Data storage system and data writing method

Abstract

The invention relates to the technical field of data processing, in particular to a data storage system and a data writing method.The data storage system comprises a traversal module used for traversing data content stored in the storage system; the classification module is used for classifying data contents stored in the data storage system based on similarity; the creating module is used for creating a data pool in the data storage system, and distinguishing and storing the data content classified by the classification module based on the data pool; according to the method, the data content in the data storage system is finely classified, the data pools are created to distinguish the stored data content, and repeated data is reduced based on data similarity analysis in the process, so that invalid occupation of the storage space in the data storage system is avoided, and meanwhile, key data information is marked, so that the data storage efficiency is improved. A data pool used for data content storage is named, reliability management of data storage is achieved, and fast and efficient services are provided for a data storage system when the data storage system is used for data retrieval and calling.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a data storage system and a data writing method. Background Art

[0002] Data writing is the process of adding new data to the storage system, and the data must be accurate and complete. Data storage is the long-term preservation of data so that it can be accessed and used at any time. A good storage system can efficiently manage data and ensure its security and availability. Writing and storage work together to provide strong support for data analysis and business decision-making.

[0003] The invention patent with application number 202311170097.6 discloses a data writing method, which is applied to a processor and is characterized in that it includes: reading the stored data currently stored in each erase unit of the storage; identifying the target data to be written into each erase unit in the storage; comparing the stored data with the target data to generate a comparison result; sending a control instruction corresponding to the comparison result to the storage; the control instruction includes a first instruction and a second instruction, the first instruction is used by the processor to control the storage not to perform an erase operation on the erase unit, and write the corresponding target data into the erase unit; the second instruction is used by the processor to control the storage to perform an erase operation on the erase unit, and then write the corresponding target data into the erase unit after the erase operation.

[0004] This application aims to solve the problem that "when erasing data in a storage device, different storage devices will have different minimum erasing units, and each erasing operation needs to be performed in units of the minimum erasing unit. Each erasing operation usually takes a long time, ranging from several milliseconds to hundreds of milliseconds, while the time required to write data to the storage device is usually short, only tens of microseconds to hundreds of microseconds. In actual production and manufacturing, when the data in the storage device needs to be frequently modified or the data needs to be written to the storage device in large quantities, multiple erasing operations are required, which results in a slower data writing speed, thereby extending the production cycle."

[0005] However, current data storage systems often manage stored data by setting simple data storage intervals and distinguishing and storing the data. The application of distinguished storage intervals is limited to user-defined storage locations and does not have good storage management conditions. Therefore, when a large amount of data is stored in the data storage system, the process of retrieving the data content in the data storage system is more difficult.

[0006] To this end, we propose a data storage system. Summary of the invention

[0007] In view of the above shortcomings of the prior art, the present invention provides a data storage system and a data writing method, which solve the technical problems raised in the above background technology.

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0009] In a first aspect, a data storage system comprises:

[0010] A traversal module, used to traverse the data content stored in the storage system;

[0011] A classification module, used to classify the data content stored in the data storage system based on similarity;

[0012] A creation module is used to create a data pool in the data storage system, and to differentiate and store the data content classified by the classification module based on the data pool;

[0013] A retrieval module, used for editing text information and performing a data retrieval operation in a data storage system using the edited text information as retrieval content;

[0014] A queue module is used to monitor the number of times data content in each data pool in the data storage system is retrieved, evaluate the activity of the data content based on the number of times the data content is retrieved, and arrange the data content in each data pool using the activity of the data content;

[0015] The control unit is used to set the operation cycle of the queue module, control the queue module to run continuously based on the operation cycle, and arrange the queues and refresh the data in each data pool in the data storage system.

[0016] Furthermore, during the operation phase of the traversal module, the storage system does not perform a data write operation, and the traversal module is internally provided with submodules, including:

[0017] The identification unit is used to run synchronously with the traversal module running phase to identify the similarities between groups of data in the storage system;

[0018] Among them, during the operation phase of the traversal module, the recognition unit synchronously obtains the recognition results of the similarities between each group of data, and eliminates the same data in the storage system based on the recognition results. The data contents stored in the storage system traversed by the traversal module are all data contents that have been processed by the recognition unit for similarity recognition.

[0019] Furthermore, when the identification unit identifies the similarity between each group of data in the storage system, the similarity identification logic between the two groups of data contents is expressed as follows:

[0020]

[0021] Where: SIMM(X,Y) is the similarity between data X and data Y; n is the set of elements in data X; m is the set of elements in data Y; p(x i ,y j ) is the element pair (x i ,y j ) occurs; p(x i ) is the probability of the i-th element in the data X appearing; p(y j ) is the probability of the jth element in the data Y; γ is the correction;

[0022] The larger the SIMM(X, Y) value is, the higher the similarity between the two sets of data is. When SIMM(X, Y) ≥ 95%, the two sets of data are determined to be the same data.

[0023] Furthermore, the modified γ value applied in the similarity recognition logic between the two sets of data contents obeys:

[0024]

[0025] Where: μ X , μ Y is the mean of data X and data Y; σ XY is the covariance of data X and data Y; C1 and C2 are constants; σ X , σ Y is the standard deviation of data X and data Y;

[0026] Among them, constants C1 and C2 are both greater than 0.

[0027] Furthermore, the classification module is provided with submodules at the lower level, including:

[0028] A setting unit, used for setting a similarity classification interval when the classification module is running to classify the data stored in the data storage system;

[0029] A marking unit, used for marking the data content classified by the classification module;

[0030] Among them, the similarity classification interval set in the setting unit is customized by the system user, and the initial default similarity classification interval is 40%, that is, data content with a similarity of not less than 40% is classified as similar data. When the marking unit marks the classified data content, the same part of the data in each group of data in the same data is used as the marking content, and each data in the same data is marked, and the similarity between each group of data in the same data is not less than 40%.

[0031] Furthermore, the number of data pools created in the creation module is equal to the number of data categories in the classification module, and each group of data pools is used to store a group of similar data, and each group of data pools is named based on the tag content of the data content stored therein.

[0032] Furthermore, the operation of editing text information in the retrieval module is performed by the system-side user. After the text information used for retrieval is edited, the retrieval module first uses the data pool as the retrieval target to search for the data pool with the same name as the edited text information. After determining the data pool, the system-side user applies the retrieval module to edit the text information again. The retrieval module performs a retrieval operation of the data content in the determined data pool based on the text information edited again by the system-side user.

[0033] Furthermore, the activity of data content in the queue module is expressed as:

[0034]

[0035] Where: f(a) is the activity of data content a; q is the number of times data content a is retrieved during the operation cycle of the queue module; t is the average time for data content a to be retrieved once during the operation cycle of the queue module;

[0036] Based on the above formula, the corresponding activity of each data content in the data pool is obtained. The queue module arranges the data content in the data pool in descending order based on the activity of the data content, so that the data content with high activity is at the front position of the queue, and the data content with low activity is at the back position of the queue;

[0037] The queue module operation cycle set during the operation phase of the control unit is subject to:

[0038] When the front-most data content in the data content queue is retrieved less times than the number of times in the previous operation cycle within the same time threshold as the previous operation cycle, the queue module refreshes and runs once.

[0039] Furthermore, the traversal module is interactively connected to an identification unit via a wireless network, the traversal module is interactively connected to a classification module via a wireless network, the classification module is interactively connected to a setting unit and a marking unit via a wireless network, the classification module is interactively connected to a creation module via a wireless network, the creation module is interactively connected to the marking unit via a wireless network, the creation module is interactively connected to a retrieval module via a wireless network, and the creation module is interactively connected to a queue module and a control unit via a wireless network.

[0040] In a second aspect, a data writing method comprises the following steps:

[0041] Traverse the data content that needs to be written into the data storage system, and identify the data types in the data content, that is, the format of the data corresponding to each data type;

[0042] Identify the data size ratio of each format of data in the data corresponding to each data content;

[0043] Select the data format with the largest proportion among the same type of data as the data conversion format, and perform format conversion on the corresponding data of all data formats except the data format with the largest proportion among the same type of data, so that the data of all the same type of data has a unified data format;

[0044] Setting data content writing logic, and executing the operation of writing data content into the data storage system after format conversion of each data content of the same type is completed based on the data content writing logic;

[0045] Among them, the types of data content include video, picture, audio, and characters. The formats of video data include: MP4, AVI, MOV, WMV, FLV, MKV, and 3GP; the formats of picture data include: JPEG / JPG, PNG, GIF, BMP, TIFF / TIF, PSD, SVG, WebP, ICO, and PCX; the formats of audio data include: MP3, WMA, WAV, ASF, AAC, VQF, FLAC, APE, MIDI, and AMR; the formats of character data include: ASCII, UTF-8, UTF-16, UTF-32, GBK, Latin-1, Unicode, Base64, and URL encoding; the data content writing logic includes: in the data content classified based on the data type, data with small data volume is written first, or characters are written first before audio, audio is written first before pictures, and pictures are written first before video.

[0046] Compared with the known public technology, the technical solution provided by the present invention has the following beneficial effects:

[0047] The present invention provides a data storage system and a data writing method. During operation, the system creates a data pool to distinguish and store data content through fine classification of data content in the data storage system, and in this process, eliminates duplicate data based on data similarity analysis, thereby avoiding invalid occupation of storage space in the data storage system. At the same time, key data information is marked and the data pool used for data content storage is named, thereby realizing reliability management of data storage and providing a faster and more efficient service for the data storage system when used for data retrieval and retrieval. At the same time, based on the data writing method, the data storage system is controlled through data management during the data writing stage to ensure that the data written into the data storage system can be written and stored more quickly. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0049] Figure 1 It is a structural schematic diagram of a data storage system;

[0050] Figure 2 The figure is a flowchart of a data writing method. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution 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 creative work are within the scope of protection of the present invention.

[0052] The present invention will be further described below in conjunction with the embodiments.

[0053] Embodiment 1:

[0054] A data storage system of this embodiment, such as Figure 1 As shown, including:

[0055] A traversal module, used to traverse the data content stored in the storage system;

[0056] During the traversal module operation phase, the storage system does not perform data writing operations. The traversal module is internally provided with submodules, including:

[0057] The identification unit is used to run synchronously with the traversal module running phase to identify the similarities between groups of data in the storage system;

[0058] Among them, during the operation phase of the traversal module, the recognition unit synchronously obtains the recognition results of the similarity between each group of data, and eliminates the same data in the storage system based on the recognition results. The data content stored in the storage system traversed by the traversal module is all data content that has been processed by the recognition unit for similarity recognition;

[0059] When the recognition unit recognizes the similarity between each group of data in the storage system, the similarity recognition logic between the two groups of data contents is expressed as follows:

[0060]

[0061] Where: SIMM(X,Y) is the similarity between data X and data Y; n is the set of elements in data X; m is the set of elements in data Y; p(x i ,y j ) is the element pair (x i ,y j ) occurs; p(x i ) is the probability of the i-th element in the data X appearing; p(y j ) is the probability of the jth element in the data Y; γ is the correction;

[0062] The larger the SIMM(X,Y) value is, the higher the similarity between the two sets of data is. When SIMM(X,Y)≥95%, the two sets of data are judged to be the same data.

[0063] The modified γ value used in the similarity recognition logic between two sets of data contents obeys:

[0064]

[0065] Where: μ X , μ Y is the mean of data X and data Y; σ XY is the covariance of data X and data Y; C1 and C2 are constants; σ X , σ Y is the standard deviation of data X and data Y;

[0066] Among them, constants C1 and C2 are both greater than 0

[0067] A classification module, used to classify the data content stored in the data storage system based on similarity;

[0068] The classification module is divided into submodules, including:

[0069] A setting unit, used for setting a similarity classification interval when the classification module is running to classify the data stored in the data storage system;

[0070] A marking unit, used for marking the data content classified by the classification module;

[0071] The similarity classification interval set in the setting unit is customized by the system end user, and the initial default similarity classification interval is 40%, that is, data content with a similarity of not less than 40% is classified as similar data. When the marking unit marks the classified data content, the same part of the data of each group of data in the same data is used as the marking content to mark each data in the same data, and the similarity between each group of data in the same data is not less than 40%;

[0072] A creation module is used to create a data pool in the data storage system, and to differentiate and store the data content classified by the classification module based on the data pool;

[0073] A retrieval module, used for editing text information and performing a data retrieval operation in a data storage system using the edited text information as retrieval content;

[0074] A queue module is used to monitor the number of times data content in each data pool in the data storage system is retrieved, evaluate the activity of the data content based on the number of times the data content is retrieved, and arrange the data content in each data pool using the activity of the data content;

[0075] The data content activity in the queue module is expressed as:

[0076]

[0077] Where: f(a) is the activity of data content a; q is the number of times data content a is retrieved during the operation cycle of the queue module; t is the average time for data content a to be retrieved once during the operation cycle of the queue module;

[0078] Based on the above formula, the corresponding activity of each data content in the data pool is obtained. The queue module arranges the data content in the data pool in descending order based on the activity of the data content, so that the data content with high activity is at the front position of the queue, and the data content with low activity is at the back position of the queue;

[0079] The queue module operation cycle set during the control unit operation phase follows:

[0080] When the data content at the front position of the data content queue is retrieved less than the number of times it was retrieved in the previous operation cycle within the same time threshold as the previous operation cycle, the queue module refreshes and runs once;

[0081] A control unit, used to set the queue module operation cycle, control the queue module to run continuously based on the operation cycle, and queue and refresh the data in each data pool in the data storage system;

[0082] The traversal module is interactively connected to an identification unit through a wireless network, the traversal module is interactively connected to a classification module through a wireless network, the classification module is interactively connected to a setting unit and a marking unit through a wireless network, the classification module is interactively connected to a creation module through a wireless network, the creation module is interactively connected to the marking unit through a wireless network, the creation module is interactively connected to a retrieval module through a wireless network, and the creation module is interactively connected to a queue module and a control unit through a wireless network.

[0083] In this embodiment, the traversal module runs to traverse the data content stored in the storage system, the identification unit runs synchronously in the traversal module running stage, identifies the similarity between each group of data in the storage system, the classification module further classifies the data content stored in the data storage system based on the similarity, the setting unit synchronously sets the similarity classification interval when the classification module runs to classify the data stored in the data storage system, the marking unit marks the data content classified by the classification module in real time, and then the creation module creates a data pool in the data storage system, and the data content classified by the classification module is distinguished and stored based on the data pool. The retrieval module further edits the text information, and performs the data retrieval operation in the data storage system with the edited text information as the retrieval content. The queue module post-operates to monitor the number of times the data content in each data pool in the data storage system is retrieved, the data content activity is evaluated based on the number of times the data content is retrieved, and the data content in each data pool is arranged using the data content activity, and the queue module operation cycle is synchronously set by the control unit, and the queue module is controlled to run continuously based on the operation cycle to arrange the queue refresh of the data in each data pool in the data storage system.

[0084] Through the operation of the system in the above embodiment, further refined data storage logic is provided to the data storage system, ensuring that the data stored in the data storage system is easier to retrieve and manage.

[0085] Embodiment 2:

[0086] The number of data pools created in the creation module is equal to the number of data categories in the classification module. Each group of data pools is used to store a group of similar data. Each group of data pools is named based on the tag content of the data content stored therein;

[0087] The operation of editing text information in the retrieval module is performed by the system-side user. After the text information used for retrieval is edited, the retrieval module first uses the data pool as the retrieval target to search for the data pool with the same name as the edited text information. After determining the data pool, the system-side user applies the retrieval module to edit the text information again. The retrieval module performs the retrieval operation of the data content in the determined data pool based on the text information edited again by the system-side user.

[0088] In this embodiment, through the above settings, further operation data support is provided for the operation of the system in Embodiment 1, ensuring the stability of the operation of the system in Embodiment 1.

[0089] Embodiment 3:

[0090] A data writing method comprises the following steps:

[0091] Traverse the data content that needs to be written into the data storage system, and identify the data types in the data content, that is, the format of the data corresponding to each data type;

[0092] Identify the data size ratio of each format of data in the data corresponding to each data content;

[0093] Select the data format with the largest proportion among the same type of data as the data conversion format, and perform format conversion on the corresponding data of all data formats except the data format with the largest proportion among the same type of data, so that the data of all the same type of data has a unified data format;

[0094] Setting data content writing logic, and executing the operation of writing data content into the data storage system after format conversion of each data content of the same type is completed based on the data content writing logic;

[0095] Among them, the types of data content include video, picture, audio, and character. The formats of video data include: MP4, AVI, MOV, WMV, FLV, MKV, and 3GP; the formats of picture data include: JPEG / JPG, PNG, GIF, BMP, TIFF / TIF, PSD, SVG, WebP, ICO, and PCX; the formats of audio data include: MP3, WMA, WAV, ASF, AAC, VQF, FLAC, APE, MIDI, and AMR; the formats of character data include: ASCII, UTF-8, UTF-16, UTF-32, GBK, Latin-1, Unicode, Base64, and URL encoding; the data content writing logic includes: in the data content classified based on the data type, data with small data volume is written first, or characters are written first before audio, audio is written first before pictures, and pictures are written first before video.

[0096] In summary, in the above embodiment, the system creates a data pool to distinguish the stored data content through fine-grained classification of the data content in the data storage system, and in this process, eliminates duplicate data based on data similarity analysis, thereby avoiding invalid occupation of storage space in the data storage system. At the same time, key data information is marked and the data pool used for data content storage is named, thereby realizing reliability management of data storage, providing the data storage system with faster and more efficient services when used for data retrieval and retrieval, and based on the data writing method, the data storage system is controlled through data management during the data writing stage to ensure that the data written into the data storage system can be written and stored more quickly.

[0097] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A data storage system, characterized in that: include: A traversal module, used to traverse the data content stored in the storage system; A classification module, used to classify the data content stored in the data storage system based on similarity; A creation module is used to create a data pool in the data storage system, and to differentiate and store the data content classified by the classification module based on the data pool; A retrieval module, used for editing text information and performing a data retrieval operation in a data storage system using the edited text information as retrieval content; A queue module is used to monitor the number of times data content in each data pool in the data storage system is retrieved, evaluate the activity of the data content based on the number of times the data content is retrieved, and arrange the data content in each data pool using the activity of the data content; The control unit is used to set the operation cycle of the queue module, control the queue module to run continuously based on the operation cycle, and arrange the queues and refresh the data in each data pool in the data storage system.

2. A data storage system according to claim 1, characterized in that: During the operation phase of the traversal module, the storage system does not perform data writing operations. The traversal module is internally provided with submodules, including: The identification unit is used to run synchronously with the traversal module running phase to identify the similarities between groups of data in the storage system; Among them, during the operation phase of the traversal module, the recognition unit synchronously obtains the recognition results of the similarities between each group of data, and eliminates the same data in the storage system based on the recognition results. The data contents stored in the storage system traversed by the traversal module are all data contents that have been processed by the recognition unit for similarity recognition.

3. A data storage system according to claim 2, characterized in that: When the identification unit identifies the similarity between each group of data in the storage system, the similarity identification logic between the two groups of data contents is expressed as follows: Where: SIMM(X,Y) is the similarity between data X and data Y; n is the set of elements in data X; m is the set of elements in data Y; p(x i ,y j ) is the element pair (x i ,y j ) occurs; p(x i ) is the probability of the i-th element in the data X appearing; p(y j ) is the probability of the jth element in the data Y; γ is the correction; The larger the SIMM(X, Y) value is, the higher the similarity between the two sets of data is. When SIMM(X, Y)≥95%, the two sets of data are determined to be the same data.

4. A data storage system according to claim 3, characterized in that: The modified γ value used in the similarity recognition logic between the two sets of data contents obeys: Where: μ X , μ Y is the mean of data X and data Y; σ XY is the covariance of data X and data Y; C1 and C2 are constants; σ X , σ Y is the standard deviation of data X and data Y; Among them, constants C1 and C2 are both greater than 0.

5. A data storage system according to claim 1, characterized in that: The classification module is provided with submodules at the lower level, including: A setting unit, used for setting a similarity classification interval when the classification module is running to classify the data stored in the data storage system; A marking unit, used for marking the data content classified by the classification module; Among them, the similarity classification interval set in the setting unit is customized by the system user, and the initial default similarity classification interval is 40%, that is, data content with a similarity of not less than 40% is classified as similar data. When the marking unit marks the classified data content, the same part of the data in each group of data in the same data is used as the marking content, and each data in the same data is marked, and the similarity between each group of data in the same data is not less than 40%.

6. A data storage system according to claim 1, characterized in that: The number of data pools created in the creation module is equal to the number of data categories in the classification module. Each group of data pools is used to store a group of similar data, and each group of data pools is named based on the tag content of the data content stored therein.

7. A data storage system according to claim 1, characterized in that: The operation of editing text information in the retrieval module is performed by the system-side user. After the text information used for retrieval is edited, the retrieval module first uses the data pool as the retrieval target to search for the data pool with the same name as the edited text information. After determining the data pool, the system-side user applies the retrieval module to edit the text information again. The retrieval module performs a retrieval operation on the data content in the determined data pool based on the text information edited again by the system-side user.

8. A data storage system according to claim 1, characterized in that: The data content activity in the queue module is expressed as: Where: f(a) is the activity of data content a; q is the number of times data content a is retrieved during the operation cycle of the queue module; t is the average time for data content a to be retrieved once during the operation cycle of the queue module; Based on the above formula, the corresponding activity of each data content in the data pool is obtained. The queue module arranges the data content in the data pool in descending order based on the activity of the data content, so that the data content with high activity is at the front position of the queue, and the data content with low activity is at the back position of the queue; The queue module operation cycle set during the operation phase of the control unit is subject to: When the front-most data content in the data content queue is retrieved less times than the number of times in the previous operation cycle within the same time threshold as the previous operation cycle, the queue module refreshes and runs once.

9. A data storage system according to claim 1, characterized in that: The traversal module is interactively connected to an identification unit via a wireless network, the traversal module is interactively connected to a classification module via a wireless network, the classification module is interactively connected to a setting unit and a marking unit via a wireless network, the classification module is interactively connected to a creation module via a wireless network, the creation module is interactively connected to the marking unit via a wireless network, the creation module is interactively connected to a retrieval module via a wireless network, and the creation module is interactively connected to a queue module and a control unit via a wireless network.

10. A data writing method, the method being an implementation method of a data storage system as claimed in any one of claims 1 to 9, characterized in that: The following steps are involved: Traverse the data content that needs to be written into the data storage system, and identify the data types in the data content, that is, the format of the data corresponding to each data type; Identify the data size ratio of each format of data in the data corresponding to each data content; Select the data format with the largest proportion among the same type of data as the data conversion format, and perform format conversion on the corresponding data of all data formats except the data format with the largest proportion among the same type of data, so that the data of all the same type of data has a unified data format; Setting data content writing logic, and executing the operation of writing data content into the data storage system after format conversion of each data content of the same type is completed based on the data content writing logic; Among them, the types of data content include video, picture, audio, and characters. The formats of video data include: MP4, AVI, MOV, WMV, FLV, MKV, and 3GP; the formats of picture data include: JPEG / JPG, PNG, GIF, BMP, TIFF / TIF, PSD, SVG, WebP, ICO, and PCX; the formats of audio data include: MP3, WMA, WAV, ASF, AAC, VQF, FLAC, APE, MIDI, and AMR; the formats of character data include: ASCII, UTF-8, UTF-16, UTF-32, GBK, Latin-1, Unicode, Base64, and URL encoding; the data content writing logic includes: in the data content classified based on the data type, data with small data volume is written first, or characters are written first before audio, audio is written first before pictures, and pictures are written first before video.

Images