Logging image definition recognition method and device, medium, and electronic equipment

1 . A logging image definition recognition method, comprising:

establishing a logging image sample library comprising a plurality of logging images;

acquiring actual definition information corresponding to the respective logging images, the actual definition information being an actual definition sorting sequence number or an actual normalized definition;

acquiring a plurality of definitions corresponding to the respective logging images, the plurality of definitions being generated by a plurality of target image definition determination algorithms respectively performing definition calculations on the logging images;

determining target weights corresponding to the respective target image definition determination algorithms according to the plurality of definitions and the actual definition information corresponding to the respective logging images; and

determining a definition of a target logging image by the respective target image definition determination algorithms and the target weights corresponding to the respective target image definition determination algorithms.

2 . The method according to claim 1 , wherein the determining the target weights corresponding to the respective target image definition determination algorithms according to the plurality of definitions and the actual definition information corresponding to the respective logging images comprises:

establishing definition vectors corresponding to the respective logging images by the plurality of definitions corresponding to the respective logging images, the definition vectors comprising normalized definitions corresponding to the respective target image definition determination algorithms;

initializing a weight vector, the weight vector comprising weights corresponding to the respective target image definition determination algorithms;

constructing a plurality of training units by the weight vector and the definition vectors corresponding to the respective logging images, each of the training units comprising definition vectors corresponding to two logging images respectively and the weight vector;

executing a weight adjustment step, the weight adjustment step comprising: for each of the training units, comparing elements in the definition vectors in each of the training units, and adjusting weights of the weight vector in the training units according to the comparison results;

determining an average of the weight vectors in the respective training units to obtain a final weight vector;

calculating and obtaining final definitions corresponding to the respective logging images by the final weight vector and the definition vectors corresponding to the respective logging images;

determining whether the final definitions corresponding to the respective logging images match the actual definition information corresponding to the respective logging images;

if yes, utilizing the weights in the final weight vector as the target weights corresponding to the respective target image definition determination algorithms, otherwise executing the weight adjustment step and subsequent steps until the final definitions corresponding to the respective logging images match the actual definition information corresponding to the respective logging images.

3 . The method according to claim 2 , wherein the training units comprise a first definition vector and a second definition vector, the weight adjustment step further comprises:

if a number, in which elements in the first definition vector are greater than corresponding elements in the second definition vector, is more than half of a total number of elements in the first definition vector or the second definition vector, then increasing the respective target weights in the weight vector or decreasing respective non-target weights in the weight vector, until the sum of the products of the weight vector and corresponding elements in the first definition vector is greater than the sum of the products of the weight vector and corresponding elements in the second definition vector, wherein the elements in the first definition vector that are larger than the corresponding elements of the second definition vector correspond to the target image definition determination algorithms corresponding to the target weights, and the non-target weights are weights in the weight vector other than the target weights.

4 . The method according to claim 3 , wherein the establishing the logging image sample library comprising the plurality of logging images comprises:

acquiring unrepaired logging images and repaired logging images corresponding to the respective unrepaired logging images;

establishing the logging image sample library by the unrepaired logging images and the repaired logging images.

5 . The method according to claim 2 , wherein the actual definition information is an actual definition sorting sequence number, and the determining whether the final definitions corresponding to the respective logging images match the actual definition information corresponding to the respective logging images comprises:

sorting the final definitions corresponding to the respective logging images to obtain a final definition sorting sequence numbers corresponding to the respective logging images;

if, for any logging image, the final definition sorting sequence number corresponding to that logging image is consistent with the actual definition sorting sequence number corresponding to that logging image, then determining that the final definitions corresponding to the respective logging images match the actual definition information corresponding to the respective logging images, otherwise determining that the final definitions corresponding to the respective logging images do not match the actual definition information corresponding to the respective logging images.

6 . The method according to claim 5 , wherein before acquiring the plurality of definitions corresponding to the respective logging images, the method further comprises:

for each logging image in the logging image sample library, respectively determining definitions of that logging image by a plurality of image definition determination algorithms;

selecting a plurality of target image definition determination algorithms among the plurality of image definition determination algorithms according to the definitions determined by a plurality of image definition determination algorithms and the actual definition sorting sequence numbers corresponding to the respective logging images.

7 . The method according to claim 6 , wherein the establishing the logging image sample library comprising the plurality of logging images comprises:

acquiring unrepaired logging images and repaired logging images corresponding to the respective unrepaired logging images;

establishing the logging image sample library by the unrepaired logging images and the repaired logging images.

8 . The method according to claim 5 , wherein the establishing the logging image sample library comprising the plurality of logging images comprises:

acquiring unrepaired logging images and repaired logging images corresponding to the respective unrepaired logging images;

establishing the logging image sample library by the unrepaired logging images and the repaired logging images.

9 . The method according to claim 2 , wherein the establishing the logging image sample library comprising the plurality of logging images comprises:

acquiring unrepaired logging images and repaired logging images corresponding to the respective unrepaired logging images;

establishing the logging image sample library by the unrepaired logging images and the repaired logging images.

10 . The method according to claim 2 , wherein the plurality of target image definition determination algorithms comprise one or more of the following algorithms: a Brenner algorithm, a Tenengrad algorithm, a Laplacian algorithm, an SMD algorithm, an SMD2 algorithm, a variance algorithm, an energy algorithm, a Vollath algorithm.

11 . The method according to claim 1 , wherein the establishing the logging image sample library comprising the plurality of logging images comprises:

acquiring unrepaired logging images and repaired logging images corresponding to the respective unrepaired logging images;

establishing the logging image sample library by the unrepaired logging images and the repaired logging images.

12 . The method according to claim 1 , wherein the plurality of target image definition determination algorithms comprise one or more of the following algorithms: a Brenner algorithm, a Tenengrad algorithm, a Laplacian algorithm, an SMD algorithm, an SMD2 algorithm, a variance algorithm, an energy algorithm, a Vollath algorithm.

13 . An electronic equipment, wherein the electronic equipment comprises:

at least one processor;

a memory, on which computer-readable instructions are stored, the computer-readable instructions, when executed by the processor, cause the at least one processor to perform operations, the operations comprising:

establishing a logging image sample library comprising a plurality of logging images;

acquiring actual definition information corresponding to the respective logging images, the actual definition information being an actual definition sorting sequence number or an actual normalized definition;

acquiring a plurality of definitions corresponding to the respective logging images, the plurality of definitions being generated by a plurality of target image definition determination algorithms respectively performing definition calculations on the logging images;

determining target weights corresponding to the respective target image definition determination algorithms according to the plurality of definitions and the actual definition information corresponding to the respective logging images; and

determining a definition of a target logging image by the respective target image definition determination algorithms and the target weights corresponding to the respective target image definition determination algorithms.

14 . The electronic equipment according to claim 13 , wherein the determining the target weights corresponding to the respective target image definition determination algorithms according to the plurality of definitions and the actual definition information corresponding to the respective logging images comprises:

establishing definition vectors corresponding to the respective logging images by the plurality of definitions corresponding to the respective logging images, the definition vectors comprising normalized definitions corresponding to the respective target image definition determination algorithms;

initializing a weight vector, the weight vector comprising weights corresponding to the respective target image definition determination algorithms;

constructing a plurality of training units by the weight vector and the definition vectors corresponding to the respective logging images, each of the training units comprising definition vectors corresponding to two logging images respectively and the weight vector;

executing a weight adjustment step, the weight adjustment step comprising: for each of the training units, comparing elements in the definition vectors in each of the training units, and adjusting weights of the weight vector in the training units according to the comparison results;

determining an average of the weight vectors in the respective training units to obtain a final weight vector;

calculating and obtaining final definitions corresponding to the respective logging images by the final weight vector and the definition vectors corresponding to the respective logging images;

determining whether the final definitions corresponding to the respective logging images match the actual definition information corresponding to the respective logging images;

if yes, utilizing the weights in the final weight vector as the target weights corresponding to the respective target image definition determination algorithms, otherwise executing the weight adjustment step and subsequent steps until the final definitions corresponding to the respective logging images match the actual definition information corresponding to the respective logging images.

15 . The electronic equipment according to claim 14 , wherein the training units comprise a first definition vector and a second definition vector, the weight adjustment step further comprises:

if a number, in which elements in the first definition vector are greater than corresponding elements in the second definition vector, is more than half of a total number of elements in the first definition vector or the second definition vector, then increasing the respective target weights in the weight vector or decreasing respective non-target weights in the weight vector, until the sum of the products of the weight vector and corresponding elements in the first definition vector is greater than the sum of the products of the weight vector and corresponding elements in the second definition vector, wherein the elements in the first definition vector that are larger than the corresponding elements of the second definition vector correspond to the target image definition determination algorithms corresponding to the target weights, and the non-target weights are weights in the weight vector other than the target weights.

16 . The electronic equipment according to claim 14 , wherein the actual definition information is an actual definition sorting sequence number, and the determining whether the final definitions corresponding to the respective logging images match the actual definition information corresponding to the respective logging images comprises:

sorting the final definitions corresponding to the respective logging images to obtain a final definition sorting sequence numbers corresponding to the respective logging images;

if, for any logging image, the final definition sorting sequence number corresponding to that logging image is consistent with the actual definition sorting sequence number corresponding to that logging image, then determining that the final definitions corresponding to the respective logging images match the actual definition information corresponding to the respective logging images, otherwise determining that the final definitions corresponding to the respective logging images do not match the actual definition information corresponding to the respective logging images.

17 . The electronic equipment according to claim 16 , wherein before acquiring the plurality of definitions corresponding to the respective logging images, the method further comprises:

for each logging image in the logging image sample library, respectively determining definitions of that logging image by a plurality of image definition determination algorithms;

selecting a plurality of target image definition determination algorithms among the plurality of image definition determination algorithms according to the definitions determined by a plurality of image definition determination algorithms and the actual definition sorting sequence numbers corresponding to the respective logging images.

18 . The electronic equipment according to claim 13 , wherein the establishing the logging image sample library comprising the plurality of logging images comprises:

acquiring unrepaired logging images and repaired logging images corresponding to the respective unrepaired logging images;

establishing the logging image sample library by the unrepaired logging images and the repaired logging images.

19 . The electronic equipment according to claim 13 , wherein the plurality of target image definition determination algorithms comprise one or more of the following algorithms: a Brenner algorithm, a Tenengrad algorithm, a Laplacian algorithm, an SMD algorithm, an SMD2 algorithm, a variance algorithm, an energy algorithm, a Vollath algorithm.

20 . A non-volatile computer-readable storage medium, in which at least one executable instruction is stored, the executable instruction being configured to cause a processor to perform operations of:

establishing a logging image sample library comprising a plurality of logging images;

acquiring actual definition information corresponding to the respective logging images, the actual definition information being an actual definition sorting sequence number or an actual normalized definition;

acquiring a plurality of definitions corresponding to the respective logging images, the plurality of definitions being generated by a plurality of target image definition determination algorithms respectively performing definition calculations on the logging images;

determining target weights corresponding to the respective target image definition determination algorithms according to the plurality of definitions and the actual definition information corresponding to the respective logging images; and

determining a definition of a target logging image by the respective target image definition determination algorithms and the target weights corresponding to the respective target image definition determination algorithms.