A computer hard disk data recovery method

Abstract

The application discloses a computer hard disk data recovery method and relates to the technical field of computers. The method comprises the following steps: if the lost part of data in target hard disk data is continuous lost data, then the importance weight ratio of the continuous lost data and the to-be-tested ratio of the data length to which the continuous lost data belongs are counted; the non-lost data adjacent to the position of the continuous lost data is extracted from the target hard disk data; and the associated data of the continuous lost data is speculated; if the lost part of data in the target hard disk data is non-continuous lost data, and the non-continuous lost data is a feature in the weight data segment set and the auxiliary hardware is not more than two, then the associated data of the continuous lost data is speculated according to the non-lost data adjacent to the position of the non-continuous lost data in the target hard disk data. The computer hard disk data recovery method provided by the application can improve the data recovery accuracy of the computer hard disk.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computers, in particular to a computer hard disk data recovery method. BACKGROUND

[0002] In the information age, the important assets of individuals and enterprises gradually change from entities to digital forms (such as documents, videos, databases, etc.), and once data loss occurs, it may cause great losses.

[0003] In the traditional technology, the way of computer hard disk data recovery is relatively single, which is often to repair the missing data by obtaining backup data from other backup hardware, the processing way is relatively single, and in order to do further repair data verification, it is easy to cause large error of data recovery, resulting in low efficiency of final data recovery work. SUMMARY

[0004] In order to overcome the shortcomings of the prior art, the present application provides a computer hard disk data recovery method.

[0005] The computer hard disk data recovery method provided by the present application comprises:

[0006] Step S1, obtaining target hard disk data, if the missing part of the target hard disk data is continuous missing data, calculating the importance weight ratio of the continuous missing data in the target hard disk data to obtain a to-be-tested weight value, and calculating the to-be-tested ratio of the data length of the continuous missing data to the total length of the target hard disk data, extracting the non-missing data adjacent to the position of the continuous missing data from the target hard disk data according to the to-be-tested weight value and the to-be-tested ratio, and performing associated data speculation on the continuous missing data to obtain first pre-processing repair data;

[0007] Step S2, if the missing part of the target hard disk data is non-continuous missing data, obtaining other associated hardware, extracting auxiliary measurement hardware from the other associated hardware, which is not covered by new data when the original data transmission of the target hard disk data is stopped immediately under the condition of data loss of the target hard disk data, performing weight data segment distribution feature detection on the non-continuous missing data, if it is a weight data segment centralized feature, and when the auxiliary measurement hardware is not more than two, combining the pre-processing reference data one of the auxiliary measurement hardware and the non-missing data adjacent to the position of the non-continuous missing data in the target hard disk data, performing associated data speculation on the continuous missing data to obtain second pre-processing repair data;

[0008] Step S3, if it is a weight data segment decentralized feature, and when the auxiliary measurement hardware is not less than three, combining the pre-processing reference data two of the auxiliary measurement hardware to perform associated data speculation on the continuous missing data to obtain third pre-processing repair data.

[0009] Preferably, the target hard drive data where data loss has occurred is acquired. If the lost data in the target hard drive data is continuously lost, the continuous data is weighted. If the continuously lost data is weighted, the test situation is output.

[0010] Based on the test conditions, if it is detected that after immediately stopping data writing in the event of data loss on the target hard drive, no data is overwritten in the target hard drive, then the importance weight ratio of continuously lost data to the target hard drive data is calculated to obtain the test weight value.

[0011] Preferably, the ratio of the length of the continuously lost data to the total length of the target hard disk data is calculated to obtain the ratio to be tested;

[0012] The average value of the weight to be measured and the ratio to be measured are averaged to obtain the average ratio. A preset discrimination ratio threshold is set. If the average ratio is less than the discrimination ratio threshold, the average ratio is increased by a multiple to obtain the expansion ratio.

[0013] Preferably, the ratio of the length of the continuously lost data to the total length of the target hard disk data is calculated to obtain the ratio to be tested;

[0014] The average value of the weight to be measured and the ratio to be measured are averaged to obtain the average ratio. A preset discrimination ratio threshold is set. If the average ratio is less than the discrimination ratio threshold, the average ratio is increased by a multiple to obtain the expansion ratio.

[0015] Preferably, based on the expansion ratio, the non-lost data adjacent to the location of the continuously lost data is extracted from the target hard disk data, and the first adjacent non-lost data segment is output;

[0016] Based on the first adjacent unlost data segment, the continuously lost data is correlated and inferred to obtain the first preprocessed repair data.

[0017] Preferably, if the lost data in the target hard disk is non-continuous, other associated hardware that immediately stops data writing when data loss occurs in the target hard disk is detected is acquired. Each of the other associated hardware processes the original data transmission of the target hard disk, and the timing of data processing by each of the other associated hardware is affected by the latency of each hardware.

[0018] Auxiliary testing hardware is extracted from the other associated hardware. Auxiliary testing hardware refers to hardware that ensures the original data transmission of the target hard disk is not overwritten by new data when data writing is immediately stopped in the event of data loss on the target hard disk.

[0019] Preferably, the weighted data segment distribution characteristics are detected for non-continuous lost data. If the weighted data segment is concentrated and there are no more than two auxiliary testing hardware, the original data of the target hard disk data transmitted in the auxiliary testing hardware is processed by data transmission to obtain preprocessed reference data one. Based on preprocessed reference data one, the non-continuous lost data is repaired to obtain repair result one.

[0020] Extract the non-lost data adjacent to the non-contiguous lost data from the target hard disk data to obtain the second adjacent non-lost data. Based on the second adjacent non-lost data, perform correlation data inference on the non-contiguous lost data to obtain the second repair result.

[0021] The difference data of the first repair result and the second repair result are corrected to obtain the second preprocessed repair data.

[0022] Preferably, the weighted data segment distribution characteristics are detected for non-continuous lost data. If the weighted data segment is dispersed and there are no less than three auxiliary testing hardware devices, the original data of the target hard disk data transmitted in the auxiliary testing hardware is processed by data transmission to obtain preprocessed reference data two.

[0023] Select identical reference data from the second preprocessed reference data to obtain preprocessed backup data. Based on the preprocessed backup data, perform data comparison and repair on non-continuous lost data to obtain third preprocessed repair data.

[0024] Compared with the prior art, the present invention has the following characteristics and beneficial effects:

[0025] By analyzing the continuity of lost data segments on the target hard drive, the difficulty of data recovery can be assessed based on the continuity of these segments. If the data loss is continuous, the importance weight of the continuously lost data within the target hard drive's data is determined. This allows for the selection of corresponding length segments of unlost data based on the local characteristics of the continuously lost data, avoiding the need for comprehensive correlation analysis of all unlost data, which would lead to complex data processing, increased interference between data, and increased processing errors. Therefore, a certain length of data segment with the strongest correlation to the continuously lost data is extracted from the target hard drive data for correlation analysis to perform data recovery processing in one scenario. For non-continuous data loss, such as data segments with specific weight distribution characteristics... The characteristics of weighted data segment concentration and weighted data segment dispersion are used to differentiate between non-continuous data loss under these two conditions. The data recovery conditions for weighted data segment dispersion are more stringent because the weighted data segments are not concentrated enough, which increases the error in data recovery. Therefore, more accurate backup data is required for comparison and repair. This requires the use of other related hardware, that is, data backup can be performed on other related hardware, and the backup data of the related hardware is analyzed and compared according to the two weighted data segment distribution characteristics of weighted data segment concentration and weighted data segment dispersion. Through the above processing method, in the case of hard disk data loss, a variety of recovery solutions can be selected, expanding the choice and making full use of the correlation between hardware. Attached Figure Description

[0026] Figure 1 This is a flowchart illustrating the steps of a computer hard drive data recovery method, which is the main feature of this embodiment. Detailed Implementation

[0027] The present invention will be further described in detail below with reference to the following embodiments.

[0028] Reference Figure 1 A method for recovering data from a computer hard drive, comprising the following steps:

[0029] Step S1: Obtain target hard disk data. If the lost data in the target hard disk data is continuously lost, calculate the importance weight ratio of continuously lost data to the target hard disk data to obtain the weight value to be tested. Calculate the ratio of the length of the continuously lost data to the total length of the target hard disk data to the weight value to be tested. Based on the weight value to be tested and the ratio to be tested, extract the non-lost data adjacent to the location of the continuously lost data from the target hard disk data. Perform correlation data inference on the continuously lost data to obtain the first preprocessed repair data.

[0030] Step S2: If the lost data in the target hard disk is non-continuous data loss, then obtain other associated hardware, extract the auxiliary testing hardware that, when data writing is immediately stopped in the event of data loss in the target hard disk, ensures that the original data transmission of the target hard disk data is not overwritten by new data, and perform weighted data segment distribution feature detection on the non-continuous data loss. If it is a weighted data segment concentration feature, and there are no more than two auxiliary testing hardware, then combine the preprocessing reference data one to which the auxiliary testing hardware belongs with the target hard disk data to extract the non-lost data adjacent to the non-continuous data loss position, and perform associated data inference on the continuous data loss to obtain the second preprocessing repair data.

[0031] Step S3: If the weighted data segment is scattered and there are at least three auxiliary test hardware devices, then combine the preprocessed reference data two to which the auxiliary test hardware belongs to perform correlation data inference on the continuously lost data to obtain the third preprocessed repair data.

[0032] Specifically, by analyzing the continuity of lost data segments on the target hard drive, the difficulty of data recovery can be assessed based on the continuity of these segments. If the data loss is continuous, the importance weight of this continuous data segment relative to the target hard drive's data is determined. This allows for the selection of corresponding length segments of unlost data based on the local characteristics of the continuously lost data, avoiding the need for comprehensive correlation analysis of all unlost data, which would lead to complex data processing, increased interference between data, and increased processing errors. Therefore, a certain length of data segment with the strongest correlation to the continuously lost data is extracted from the target hard drive data for correlation analysis to perform data recovery processing under one scenario. For non-continuous data loss, such as weighted data segments... The distribution characteristics are categorized into concentrated and dispersed weighted data segments. Differential analysis is performed on non-continuous data loss under these two conditions. Data recovery under dispersed weighted data segments faces stricter limitations because the insufficient concentration of weighted data segments increases recovery errors. Therefore, more accurate backup data is required for comparison and repair. This necessitates the use of other associated hardware, allowing for data backup on these devices. The analysis and comparison of backup data from associated hardware is then performed based on the two weighted data segment distribution characteristics (concentrated and dispersed). This approach enhances the selection of diverse recovery solutions in the event of hard drive data loss, expanding the options and fully utilizing the interrelationships between hardware components.

[0033] The specific step S1 includes the following sub-steps:

[0034] Obtain the target hard drive data where data loss has occurred. If the lost data in the target hard drive is continuous, then determine the weight of the continuous data. If the continuous data loss is weighted data, then output the test situation.

[0035] Based on the test conditions, if it is detected that after immediately stopping data writing in the event of data loss on the target hard drive, no data is overwritten in the target hard drive, then the importance weight ratio of continuously lost data to the target hard drive data is calculated to obtain the test weight value.

[0036] The ratio of the length of continuously lost data to the total length of data on the target hard drive is calculated to obtain the ratio to be measured.

[0037] The average ratio is obtained by averaging the weight value to be tested and the ratio to be tested. A threshold for the discrimination ratio is preset. If the average ratio is less than the threshold for the discrimination ratio, the average ratio is increased by a multiple to obtain the expansion ratio.

[0038] Based on the expansion ratio, extract the unlost data adjacent to the location of the continuously lost data from the target hard disk data, and output the first adjacent unlost data segment.

[0039] Based on the first adjacent unlost data segment, the continuously lost data is correlated and inferred to obtain the first preprocessed repair data.

[0040] Specifically, this could involve data loss on the target hard drive (e.g., office documents, logs, images, etc., partially lost due to hard drive failure, etc.), continuously lost data (e.g., market research statistics, such as W1, W2, W3, W4, W5, W6; if W4 and W5 are lost, and they are adjacent, this is considered continuously lost data), conditions to be tested (e.g., based on the weight of the keywords related to the data's theme (the percentage of occurrences and the percentage of data contained; the higher the percentage of occurrences and the higher the percentage of data contained, the higher the weight ratio)), and the weight value to be tested (e.g., if data writing is immediately stopped after data loss is detected on the target hard drive, the target hard drive data...). The data in question is not overwritten: for example, when a file is deleted or a partition is formatted, the operating system only marks the corresponding space as "available" and does not immediately erase the data. In this case, as long as it is not completely overwritten by new data, recovery can still be achieved by scanning for residual information fragments. However, once partial overwriting occurs (such as when a new file is written to the same area), the integrity of the original data is compromised, making it difficult to directly restore using traditional methods. Therefore, this invention primarily addresses data recovery processing under the condition that lost data has not been overwritten by new data. The weight value to be measured refers to the sum of the percentage of occurrences and the percentage of data contained in W4 and W5. If it is h, then similarly, a comprehensive statistical analysis is performed on W1, W2, W3, W4, W5, and W6. If it is... H), then h / H is the weight value to be measured), the ratio to be measured (for example, the length of the data segment of W4 and W5 is d1+d2, if it is c, and the total length of W1, W2, W3, W4, W5, and W6 is L, then c / L is the ratio to be measured), the average ratio (i.e., (h / H+c / L) / 2, if it is z / Z), the preset discrimination ratio threshold (e.g., K, used to determine whether the length needs to be increased or decreased when selecting some data segments of the non-lost data segments later, which can be set by yourself), the expansion ratio (e.g., the multiplier expansion refers to a multiple not exceeding two times, i.e., 2z / Z, it should be noted that 2z is less than Z), the first adjacent non-lost data segment (e.g., according to the ratio value of 2z / Z, from the target hard disk data (W1, W2, W3, W6) Select the unlost data segments of length (2z / Z)*C. If these are W2 and W3, it should be noted that the total length of W2 and W3 is not less than c. Since W4 and W5 are weighted data, in order to improve the correlation accuracy of subsequent data recovery processing, the length of the selected data segments should not be less than W4 and W5, nor should the selected data segments be too long, which will cause data redundancy and interfere with data processing. The first preprocessing repair data (for example, when a file is deleted or a partition is formatted, the original data on its physical sectors does not disappear immediately; the operating system only marks the area as "available." If it is not overwritten by new data, the content of the lost part can be inferred by analyzing the surrounding undamaged data blocks (such as file headers, tail checksums, adjacent byte patterns, etc.).For example, some documents with fixed formats (such as Excel spreadsheets) have a preset structural framework, and even if a section is damaged, the missing content can be reconstructed based on the header row and column width information. For instance, a scan first reveals that the beginning of the document contains a table of contents structure "Chapter 1 Introduction," and the end contains a concluding section "In conclusion...". There are numerous consecutive whitespace clusters between them (marked as available space). Then, based on Chinese typesetting habits, it is inferred that the missing part should be multiple paragraphs of argumentative text. By comparing the frequency of keywords (such as "sample" and "data analysis"), punctuation usage patterns, and the number of characters per line in the preceding and following paragraphs, a probabilistic model is constructed to generate candidate text (historical data, textual information, keywords, and other feature information can be used as training text to construct the probabilistic model). Here, we consider data information, and so on.

[0041] The specific step S2 includes the following sub-steps:

[0042] If the lost data in the target hard drive is non-continuous, then other associated hardware that immediately stops writing data when data loss occurs in the target hard drive is detected. Each of the other associated hardware processes the original data transmission of the target hard drive, and the timing of data processing by each of the other associated hardware is affected by the latency of its own hardware.

[0043] Auxiliary testing hardware is extracted from other related hardware. Auxiliary testing hardware refers to hardware that ensures the original data transmission of the target hard drive is not overwritten by new data when data writing is immediately stopped in the event of data loss on the target hard drive.

[0044] For non-continuously lost data, the distribution characteristics of weighted data segments are detected. If the weighted data segments are concentrated and there are no more than two auxiliary testing hardware devices, the original data of the target hard disk data transmitted in the auxiliary testing hardware is processed for data transmission to obtain preprocessed reference data one. Based on preprocessed reference data one, the non-continuously lost data is repaired to obtain repair result one.

[0045] Extract the non-lost data adjacent to the non-contiguous lost data from the target hard drive data to obtain the second adjacent non-lost data. Based on the second adjacent non-lost data, perform correlation data inference on the non-contiguous lost data to obtain the second repair result.

[0046] The difference data of repair result 1 and repair result 2 are corrected to obtain the second preprocessed repair data.

[0047] Specifically, for example, non-contiguous data loss (e.g., W2, W4, and W6 are lost data, and their positions are not consecutively adjacent, thus constituting non-contiguous data loss), other related hardware (such as USB-HDD / SSD adapters (used when the original hard drive cannot be directly connected to the host (e.g., internal interface damage), converting it to a USB interface for easy connection to the computer for data reading or recovery), hard drive data recovery cards, etc., labeled A, B, C, and D. Because each piece of hardware has different degrees of latency during data processing, the latency characteristics of historical latency data (e.g., how much data capacity corresponds to how long) are considered. The latency duration (based on historical latency data and statistical analysis) is used to back up the original data (in terms of format, etc.) of the target hard drive into files A, B, C, and D. If the original data (in terms of format, etc.) is w1, w2, w3, w4, w5, and w6, these files are backed up as original data and stored in A, B, C, and D respectively (storing them in their original form improves data security). Based on the latency characteristics of A, B, C, and D (e.g., the corresponding latency duration based on data size, i.e., statistical analysis based on historical latency data), and the capacity of w1, w2, w3, w4, w5, and w6, when further processing is required... During data transmission processing, data processing is performed with an advance time t (and w1, w2, w3, w4, w5, and w6 are also updated in real time based on the target hard drive data) so that A, B, C, and D are processed simultaneously to obtain the latest backup data. Auxiliary testing hardware (if A and B, this means that w1, w2, w3, w4, w5, and w6 in A and B are all complete) is performed. Weighted data segment cluster characteristics (if W2, W4, and W6 are missing data, and W2 is weighted data, then it is non-continuous missing data) are tested. Preprocessing reference data one (i.e., A and B respectively process w1, w2, w3, and w4...) Data formats such as W1, W2, W3, W4, W5, and W6 are converted, compressed, and simplified to obtain data information in the same format as W1, W2, W3, W4, W5, and W6. Repair result one (i.e., the data processed by A and B are compared and filled with the content of W2, W4, and W6 respectively, resulting in two repair results G1 and G2). Repair result two (the same interpretation as the first preprocessed repair data, if it is G3). Second preprocessed repair data (i.e., the three data repair results G1, G2, and G3 are comprehensively compared and corrected (e.g., the same data is retained, and the different data is neutralized) to obtain the comprehensive repair result.

[0048] The specific step S3 includes the following sub-steps:

[0049] For non-continuously lost data, the distribution characteristics of weighted data segments are detected. If the weighted data segments are dispersed and there are at least three auxiliary testing hardware devices, the original data of the target hard disk data transmitted in the auxiliary testing hardware is processed to obtain preprocessed reference data two.

[0050] Select identical reference data from the second preprocessed reference data to obtain preprocessed backup data. Based on the preprocessed backup data, perform data comparison and repair on non-continuous lost data to obtain the third preprocessed repair data.

[0051] Specifically, the preprocessing parameters include: weighted data segment dispersion characteristics (e.g., W2, W4, and W6 are all weighted data, but they are not adjacent and continuous, which is the weighted data segment dispersion characteristic); preprocessed reference data two (if it is A, B, and C); preprocessed reference data two (with the same explanation as preprocessed reference data one); preprocessed backup data (if the results of processing B and C are the same, and the data of processing A has a deviation (it is highly likely that there is a data processing abnormality or data omission problem in some step of the data processing process), in order to reduce the error probability of subsequent data recovery, the data processed by B and C is selected); and third preprocessed repair data (i.e., the data processed by B and C are compared and filled with W2, W4, and W6 respectively, which yields two repair results. Finally, the two repair results are compared and corrected comprehensively (e.g., the same data is retained, and the different data is neutralized) to obtain the comprehensive repair result).

[0052] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A method for recovering data from a computer hard drive, characterized in that, Includes the following steps: Step S1: Obtain target hard disk data. If the lost data in the target hard disk data is continuously lost, calculate the importance weight ratio of continuously lost data to the target hard disk data to obtain the weight value to be tested. Calculate the ratio of the length of the continuously lost data to the total length of the target hard disk data to the weight value to be tested. Based on the weight value to be tested and the ratio to be tested, extract the non-lost data adjacent to the location of the continuously lost data from the target hard disk data. Perform correlation data inference on the continuously lost data to obtain the first preprocessed repair data. Step S2: If the lost data in the target hard disk is non-continuous data loss, then other associated hardware is obtained. From the other associated hardware, auxiliary testing hardware is extracted that immediately stops data writing when data loss occurs in the target hard disk and performs original data transmission of the target hard disk data without being overwritten by new data. Weighted data segment distribution feature detection is performed on the non-continuous data loss. If it is a weighted data segment concentration feature and there are no more than two auxiliary testing hardware, then the preprocessing reference data one to which the auxiliary testing hardware belongs and the target hard disk data are combined to extract the non-lost data adjacent to the non-continuous data loss position. Associated data inference is performed on the non-continuous data loss to obtain the second preprocessing repair data. Step S3: If the weighted data segment is scattered and there are at least three auxiliary test hardwares, then the non-continuous lost data is compared and repaired by combining the preprocessing reference data two to which the auxiliary test hardware belongs, and the third preprocessing repair data is obtained. Preprocessing reference data one and preprocessing reference data two refer to the data obtained after transmitting and processing the original data of the target hard disk data in the auxiliary test hardware.

2. The computer hard drive data recovery method according to claim 1, characterized in that, Step S1 includes: Acquire the target hard drive data where data loss has occurred. If the lost data in the target hard drive data is continuous data loss, then determine the weight data of the continuous data. If the continuous data loss is weight data, then output the test situation. Based on the test conditions, if it is detected that after immediately stopping data writing in the event of data loss on the target hard drive, no data is overwritten in the target hard drive, then the importance weight ratio of continuously lost data to the target hard drive data is calculated to obtain the test weight value.

3. The computer hard drive data recovery method according to claim 2, characterized in that, Step S1 also includes: The ratio of the length of continuously lost data to the total length of data on the target hard drive is calculated to obtain the ratio to be measured. The average value of the weight to be measured and the ratio to be measured are averaged to obtain the average ratio. A preset discrimination ratio threshold is set. If the average ratio is less than the discrimination ratio threshold, the average ratio is increased by a multiple to obtain the expansion ratio.

4. The computer hard drive data recovery method according to claim 3, characterized in that, Step S1 also includes: Based on the expansion ratio, extract the non-lost data adjacent to the location of the continuously lost data from the target hard disk data, and output the first adjacent non-lost data segment; Based on the first adjacent unlost data segment, the continuously lost data is correlated and inferred to obtain the first preprocessed repair data.

5. A computer hard drive data recovery method according to claim 4, characterized in that, Step S2 includes: If the lost data in the target hard drive is non-continuous, then other associated hardware that immediately stops data writing when data loss occurs in the target hard drive is detected is obtained. Each of the other associated hardware processes the original data transmission of the target hard drive, and the timing of data processing by each of the other associated hardware is affected by the latency of its own hardware. Auxiliary testing hardware is extracted from the other associated hardware. Auxiliary testing hardware refers to hardware that ensures the original data transmission of the target hard disk is not overwritten by new data when data writing is immediately stopped in the event of data loss on the target hard disk.

6. A method for recovering computer hard drive data according to claim 5, characterized in that, Step S2 also includes: For non-continuous data loss, the weighted data segment distribution characteristics are detected. If the weighted data segment is concentrated and there are no more than two auxiliary test hardware, the original data of the target hard disk data transmitted in the auxiliary test hardware is processed for data transmission to obtain preprocessed reference data one. Based on preprocessed reference data one, the non-continuous data loss is repaired to obtain repair result one. Extract the non-lost data adjacent to the non-contiguous lost data from the target hard disk data to obtain the second adjacent non-lost data. Based on the second adjacent non-lost data, perform correlation data inference on the non-contiguous lost data to obtain the second repair result. The difference data of the first repair result and the second repair result are corrected to obtain the second preprocessed repair data.

7. A computer hard drive data recovery method according to claim 6, characterized in that, Step S3 includes: For non-continuous lost data, the weighted data segment distribution characteristics are detected. If the weighted data segment is dispersed and there are no less than three auxiliary test hardware, the original data of the target hard disk data transmitted in the auxiliary test hardware is processed by data transmission to obtain preprocessed reference data two. Select identical reference data from the second preprocessed reference data to obtain preprocessed backup data. Based on the preprocessed backup data, perform data comparison and repair on non-continuous lost data to obtain third preprocessed repair data.

Images