating Euclidean distance, cosine similarity or chi-square distance, but the performance is low and the spatial relationship between adjacent pixels cannot be accurately captured, resulting in inaccurate results. For example, CN110175591B "A method for obtaining video similarity" calculates the chi-square distance of the color histogram of adjacent video frame images as the similarity between adjacent video frame images. The results of methods such as chi-square distance are intuitive and easy to understand, but they have the following disadvantages:

[0004] (1) In the case of small samples, the stability of the test is poor, which may lead to inaccurate results. Therefore, when calculating the similarity of video frames, if the difference between frames is small or rich in details, subtle changes may not be accurately captured.

[0005] (2) In a video frame, the spatial relationship between adjacent pixels is very important for understanding the overall scene. Existing methods mainly focus on the differences between distributions and ignore the spatial relationship between pixels.

[0006] In view of this, the applicant filed this application after studying the existing technology. Summary of the invention

[0007] The present invention aims to provide a similarity-based automatic video segmentation method, device, equipment and medium to solve the shortcomings of existing methods such as being unable to accurately capture subtle changes in samples and ignoring the spatial relationship between pixels.

[0008] In order to solve the above technical problems, the present invention is implemented through the following technical solutions:

[0009] A video automatic segmentation method based on similarity, comprising:

[0010] S1, obtain the video to be segmented;

[0011] S2, extracting the color histogram features of each video frame of the video to be segmented;

[0012] S3, according to the color histogram features, combined with the distance and time information of the video frames, based on the earth porter distance, calculating the similarity of adjacent video frames, and obtaining a similarity set arranged in the order of video time;

[0013] S4, calculating the difference between the previous and next similarities according to the similarity set to obtain a similarity difference set;

[0014] S5, arranging the similarity difference value set in descending order, finding the maximum value, obtaining the maximum difference value, and storing the video frame corresponding to the maximum difference value in the shot point list;

[0015] S6, setting a minimum change threshold according to the maximum difference, and storing the video frames in the similarity set that exceed the minimum change threshold into the shot point list;

[0016] S7, dividing the video by using the splitting points in the splitting point list as splitting points, so as to realize automatic splitting of the video.

[0017] Preferably, the calculation formula for calculating the similarity of adjacent video frames is:

[0018]

[0019] Among them, SM i,j is the similarity between two adjacent video frames i and j; H i , H j Respectively represent the color histogram features of the i-th frame and the j-th frame; EMD(.) represents the earth porter distance function; α and β are adjustment parameters;

[0020] N is the total number of frames of the video to be segmented; t i ,t j They respectively represent the time points corresponding to the i-th frame and the j-th frame; T is the total duration of the video to be segmented.

[0021] Preferably, the calculation formula of the minimum change threshold is:

[0022]

[0023] Among them, M is the minimum change threshold; represents the maximum value in the similarity difference set D, that is, the maximum difference; k1 and k2 are the set adjustment coefficients.

[0024] Preferably, the method further comprises: adding the starting frame and the last frame of the video to be segmented into the shot point list to form a complete shot point list, and then merging adjacent shot points in the complete shot point list into a time period for segmenting the video.

[0025] The present invention also provides a similarity-based automatic video splitting device, comprising:

[0026] A data acquisition unit, used for acquiring the video to be segmented;

[0027] A feature extraction unit, used to extract color histogram features of each video frame of the video to be segmented;

[0028] A similarity unit, for calculating the similarity of adjacent video frames based on the color histogram features, combined with the distance and time information of the video frames, and based on the earth porter distance, to obtain a similarity set arranged in the order of video time;

[0029] A similarity difference unit, used for calculating the difference between two similarities before and after the similarity set to obtain a similarity difference set;

[0030] A maximum difference point unit is used to sort the similarity difference set in descending order, find the maximum value, obtain the maximum difference, and store the video frame corresponding to the maximum difference in the shot point list;

[0031] A minimum change threshold unit, used to set a minimum change threshold according to the maximum difference, and store the video frames in the similarity set that exceed the minimum change threshold into the shot point list;

[0032] The video segmentation unit is used to segment the video using the segmentation points in the segmentation point list as segmentation points, thereby realizing automatic segmentation of the video.

[0033] The present invention also provides a similarity-based automatic video segmentation device, including a processor and a memory, wherein the memory stores a computer program, and the computer program can be executed by the processor to implement the similarity-based automatic video segmentation method as described above.

[0034] The present invention also provides a computer-readable storage medium, on which computer-readable instructions are stored. When the computer-readable instructions are executed by a processor of a device where the computer-readable storage medium is located, the above-mentioned method for automatically splitting videos based on similarity is implemented.

[0035] In summary, compared with the prior art, the present invention has the following beneficial effects:

[0036] The present invention combines the distance and time information of the video frames, calculates the similarity of adjacent video frames based on the earth porter distance, determines the splitting point by comparing the color histogram similarity between adjacent front and back frames, considers the relationship between the similarity between the two distributions and the geometric distance, and better captures the similarity between images. In addition, when calculating the similarity, the influence of the distance and time between the video frames on the similarity is also considered. The greater the distance between the video frames, the smaller the similarity may be; the greater the time interval, the smaller the similarity may be, so as to reduce the error when calculating the similarity of the color histogram.

[0037] In addition, in order to prevent small color changes from being misjudged as split-shot points, the minimum change threshold is dynamically set through a calculation formula to improve the accuracy of split-shot point detection. Compared with traditional video image processing methods (such as Euclidean distance, cosine similarity, etc.), it has good stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0039] Figure 1 A schematic diagram of a similarity-based automatic video segmentation method provided in Example 1 of the present invention.

[0040] Figure 2 A schematic diagram of a similarity-based automatic video segmentation device provided in Embodiment 2 of the present invention.

[0041] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the invention claimed for protection, but merely represents the selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0043] Embodiment 1

[0044] Embodiment 1 of the present invention provides a similarity-based automatic video segmentation method, which can be implemented by a similarity-based automatic video segmentation device (hereinafter referred to as the segmentation device), and in particular, executed by one or more processors in the segmentation device.

[0045] In this embodiment, the storyboard device may be an electronic device equipped with a processor, the processor having a computer program of the similarity-based automatic video storyboarding method and the computer program can be executed, such as a computer, a smart phone, a smart tablet, etc., without limitation.

[0046] like Figure 1 As shown, a method for automatically splitting a video based on similarity includes steps S1 to S7.

[0047] S1, obtain the video to be segmented.

[0048] In this embodiment, the video is processed according to the time sequence of the video to be segmented to obtain frame sequence pictures in the video.

[0049] S2, extracting the color histogram features of each video frame of the video to be segmented.

[0050] In this embodiment, the color histogram is a statistical feature of the color distribution in the image. In the same shot, if no person or object enters or leaves the scene, the color distribution in the video frame will basically not change. Based on the color histogram feature, if there is a large fluctuation between adjacent frames, it can be determined that the frame is one of the split-shot points.

[0051] S3, according to the color histogram features, combined with the distance and time information of the video frames, based on the earth porter distance, calculate the similarity of adjacent video frames to obtain a similarity set arranged in video time order.

[0052] In this embodiment, the Earth Mover's distance, also known as the Wasserstein distance or the bulldozer distance, is a mathematical method for measuring the distance between two probability distributions. It takes into account the relationship between the similarity between the distributions and the geometric distance, and is measured by calculating the minimum total cost (or "bulldozer cost") required to transform one distribution into another.

[0053] Furthermore, the formula for calculating the similarity between adjacent video frames is:

[0054]

[0055] Among them, SM i,j is the similarity between two adjacent video frames i and j; H i , H j Respectively represent the color histogram features of the i-th frame and the j-th frame; EMD(.) represents the earth porter distance function; α and β are adjustment parameters;

[0056] N is the total number of frames of the video to be segmented; t i ,t j They respectively represent the time points corresponding to the i-th frame and the j-th frame; T is the total duration of the video to be segmented.

[0057] This formula not only considers the EMD distance similarity of the color histogram, but also considers a second factor, that is, the effect of the distance between frames on the similarity. The greater the distance between frames, the smaller the similarity may be. α is an adjustment parameter for the distance factor between frames.

[0058] The third factor considered is the effect of time on similarity. The longer the time interval, the smaller the similarity may be. β is the adjustment parameter for the time factor.

[0059] At this time, the similarity set SM={SM 1,2 ,SM 2,3 ,...,SM i,j ...SM N-1,N}.

[0060] S4, according to the similarity set, calculating the difference between the two similarities before and after, to obtain a similarity difference set.

[0061] Assume that the set of similarity differences is D, then D = {d1, d2, ..., d i ..d m}, where d1 is SM 1,2 With SM 2,3 The difference of i For SM i,i+1 With SM i+1,i+2 difference; m = |D| = N-2, is the dimension of the similarity difference set.

[0062] In this embodiment, the similarity difference reflects the degree of color change between adjacent frames.

[0063] S5, arranging the similarity difference set in descending order, finding the maximum value, obtaining the maximum difference, and storing the video frame corresponding to the maximum difference into the shot point list.

[0064] In this embodiment, the maximum value in the similarity difference value set, ie, the maximum difference value, is found, and the video frame corresponding to the maximum difference value is considered as a potential shot separation point.

[0065] S6, setting a minimum change threshold according to the maximum difference, and storing the video frames in the similarity set that exceed the minimum change threshold into the shot point list;

[0066] In order to avoid misjudging slight color changes as split-shot points, a minimum change threshold M is set, and frames below this threshold are not considered split-shot points.

[0067] Preferably, the calculation formula of the minimum change threshold is:

[0068]

[0069] Among them, M is the minimum change threshold; max(D) represents the maximum value in the similarity difference set D, that is, the maximum difference; k1 and k2 are set adjustment coefficients, generally based on experience, k1=1.5 and k2=0.4 are set.

[0070] For example, if the similarity set SM = {0.95, 0.94, 0.93, 0.84}, then the similarity difference set is:

[0071] D = {0.01, 0.01, 0.09},

[0072] The similarity difference set after descending order is: {0.09, 0.01, 0.01}, then, max(D) = 0.09, which is the maximum difference, and the two elements corresponding to the maximum difference are 0.93 and 0.84.

[0073] Calculated That is, the minimum change threshold is 0.4.

[0074] Specifically, the values ​​of k1 and k2 may be adjusted again according to the calculation results.

[0075] S7, dividing the video by using the splitting points in the splitting point list as splitting points, so as to realize automatic splitting of the video.

[0076] In another preferred embodiment, it also includes: adding the starting frame and the last frame of the video to be segmented to the shot point list to form a complete shot point list, and then merging adjacent shot points in the complete shot point list into a time period for segmenting the video.

[0077] In another preferred embodiment, the method is mainly applied to the following aspects:

[0078] (1) After using this method to storyboard a video, short video bloggers or merchants can learn the video structure of other bloggers in the storyboard video and shoot accordingly.

[0079] (2) Since the video capacity is large and it takes a certain amount of time to load, after the video is storyboarded using this method, it is convenient to conduct targeted analysis and understanding of the storyboarded video segments. Later, the storyboarded video segments can be selectively synthesized according to one's own copy.

[0080] In summary, compared with the prior art, the present invention has the following beneficial effects:

[0081] The present invention combines the distance and time information of the video frames, calculates the similarity of adjacent video frames based on the earth porter distance, determines the splitting point by comparing the color histogram similarity between adjacent front and back frames, considers the relationship between the similarity between the two distributions and the geometric distance, and better captures the similarity between images. In addition, when calculating the similarity, the influence of the distance and time between the video frames on the similarity is also considered. The greater the distance between the video frames, the smaller the similarity may be; the greater the time interval, the smaller the similarity may be, so as to reduce the error when calculating the similarity of the color histogram.

[0082] In addition, in order to prevent small color changes from being misjudged as split-shot points, the minimum change threshold is dynamically set through a calculation formula to improve the accuracy of split-shot point detection. Compared with traditional video image processing methods (such as Euclidean distance, cosine similarity, etc.), it has good stability.

[0083] Embodiment 2

[0084] like Figure 2 As shown, the second embodiment of the present invention further provides a similarity-based automatic video splitting device, comprising:

[0085] A data acquisition unit, used for acquiring the video to be segmented;

[0086] A feature extraction unit, used to extract color histogram features of each video frame of the video to be segmented;

[0087] A similarity unit, for calculating the similarity of adjacent video frames based on the color histogram features, combined with the distance and time information of the video frames, and based on the earth porter distance, to obtain a similarity set arranged in the order of video time;

[0088] A similarity difference unit, used for calculating the difference between two similarities before and after the similarity set to obtain a similarity difference set;

[0089] A maximum difference point unit is used to sort the similarity difference set in descending order, find the maximum value, obtain the maximum difference, and store the video frame corresponding to the maximum difference in the shot point list;

[0090] A minimum change threshold unit, used to set a minimum change threshold according to the maximum difference, and store the video frames in the similarity set that exceed the minimum change threshold into the shot point list;

[0091] The video segmentation unit is used to segment the video using the segmentation points in the segmentation point list as segmentation points, thereby realizing automatic segmentation of the video.

[0092] Preferably, the calculation formula for calculating the similarity of adjacent video frames in the similarity unit is:

[0093]

[0094] Among them, SM i,j is the similarity between two adjacent video frames i and j; H i , H j Respectively represent the color histogram features of the i-th frame and the j-th frame; EMD(.) represents the earth porter distance function; α and β are adjustment parameters;

[0095] N is the total number of frames of the video to be segmented; t i ,t j They respectively represent the time points corresponding to the i-th frame and the j-th frame; T is the total duration of the video to be segmented.

[0096] Preferably, the calculation formula of the minimum change threshold in the minimum change threshold unit is:

[0097]

[0098] Among them, M is the minimum change threshold; max(D) represents the maximum value in the similarity difference set D, that is, the maximum difference; k1 and k2 are set adjustment coefficients.

[0099] Preferably, it also includes: adding the starting frame and the last frame of the video to be segmented into the shot point list to form a complete shot point list, and then merging adjacent shot points in the complete shot point list into a time period for segmenting the video.

[0100] Embodiment 3

[0101] The third embodiment of the present invention further provides a similarity-based automatic video segmentation device, which includes a memory and a processor, wherein the memory stores a computer program, and the computer program can be executed by the processor to implement the similarity-based automatic video segmentation method as described above.

[0102] Embodiment 4

[0103] The fourth embodiment of the present invention further provides a computer-readable storage medium, on which computer-readable instructions are stored. When the computer-readable instructions are executed by a processor of a device where the computer-readable storage medium is located, the above-mentioned similarity-based automatic video segmentation method is implemented.

[0104] In several embodiments provided in the embodiments of the present invention, it should be understood that the disclosed apparatus and method can also be implemented in other ways. The apparatus and method embodiments described above are merely schematic. For example, the flowcharts in the accompanying drawings show the possible architecture, functions and operations of the apparatus, method and computer program product according to multiple embodiments of the present invention. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, program segment or a part of the code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flow chart, and the combination of boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0105] In addition, the functional modules in the various embodiments of the present invention may be integrated together to form an independent part, or each module may exist independently, or two or more modules may be integrated to form an independent part.

[0106] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, an electronic device, or a network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk. It should be noted that, in this article, the terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements that are not explicitly listed, or also includes elements inherent to such a process, method, article or device. Without more constraints, an element defined by the phrase "comprising a..." does not exclude the existence of other identical elements in the process, method, article or apparatus comprising the element.

[0107] The terms used in the embodiments of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. The singular forms "a", "said" and "the" used in the embodiments of the present invention and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings.

[0108] It should be understood that the term "and / or" used in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0109] The word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrases "if it is determined" or "if (stated condition or event) is detected" may be interpreted as "when it is determined" or "in response to determining" or "when detecting (stated condition or event)" or "in response to detecting (stated condition or event)", depending on the context.

[0110] The "first\second" mentioned in the embodiments is only to distinguish similar objects, and does not represent a specific order for the objects. It is understandable that the "first\second" can be interchanged with the specific order or sequence where permitted. It should be understood that the objects distinguished by "first\second" can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than those illustrated or described herein.

[0111] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A video automatic segmentation method based on similarity, characterized in that: include: S1, obtain the video to be segmented; S2, extracting the color histogram features of each video frame of the video to be segmented; S3, according to the color histogram features, combined with the distance and time information of the video frames, based on the earth porter distance, calculating the similarity of adjacent video frames, and obtaining a similarity set arranged in the order of video time; S4, calculating the difference between the previous and next similarities according to the similarity set to obtain a similarity difference set; S5, arranging the similarity difference value set in descending order, finding the maximum value, obtaining the maximum difference value, and storing the video frame corresponding to the maximum difference value in the shot point list; S6, setting a minimum change threshold according to the maximum difference, and storing the video frames in the similarity set that exceed the minimum change threshold into the shot point list; S7, dividing the video by using the splitting points in the splitting point list as splitting points, so as to realize automatic splitting of the video.

2. The method for automatically splitting a video based on similarity according to claim 1, characterized in that ,The formula for calculating the similarity of adjacent video frames is : Among them, SM i,j is the similarity between two adjacent video frames i and j; H i , H j Respectively represent the color histogram features of the i-th frame and the j-th frame; EMD(.) represents the earth porter distance function; α and β are adjustment parameters; N is the total number of frames of the video to be segmented; t i ,t j They respectively represent the time points corresponding to the i-th frame and the j-th frame; T is the total duration of the video to be segmented.

3. The method for automatically splitting a video based on similarity according to claim 2, characterized in that , the calculation formula of the minimum change threshold is: Among them, M is the minimum change threshold; max(D) represents the maximum value in the similarity difference set D, that is, the maximum difference; k1 and k2 are set adjustment coefficients.

4. The method for automatically splitting a video based on similarity according to claim 1, characterized in that , also includes: adding the starting frame and the last frame of the video to be segmented to the split shot point list to form a complete split shot point list, and then merging the adjacent split shot points in the complete split shot point list into the time period of the split video.

5. A similarity-based automatic video splitting device, characterized in that: include: A data acquisition unit, used for acquiring the video to be segmented; A feature extraction unit, used to extract color histogram features of each video frame of the video to be segmented; A similarity unit, used to calculate the similarity of adjacent video frames based on the color histogram features, combined with the distance and time information of the video frames, based on the earth porter distance, to obtain a similarity set arranged in the order of video time; A similarity difference unit, used for calculating the difference between two similarities before and after the similarity set to obtain a similarity difference set; A maximum difference point unit is used to sort the similarity difference set in descending order, find the maximum value, obtain the maximum difference, and store the video frame corresponding to the maximum difference in the shot point list; A minimum change threshold unit, used to set a minimum change threshold according to the maximum difference, and store the video frames in the similarity set that exceed the minimum change threshold into the shot point list; The video segmentation unit is used to segment the video using the segmentation points in the segmentation point list as segmentation points, thereby realizing automatic segmentation of the video.

6. The video automatic segmentation device based on similarity according to claim 5, characterized in that ,The formula for calculating the similarity of adjacent video frames is : Among them, SM i,j is the similarity between two adjacent video frames i and j; H i , H j Respectively represent the color histogram features of the i-th frame and the j-th frame; EMD(.) represents the earth porter distance function; α and β are adjustment parameters; N is the total number of frames of the video to be segmented; t i ,t j They respectively represent the time points corresponding to the i-th frame and the j-th frame; T is the total duration of the video to be segmented.

7. The video automatic segmentation device based on similarity according to claim 6, characterized in that , the calculation formula of the minimum change threshold is: Among them, M is the minimum change threshold; max(D) represents the maximum value in the similarity difference set D, that is, the maximum difference; k1 and k2 are set adjustment coefficients.

8. The video automatic segmentation device based on similarity according to claim 5, characterized in that , also includes: adding the starting frame and the last frame of the video to be segmented to the split shot point list to form a complete split shot point list, and then merging the adjacent split shot points in the complete split shot point list into the time period of the split video.

9. A video automatic splitting device based on similarity, characterized in that: It comprises a processor and a memory, wherein the memory stores a computer program, and the computer program can be executed by the processor to implement a similarity-based automatic video segmentation method as described in any one of claims 1 to 4.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by a processor of a device where the computer-readable storage medium is located, the method for automatically splitting videos based on similarity as described in any one of claims 1 to 4 is implemented.

Images