CN111708903A - Method, device, electronic device and storage medium for time-consuming optimization of image search questions - Google Patents

Abstract

Embodiments of the present invention relate to the technical field of intelligent devices, and disclose a method, an apparatus, an electronic device and a storage medium for time-consuming optimization of a picture search question. The method includes: obtaining a target picture; when receiving a question search instruction, obtaining the size of the memory occupied by the target picture, and when the size of the memory occupied by the target picture is greater than or equal to a preset threshold Compression is performed according to a predetermined compression rate to obtain a compressed picture; the question is searched by using the compressed picture. By implementing the embodiments of the present invention, the memory occupancy rate of the picture can be reduced by compressing the picture, the corresponding speed of recognition and question search can be improved, and the user experience can be improved.

Description

Method, device, electronic device and storage medium for time-consuming optimization of image search questions

technical field

The present invention relates to the technical field of intelligent devices, in particular to a method, device, electronic device and storage medium for time-consuming optimization of image search questions.

Background technique

Most of the current electronic teaching aids have point-to-read scenes. The point-to-read scene means that when a user points to a carrier such as a book, exercise book, or test paper with his finger, the teaching aid device will take a picture of the carrier through an image acquisition device, and recognize the finger. position, so as to determine the user's intention according to the finger position, and then obtain the image corresponding to the user's intention, which is used for searching questions, etc. When searching for a question, the general method is to search for the text-related content after the OCR recognition in the resource library or the Internet after the image is recognized by OCR. When the image used in the search question occupies a large amount of memory, the recognition and search response speed is slow, which affects the effect of user experience.

SUMMARY OF THE INVENTION

In view of the above-mentioned defects, the embodiments of the present invention disclose a method, device, electronic device and storage medium for optimizing the time-consuming of image search, which can compress the memory occupancy rate of the target image and improve the corresponding speed of search.

A first aspect of the embodiments of the present invention discloses a time-consuming optimization method for image search questions, and the method includes:

Get the target image;

When the question search instruction is received, the size of the memory occupied by the target picture is obtained, and when the size of the memory occupied by the target picture is greater than or equal to a preset threshold, the target picture is compressed according to a predetermined compression rate, and the result is obtained Compress image;

Use the compressed picture to search for questions.

As an optional implementation manner, in the first aspect of the embodiment of the present invention, before the acquisition of the target picture, the method further includes:

The predetermined compression rate is determined according to the correct rate of the predetermined search questions through the training sample set.

As an optional implementation manner, in the first aspect of the embodiment of the present invention, the predetermined compression rate is determined according to the predetermined search question accuracy rate through the training sample set, including:

Each sample in the training sample set is compressed with different compression rates to obtain compressed samples;

Determine the text recognition rate of the compressed samples at each compression rate;

When the character recognition rate is equal to the character recognition rate threshold, the predetermined compression rate is determined, and the predetermined compression rate is a compression rate corresponding to a compressed sample equal to the character recognition rate threshold.

As an optional implementation manner, in the first aspect of the embodiment of the present invention, the method for determining the character recognition rate threshold includes:

Construct a linear function of text recognition rate and search accuracy rate;

The character recognition rate threshold is determined according to the preset search question accuracy rate and the linear function.

As an optional implementation manner, in the first aspect of the embodiment of the present invention, when the character recognition rate is equal to the character recognition rate threshold, determining the predetermined compression rate includes:

Determine the target compression rate when the text recognition rate of each compressed sample is equal to the text recognition rate threshold;

All target compression ratios are clustered, and a target cluster set is determined, and the target cluster set is the set of the largest target compression ratios in the clusters;

The average value of all target compression ratios in the target cluster set is obtained, and the average value is used as the predetermined compression ratio.

As an optional implementation manner, in the first aspect of the embodiment of the present invention, using the compressed picture to search for questions includes:

Perform OCR recognition on the compressed image to obtain text recognition information;

Using the character recognition information, the resource database or the Internet is searched according to the search instruction, and the search result is obtained.

As an optional implementation manner, in the first aspect of the embodiment of the present invention, the predetermined compression ratio includes a size compression ratio or/and a quality compression ratio.

A second aspect of the embodiments of the present invention discloses an apparatus for optimizing time-consuming image search questions, and the apparatus includes:

an acquisition unit, used to acquire the target image;

The compression unit is used to obtain the size of the memory occupied by the target picture when receiving the question search instruction, and when the size of the memory occupied by the target picture is greater than or equal to a preset threshold, compress the target picture according to a predetermined Compression rate to obtain a compressed image;

A question searching unit, used for searching questions by using the compressed picture.

As an optional implementation manner, in the second aspect of the embodiment of the present invention, the device further includes:

The training unit is used to determine the predetermined compression rate according to the correct rate of the predetermined search question through the training sample set.

As an optional implementation manner, in the second aspect of the embodiment of the present invention, the training unit includes:

The model building subunit is used to compress each sample in the training sample set with different compression rates to obtain compressed samples;

The recognition rate determination subunit is used to determine the text recognition rate of the compressed samples under each compression rate;

The compression rate determination subunit is configured to determine the predetermined compression rate when the character recognition rate is equal to the character recognition rate threshold, the predetermined compression rate being the compression rate corresponding to the compressed samples equal to the character recognition rate threshold.

As an optional implementation manner, in the second aspect of the embodiment of the present invention, the recognition rate determination subunit includes:

The first grandson unit is used to construct a linear function of text recognition rate and search accuracy rate;

The second grandchild unit is configured to determine the character recognition rate threshold according to the preset search question accuracy rate and the linear function.

As an optional implementation manner, in the second aspect of the embodiment of the present invention, the compression ratio determination subunit includes:

The third grandchild unit is used to determine the target compression rate when the character recognition rate of each compressed sample is equal to the character recognition rate threshold;

The fourth grandchild unit is used for clustering all target compression ratios, and determines a target cluster set, which is the set of the largest target compression ratios in the cluster;

The fifth grandchild unit is used to obtain the average value of all target compression ratios in the target cluster set, and use the average value as the predetermined compression ratio.

As an optional implementation manner, in the second aspect of the embodiment of the present invention, the question search unit includes:

an identification subunit, used to perform OCR identification on the compressed picture to obtain text identification information;

The search subunit is used to search the resource library or the Internet according to the search instruction by using the character identification information, and obtain the search result.

A third aspect of the embodiments of the present invention discloses an electronic device, comprising: a memory storing executable program codes; a processor coupled to the memory; the processor calling the executable program codes stored in the memory , which is used to execute part or all of the steps of the method for time-consuming optimization of a picture search question disclosed in the first aspect of the embodiments of the present invention.

A fourth aspect of the embodiments of the present invention discloses a computer-readable storage medium, which stores a computer program, wherein the computer program enables a computer to execute the method for time-consuming optimization of a picture search question disclosed in the first aspect of the embodiments of the present invention some or all of the steps.

A fifth aspect of the embodiments of the present invention discloses a computer program product that, when the computer program product runs on a computer, causes the computer to execute the method for optimizing time-consuming image search questions disclosed in the first aspect of the embodiments of the present invention some or all of the steps.

A sixth aspect of the embodiments of the present invention discloses an application publishing platform, and the application publishing platform is used for publishing a computer program product, wherein when the computer program product runs on a computer, the computer is made to execute the first embodiment of the present invention. In one aspect, part or all of the steps of a method for time-consuming optimization of image search questions are disclosed.

Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

In the embodiment of the present invention, a target picture is obtained; when a question search instruction is received, the size of the memory occupied by the target picture is obtained, and when the size of the memory occupied by the target picture is greater than or equal to a preset threshold, the size of the memory occupied by the target picture is obtained. The target picture is compressed according to a predetermined compression rate to obtain a compressed picture; the question is searched by using the compressed picture. It can be seen that, by implementing the embodiments of the present invention, the memory occupancy rate of the picture can be reduced by compressing the picture, the corresponding speed of identification and question search can be improved, and the user experience can be improved.

Description of drawings

In order to explain the technical solutions in the embodiments of the present invention more clearly, the following briefly introduces the accompanying drawings that need to be used in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained from these drawings without any creative effort.

1 is a schematic flowchart of a method for time-consuming optimization of picture search questions disclosed in an embodiment of the present invention;

2 is a schematic flowchart of another method for time-consuming optimization of picture search questions disclosed in an embodiment of the present invention;

3 is a schematic structural diagram of a device for optimizing time-consuming image search questions disclosed in an embodiment of the present invention;

FIG. 4 is a schematic structural diagram of an electronic device disclosed in an embodiment of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

It should be noted that the terms "first", "second", "third", "fourth", etc. in the description and claims of the present invention are used to distinguish different objects, rather than to describe specific order. The terms "comprising" and "having" and any variations thereof in the embodiments of the present invention are intended to cover non-exclusive inclusion, for example, a process, method, apparatus, product or device comprising a series of steps or units is not necessarily limited to Those steps or elements that are expressly listed may instead include other steps or elements that are not expressly listed or are inherent to the process, method, product or apparatus.

The embodiments of the present invention disclose a method, device, electronic device and storage medium for time-consuming optimization of picture search questions, which can reduce the memory occupancy rate of pictures by compressing pictures, improve the corresponding speed of identification and question search, and improve user experience , which will be described in detail below with reference to the accompanying drawings.

Example 1

Please refer to FIG. 1. FIG. 1 is a schematic flowchart of a method for optimizing time-consuming image search questions disclosed in an embodiment of the present invention. As shown in Figure 1, the time-consuming optimization method for the image search question includes the following steps:

110. Obtain a target image.

The target image is an image determined according to the user's intention. Exemplarily, in the point-and-read scenario, the user touches the carrier with his hand and triggers a corresponding photographing instruction, and the carrier is photographed by the image capture device, thereby identifying the user's intention. Recognizing user intent can be achieved through neural network algorithms, such as identifying text lines according to the PSENet algorithm, and then determining the fingertip coordinates according to the skin color segmentation algorithm, so as to determine the text line or topic line where the fingertip coordinates are located, and according to the text line where the fingertip coordinates are located. Or the subject line is used to segment the photographed image to obtain the target image.

There are various instructions for triggering a photo, which can be exemplarily implemented by voice, such as "please take a photo" or "how to pronounce this word (what does this word mean)", etc., where the former voice mode only triggers the photo-taking operation, and the subsequent User intent requires a new command, and the latter voice method is to trigger a photographing command through a question search command. It is also possible to trigger the corresponding photographing command by pressing a button (such as a mechanical button or a touch button), or you can enter a reading scene on an electronic device (mainly refers to teaching aids such as tutoring machines, learning machines, and reading machines), such as a reading app. Trigger the camera command.

The carrier is a paper book, exercise book, workbook or test paper, etc. The image acquisition device is a device that performs a photographing function, and it can be integrated on the electronic device, for example, the carrier placed in front of the electronic device can be photographed through the front camera of the electronic device, or it can be a discrete device, which can be connected with the electronic device through wired or wireless means. The electronic device establishes a communication connection, executes the photographing instruction sent by the electronic device, and sends the image obtained by photographing to the electronic device.

120. When receiving a question search instruction, obtain the size of the memory occupied by the target picture, and compress the target picture according to a predetermined compression rate when the size of the memory occupied by the target picture is greater than or equal to a preset threshold. , to get the compressed image.

The search instruction is the user's intention operation. The search instruction can be a search answer, such as "how to solve this question", or it can be a pronunciation search, such as "how to pronounce this word", or a search explanation, such as "what does this word mean? ", can also search for synonyms or antonyms, such as "what is the synonym (antonym) of this word" and so on.

Because the target picture occupies a certain amount of memory during the question search process, when the target picture occupies a large amount of memory, it will affect the response speed of the device, thereby affecting the corresponding speed of identification and question search. In the embodiment of the present invention, whether to compress the target image is determined according to the size of the memory occupied by the target image.

The preset threshold can be set according to the type of the question search instruction. When the question search instruction is to search pronunciation, explanation, synonyms or antonyms, etc., the preset threshold can be set to be relatively small, so as to respond quickly and improve the user experience. When searching for answers, etc., the preset threshold can be set relatively large. On the one hand, when searching for answers, there are generally many calculation questions, so the target image itself is large, and compression and identification itself take more time; When answering, the user has a certain tolerance for the search time, and also leaves a certain time for the user to think.

The predetermined compression rate can be set as required, and the predetermined compression rate can be combined with the text recognition rate, that is, according to the preset text recognition rate, the text recognition rate determined after corresponding compression of the target image and the preset text recognition rate set When they are equivalent, it can be considered that the corresponding compression ratio during the compression is a predetermined compression ratio. The text recognition rate is the text recognition rate of the compressed compressed image, and the ratio of the number of characters obtained by text recognition to the number of characters of the target image itself is the text recognition rate. If the search speed needs to be improved, and the accuracy of the search is not too high, the preset text recognition rate can be appropriately reduced, that is, the predetermined compression rate can be increased. On the contrary, if the search accuracy is required to be high, the search speed is required. Generally, the preset character recognition rate can be appropriately increased, that is, the predetermined compression rate can be decreased.

130. Use the compressed picture to search for a question.

For the uncompressed target picture, when executing the question search command, the question is still searched with the original target picture, and for the compressed target picture, when executing the question search command, the question is searched with the compressed picture.

The search method is similar to the prior art, that is, first perform OCR identification on the picture (original target picture or compressed picture), determine the text identification information after identification, that is, the text content, and then use the text content and search instructions in the resource library or the Internet. Search for the appropriate content.

Repositories can be pre-created based on different subjects, grades or other databases, created as needed. The content corresponding to the search is determined by the search instructions at all levels, which can be voice (when searching for how to read), answer (when searching for how to do), or characters (when searching for synonyms or antonyms).

By implementing the embodiments of the present invention, the memory occupancy rate of the picture can be reduced by compressing the picture, the corresponding speed of recognition and question search can be improved, and the user experience can be improved.

Embodiment 2

Please refer to FIG. 2 . FIG. 2 is a schematic flowchart of another method for optimizing time-consuming image search questions disclosed in an embodiment of the present invention. As shown in Figure 2, the time-consuming optimization method for the image search question includes the following steps:

210. Obtain a target image.

220. When receiving a question search instruction, obtain the size of the memory occupied by the target picture, and compress the target picture according to a predetermined compression rate when the size of the memory occupied by the target picture is greater than or equal to a preset threshold. , to get the compressed image.

230. Use the compressed picture to search for a question.

Steps 210 and 230 may be the same as steps 110 and 130 in the first embodiment, and will not be repeated here.

In step 220, the question search instruction is an operation intended by the user, and the question search instruction can be an answer search, such as "how to solve this question", or a pronunciation search, such as "how to pronounce this word", or an explanation, such as "this What does the word mean?”, it can also be a search for synonyms or antonyms, such as “what is the synonym (antonym) of this word” and so on.

Because the target picture occupies a certain amount of memory during the question search process, when the target picture occupies a large amount of memory, it will affect the response speed of the device, thereby affecting the corresponding speed of identification and question search. In the embodiment of the present invention, whether to compress the target image is determined according to the size of the memory occupied by the target image.

The preset threshold can be set according to the type of the question search instruction. When the question search instruction is to search pronunciation, explanation, synonyms or antonyms, etc., the preset threshold can be set to be relatively small, so as to respond quickly and improve the user experience. When searching for answers, etc., the preset threshold can be set relatively large. On the one hand, when searching for answers, there are generally many calculation questions, so the target image itself is large, and compression and identification itself take more time; When answering, the user has a certain tolerance for the search time, and also leaves a certain time for the user to think.

The predetermined compression rate is obtained according to pre-training, and the predetermined compression rate includes the size compression rate and/or the quality compression rate. The two are combined through the training model to obtain a compression model, and the compression model is trained to determine the size compression rate and quality compression rate. value of .

Size compression will change the size of the image, that is, compress the pixels of the width and height of the image, which is mainly used to reduce the memory size of the image. The quality compression is to change the bit depth and transparency of the picture under the premise of maintaining the pixels to compress the picture, which is mainly used to reduce the storage size of the picture. In the embodiment of the present invention, since the memory occupancy rate of the target picture is mainly reduced, when obtaining the predetermined compression rate, the proportion of the size compression rate can be set to be relatively larger, for example, P=0.95P ₁ +0.05P ₂ ; Among them, P is the predetermined compression ratio, and P1 and _P2 are the size compression ratio and the quality compression _ratio , respectively.

The compression model can be combined with existing compression algorithms such as Luban image compression algorithm, or existing quality compression algorithms such as MediaStore.Images.Media.getBitmap or BitmapFactory.decodeStream and existing size compression algorithms such as BitmapFactory.decodeFile. Set the weights to get the compressed model. Of course, the dimensional compression ratio and the mass compression ratio can also be determined through a limited number of experiments according to the above-mentioned formula for the predetermined compression ratio.

The compression model is trained through the training sample set, and the predetermined compression rate is determined according to the correct rate of the predetermined search questions. Specifically, it may include the following steps:

221. Compress each sample in the training sample set through different compression rates to obtain compressed samples.

Each sample is input into the compression model, and by implementing the set different sampling rates (for size compression ratios) and quality compression ratios (in this embodiment of the present invention, the quality compression ratio is set to 1-(the difference between the compressed image and the original image). quality ratio), for example, if the quality compression ratio is 10%, the quality of the compressed image is 90% of the original image quality) to determine the compression samples. If the offline compression ratio is used, no less than 20 samples can be set for each sample. For different compression ratios, if continuous compression ratios are used, the continuous compression ratios may be sampled so that each sample determines not less than 20 different sampling points.

222. Determine the character recognition rate of the compressed sample under each compression rate.

Character recognition rate refers to the ratio of the number of characters obtained in the compressed sample input OCR recognition model to the actual total number of characters in the compressed sample.

223. When the character recognition rate is equal to the character recognition rate threshold, determine the predetermined compression rate, where the predetermined compression rate corresponds to the compression rate of the compressed samples equal to the character recognition rate threshold.

Theoretically, the greater the image compression rate, the lower the text recognition rate and the lower the accuracy of search questions. Therefore, the image compression rate can be determined by the search question accuracy rate, and a balance point between the search question accuracy rate and the image compression rate can be sought. The linear function of the text recognition rate and the search question accuracy rate can be constructed through multiple experiments. The linear function can be fitted by the least squares method, and the linear function can be performed piecewise. According to the coordinate system corresponding to the specific recognition rate and accuracy rate The location of the midpoint sets the piecewise linear function. Users generally have certain expectations about the accuracy of search questions, which are determined according to different types of search questions. When the search questions are searched for pronunciation, explanation, synonyms or antonyms, etc., the search accuracy rate is generally higher, and for search answers, etc. , a certain error tolerance rate can be allowed, and the search question accuracy rate can be set lower. Therefore, the threshold of the text recognition rate can be determined according to the preset search question accuracy rate and the constructed linear function, and then the predetermined compression rate can be determined according to the threshold of the text recognition rate.

The text recognition rate threshold can be a range. Determine the target compression rate when the text recognition rate of each compressed sample is equal to the text recognition rate threshold, that is, the determined target compression rate for text recognition, and the number of recognized characters is the total number of characters in the target image. The proportion of the number of characters is within the range of the character recognition rate threshold.

The predetermined compression ratio can be determined by mathematical expectation (average value), that is, all target compression ratios are averaged, and then the average value is taken as the predetermined compression ratio. Under the premise that there is a certain error in the compression model, the error can also be eliminated by clustering. First, all target compression rates are clustered to determine the number of target compression rates in each cluster set. The so-called clustering is to take all target compression ratios within a range of a certain target compression ratio ± a preset value as a cluster to form a cluster set. Find the set of target compression ratios with the largest number of clusters from these cluster sets, record it as the target cluster set, calculate the average of all target compression ratios in the target cluster set, and use the average value as the predetermined compression ratio .

By implementing the embodiments of the present invention, the memory occupancy rate of the picture can be reduced by compressing the picture, the corresponding speed of recognition and question search can be improved, and the user experience can be improved.

Embodiment 3

Please refer to FIG. 3 . FIG. 3 is a schematic structural diagram of an apparatus for optimizing time-consuming image search questions according to an embodiment of the present invention. As shown in Figure 3, the device for optimizing the time-consuming image search question may include:

an obtaining unit 320, configured to obtain a target picture;

The compressing unit 330 is configured to obtain the size of the memory occupied by the target picture when receiving the question search instruction, and when the size of the memory occupied by the target picture is greater than or equal to a preset threshold, compress the target picture according to a predetermined The compression rate is compressed to obtain a compressed image;

The question search unit 340 is configured to use the compressed picture to search for questions.

As an optional implementation manner, the device further includes:

The training unit 310 is configured to determine a predetermined compression rate according to a predetermined correct rate of search questions through a training sample set.

As an optional implementation manner, the training unit 310 includes:

The model construction subunit 311 is used for compressing each sample in the training sample set with different compression rates to obtain compressed samples;

The recognition rate determination subunit 312 is used to determine the character recognition rate of the compressed samples under each compression rate;

The compression rate determination subunit 313 is configured to determine the predetermined compression rate when the character recognition rate is equal to the character recognition rate threshold, the predetermined compression rate being the compression rate corresponding to the compressed samples equal to the character recognition rate threshold.

As an optional implementation manner, the recognition rate determination subunit 312 includes:

The first grandson unit 3121 is used to construct a linear function of the character recognition rate and the accuracy rate of search questions;

The second grandchild unit 3122 is configured to determine the character recognition rate threshold according to the preset search question accuracy rate and the linear function.

As an optional implementation manner, the compression ratio determination subunit 313 includes:

The third grandchild unit 3131 is used to determine the target compression rate when the character recognition rate of each compressed sample is equal to the character recognition rate threshold;

The fourth grandchild unit 3132 is used to cluster all target compression ratios, and determine a target cluster set, which is the set of the largest target compression ratios in the cluster;

The fifth grandchild unit 3133 is configured to obtain the average value of all target compression ratios in the target cluster set, and use the average value as the predetermined compression ratio.

As an optional implementation manner, the question search unit 340 includes:

The identification subunit 341 is used to perform OCR identification on the compressed picture to obtain text identification information;

The search subunit 342 is configured to use the character recognition information to search in the resource library or the Internet according to the search instruction to obtain the search result.

As shown in FIG. 3 , the device for optimizing the time-consuming of searching questions for pictures can reduce the memory occupancy rate of the pictures by compressing the pictures, improve the corresponding speed of recognition and question searching, and improve the user experience.

Embodiment 4

Please refer to FIG. 4 , which is a schematic structural diagram of an electronic device disclosed in an embodiment of the present invention. As shown in Figure 4, the electronic device may include:

a memory 410 storing executable program code;

a processor 420 coupled to the memory 410;

Wherein, the processor 420 invokes the executable program code stored in the memory 410 to execute some or all of the steps in the method for optimizing the time consumption of a picture search question in the first embodiment.

An embodiment of the present invention discloses a computer-readable storage medium, which stores a computer program, wherein the computer program causes a computer to execute some or all of the steps in the method for time-consuming optimization of a picture search question in the first embodiment.

The embodiment of the present invention also discloses a computer program product, wherein when the computer program product runs on the computer, the computer is made to execute some or all of the steps in the method for time-consuming optimization of image search in the first embodiment.

The embodiment of the present invention also discloses an application publishing platform, wherein the application publishing platform is used for publishing a computer program product, wherein, when the computer program product runs on a computer, the computer is made to execute the time-consuming optimization of image search in the first embodiment. some or all of the steps in the method.

In various embodiments of the present invention, it should be understood that the size of the sequence numbers of the described procedures does not imply a necessary order of execution, and the execution order of each procedure should be determined by its functions and internal logic, and does not deal with the present invention. The implementation of the embodiments constitutes no limitation.

The units described as separate components may or may not be physically separated, and components displayed as units may or may not be object units, and may be located in one place or distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution in this embodiment.

In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit. The integrated unit may be implemented in the form of hardware, or may be implemented in the form of software functional units.

The integrated unit, if implemented as a software functional unit and sold or used as a stand-alone product, may be stored in a computer-accessible memory. Based on such understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a memory , including several requests to cause a computer device (which may be a personal computer, a server, or a network device, etc., specifically a processor in the computer device) to execute some or all of the steps of the methods described in the various embodiments of the present invention.

In the embodiments provided by the present invention, it should be understood that "B corresponding to A" means that B is associated with A, and B can be determined according to A. However, it should also be understood that determining B according to A does not mean that B is only determined according to A, and B may also be determined according to A and/or other information.

Those of ordinary skill in the art can understand that some or all of the steps in the various methods of the embodiments can be completed by instructing the relevant hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium includes only Read-Only Memory (ROM), Random Access Memory (RAM), Programmable Read-only Memory (PROM), Erasable Programmable Read-Only Memory, EPROM), One-time Programmable Read-Only Memory (OTPROM), Electronically Erasable Programmable Read-Only Memory (EEPROM), CD-ROM ( CompactDisc Read-Only Memory, CD-ROM) or other optical disk storage, magnetic disk storage, tape storage, or any other computer-readable medium that can be used to carry or store data.

A method, device, electronic device, and storage medium for time-consuming optimization of picture search questions disclosed in the embodiments of the present invention have been described above in detail. In this paper, specific examples are used to illustrate the principles and implementations of the present invention. The description of the example is only used to help understand the method of the present invention and its core idea; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific embodiment and the scope of application. As stated, the contents of this specification should not be construed as limiting the present invention.

Claims (15)

1. A method for optimizing time consumption of picture searching questions is characterized by comprising the following steps:

acquiring a target picture;

when a question searching instruction is received, the size of a memory occupied by the target picture is obtained, and when the size of the memory occupied by the target picture is larger than or equal to a preset threshold value, the target picture is compressed according to a preset compression rate to obtain a compressed picture;

and searching questions by adopting the compressed picture.

2. The method according to claim 1, wherein before the obtaining the target picture, further comprising:

and determining a preset compression rate according to the preset search question accuracy rate through the training sample set.

3. The method of claim 2, wherein the determining the predetermined compression rate according to the predetermined search question correction rate by the training sample set comprises:

compressing each sample in the training sample set through different compression rates to obtain a compressed sample;

determining the character recognition rate of the compressed sample under each compression rate;

and when the character recognition rate is equal to a character recognition rate threshold value, determining the preset compression rate, wherein the preset compression rate is the compression rate corresponding to the compression sample equal to the character recognition rate threshold value.

4. The method of claim 3, wherein the text recognition rate threshold determination method comprises:

constructing a linear function of the character recognition rate and the question searching accuracy rate;

and determining a character recognition rate threshold according to a preset question searching accuracy and the linear function.

5. The method of claim 3, wherein determining the predetermined compression rate when the rate of word recognition equals a rate of word recognition threshold comprises:

determining a target compression rate when the character recognition rate of each compressed sample is equal to a character recognition rate threshold;

clustering all target compression ratios, and determining a target cluster set, wherein the target cluster set is the set of the largest number of target compression ratios in the clusters;

and acquiring the average value of all target compression ratios in the target clustering set, and taking the average value as a preset compression ratio.

6. The method according to any one of claims 1-5, wherein using the compressed picture for performing a search, comprises:

performing OCR recognition on the compressed picture to obtain character recognition information;

and searching in a resource library or the Internet according to the question searching instruction by using the character recognition information to obtain a question searching result.

7. The method according to any of claims 1-5, wherein the predetermined compression ratio comprises a size compression ratio or/and a quality compression ratio.

8. The utility model provides a picture is searched for subject and is consumed time optimization's device which characterized in that includes:

an acquisition unit configured to acquire a target picture;

the compression unit is used for acquiring the size of the memory occupied by the target picture when a question searching instruction is received, and compressing the target picture according to a preset compression rate when the size of the memory occupied by the target picture is larger than or equal to a preset threshold value to obtain a compressed picture;

and the question searching unit is used for searching questions by adopting the compressed picture.

9. The apparatus of claim 8, further comprising:

and the training unit is used for determining a preset compression rate according to the preset search question accuracy rate through the training sample set.

10. The apparatus of claim 9, wherein the training unit comprises:

the model building subunit is used for compressing each sample in the training sample set through different compression rates to obtain a compressed sample;

the recognition rate determining subunit is used for determining the character recognition rate of the compressed sample under each compression rate;

and the compression rate determining subunit is configured to determine the predetermined compression rate when the text recognition rate is equal to a text recognition rate threshold, where the predetermined compression rate is a compression rate corresponding to a compressed sample equal to the text recognition rate threshold.

11. The apparatus of claim 10, wherein the recognition rate determining subunit comprises:

the first grandchild unit is used for constructing a linear function of the character recognition rate and the question searching accuracy rate;

and the second grandchild unit is used for determining a character recognition rate threshold according to the preset accuracy rate of searching the questions and the linear function.

12. The apparatus of claim 10, wherein the compression ratio determining subunit comprises:

the third grandchild unit is used for determining a target compression rate when the character recognition rate of each compressed sample is equal to the character recognition rate threshold;

the fourth sun unit is used for clustering all target compression ratios and determining a target clustering set, wherein the target clustering set is the set of the target compression ratios with the largest quantity in the clusters;

and the fifth grandchild unit is used for acquiring the average value of all target compression ratios in the target cluster set and taking the average value as the preset compression ratio.

13. The apparatus according to any one of claims 8-12, wherein the search unit comprises:

the identification subunit is used for performing OCR (optical character recognition) on the compressed picture to obtain character recognition information;

and the searching subunit is used for searching in a resource library or the Internet according to the search question instruction by using the character identification information to obtain a search question result.

14. An electronic device, comprising: a memory storing executable program code; a processor coupled with the memory; the processor calls the executable program code stored in the memory to execute a method for optimizing the time for searching the problems of the pictures according to any one of claims 1 to 7.

15. A computer-readable storage medium storing a computer program, wherein the computer program causes a computer to execute a method for optimizing picture search time according to any one of claims 1 to 7.

Images