CN110336943A - A kind of scene recognition method and device - Google Patents
A kind of scene recognition method and device Download PDFInfo
- Publication number
- CN110336943A CN110336943A CN201910596008.1A CN201910596008A CN110336943A CN 110336943 A CN110336943 A CN 110336943A CN 201910596008 A CN201910596008 A CN 201910596008A CN 110336943 A CN110336943 A CN 110336943A
- Authority
- CN
- China
- Prior art keywords
- frame image
- scene recognition
- characteristic value
- current frame
- similarity
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/10—Cameras or camera modules comprising electronic image sensors; Control thereof for generating image signals from different wavelengths
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/80—Camera processing pipelines; Components thereof
Abstract
This disclosure relates to a kind of scene recognition method and device, wherein, scene recognition method includes obtaining step, obtains the characteristic value of frame image to be identified, frame image to be identified includes current frame image and prior frame image, and prior frame image is the preceding nth frame image of current frame image;Judgment step judges whether to execute scene Recognition step, wherein similarity of the characteristic value similarity between the characteristic value of current frame image and the characteristic value of prior frame image according to comparative feature value similarity and threshold value;Scene Recognition step includes being based on current frame image, carries out scene Recognition.Pass through a kind of scene recognition method of the disclosure, it is possible to reduce scene Recognition number, and then power consumption is effectively reduced.
Description
Technical field
This disclosure relates to scene Recognition technical field, it is specifically related to a kind of scene recognition method and device.
Background technique
For some devices for having image-acquisition functions, such as have the mobile phone of shooting function, is often opening
After camera, starting camera preview, usage scenario identification technology makes classification to the scene of live preview, according to different scenes
Categorization results take different image exposure and processing mode, are presented to the preview effect that user perceives closer to human eye, with
And shoot the photo effect that human eye seems more true to nature.During camera preview, since camera is to each frame figure of capture
As data all do scene Recognition and result output, and then the power consumption of device is increased.
Summary of the invention
In order to overcome problems of the prior art, the disclosure provides a kind of scene recognition method and device.
In a first aspect, the embodiment of the present disclosure provides a kind of scene recognition method comprising, obtaining step obtains to be identified
The characteristic value of frame image, frame image to be identified include current frame image and prior frame image, and prior frame image is current frame image
Preceding nth frame image;Judgment step judges whether execution scene Recognition step according to comparative feature value similarity and threshold value,
Middle characteristic value similarity is the similarity between the characteristic value of current frame image and the characteristic value of prior frame image;Scene Recognition step
Suddenly include being based on current frame image, carry out scene Recognition.
In one example, judgment step includes, according to comparative feature value similarity and threshold value, if characteristic value similarity is greater than threshold
Value, then execute scene Recognition step;If characteristic value similarity is less than or equal to threshold value, scene Recognition step is not executed.
In one example, prior frame image is the preceding first frame image of current frame image.
In one example, obtaining step further include: obtain the acquisition moment of current frame image, and adopting according to current frame image
Collect the moment and it is previous execute scene Recognition step the execution moment, obtain current frame image the acquisition moment and previous execution scene
The time interval of identification step executed between the moment;Judgment step include: according to comparative feature value similarity and threshold value and
Compare time interval and interval threshold, judges whether to execute scene Recognition step;Scene Recognition step further include: obtain and save
Execute the execution moment of scene Recognition step.
In one example, time interval includes the first interval threshold;Judgment step include: according to comparative feature value similarity with
Threshold value, if characteristic value similarity be greater than threshold value, judge current frame image the acquisition moment and previous execution scene Recognition step
Execute the moment between time interval, if time interval be greater than or equal to the first interval threshold, execute scene Recognition step;
If time interval less than the first interval threshold, does not execute scene Recognition step.
In one example, interval threshold includes the second interval threshold;Judgment step include: according to comparative feature value similarity with
Threshold value judges that the acquisition moment of current frame image and previous execution scene know if characteristic value similarity is less than or equal to threshold value
The time interval of other step executed between the moment executes scene Recognition step if time interval is greater than the second interval threshold;
If time interval is less than or equal to the second interval threshold, scene Recognition step is not executed.
In one example, obtaining step further include: feature is carried out to current frame image and prior frame image by sample mode
Value is extracted, characteristic value of the characteristic value that will acquire as current frame image and prior frame image.
In one example, characteristic value includes the brightness value of frame image to be identified or the rgb value of frame image to be identified.
In one example, scene recognition method further include: parameter step is generated, based on the scene obtained in scene Recognition step
Recognition result generates the acquisition parameters that image taking is carried out for image collecting device.
Second aspect, the embodiment of the present disclosure provide a kind of scene Recognition device, which, which has, realizes above-mentioned first aspect
The function for the scene recognition method being related to.The function can also be executed corresponding soft by hardware realization by hardware
Part is realized.The hardware or software include one or more modules corresponding with above-mentioned function.
In one example, a kind of scene Recognition device includes obtaining module, for obtaining the characteristic value of frame image to be identified,
Frame image to be identified includes current frame image and prior frame image, and prior frame image is the preceding nth frame image of current frame image;Sentence
Disconnected module, is used for comparative feature value similarity and threshold value, and wherein characteristic value similarity similarity is the characteristic value of current frame image
Similarity between the characteristic value of prior frame image;Scene Recognition module: for the ratio according to characteristic value similarity and threshold value
Compared with as a result, the data based on current frame image, carry out scene Recognition.
The third aspect, the embodiment of the present disclosure provide a kind of electronic equipment, wherein electronic equipment includes: memory, for depositing
Storage instruction;And processor, the scene recognition method of the instruction execution first aspect for calling memory to store.
Fourth aspect, the embodiment of the present disclosure provide a kind of computer readable storage medium, wherein computer-readable storage medium
Matter is stored with computer executable instructions, and computer executable instructions when executed by the processor, execute the scene of first aspect
Recognition methods.
A kind of scene recognition method and device that the disclosure provides.In camera preview, scene recognition method passes through calculating
Characteristic value similarity between the characteristic value of current frame image and the characteristic value of prior frame image, and by this feature value similarity with
Threshold value is compared, and determines whether to carry out scene Recognition to current frame image according to the result of the comparison.If not needing to present frame
Image carries out scene Recognition, then can be directly using the scene of previous frame image as the scene of current frame image, and calls upper one
The acquisition parameters of frame image carry out current frame image shooting, to reach by the number for reducing scene Recognition and reduce camera consumption
The effect of electricity.
Detailed description of the invention
The following detailed description is read with reference to the accompanying drawings, above-mentioned and other purposes, the feature of disclosure embodiment
It will become prone to understand with advantage.In the accompanying drawings, embodiment of the present disclosure is shown by way of example rather than limitation,
Wherein:
Fig. 1 shows the scene recognition method schematic diagram of embodiment of the present disclosure offer;
Fig. 2 shows another scene recognition method schematic diagrames that the embodiment of the present disclosure provides;
Fig. 3 shows another scene recognition method schematic diagram of embodiment of the present disclosure offer;
Fig. 4 shows a kind of scene Recognition schematic device of embodiment of the present disclosure offer;
Fig. 5 shows a kind of electronic equipment schematic diagram of embodiment of the present disclosure offer.
Specific embodiment
The principle and spirit of the disclosure are described below with reference to several illustrative embodiments.It should be appreciated that providing this
A little embodiments are used for the purpose of making those skilled in the art can better understand that realizing the disclosure in turn, and be not with any
Mode limits the scope of the present disclosure.
Although being noted that the statements such as " first " used herein, " second " to describe implementation of the disclosure mode not
Same module, step and data etc., still the statement such as " first ", " second " is merely in different modules, step and data etc.
Between distinguish, and be not offered as specific sequence or significance level.In fact, the statements such as " first ", " second " are complete
It may be used interchangeably.
Fig. 1 is a kind of scene recognition method schematic diagram that the embodiment of the present disclosure provides.As shown in Figure 1, the recognition methods 100
Including obtaining step S101, judgment step S102 and scene Recognition step S103.
As a kind of possible embodiment, the characteristic value of frame image to be identified can be obtained by obtaining step S101.Its
In, frame image to be identified includes current frame image and prior frame image, and prior frame image is the preceding nth frame figure of current frame image
Picture.Using judgment step S102, i.e., according to comparative feature value similarity and threshold value, judge whether to execute scene Recognition step.Its
In, characteristic value similarity refers to, according to the characteristic value of the characteristic value of current frame image and prior frame image, obtains between the two
Characteristic value similarity, wherein characteristic value similarity can be characteristic value difference between the two.
By scene Recognition step S103, i.e., according to current frame image, scene knowledge is carried out to the scene of the current frame image
Not.
It should be noted that can be carried out by neural network recognization algorithm to current frame image in actual application
Scene Recognition.Prior frame image is the preceding nth frame image of current frame image, and " N " value of preceding nth frame therein can be according to reality
Border situation determines.In practical applications, if the precise requirements to scene Recognition are higher, " N " can suitably be reduced
The size of value, i.e. prior frame image are the preceding nth frame image closer to current frame image, for example, before third frame image or preceding the
Two frame images.If the precise requirements to scene Recognition are higher, the size of " N " value can be properly increased, is had to reduce
Close the calculation amount of characteristic value similarity.
In judgment step S102, characteristic value, which refers to, can be used to characterize each frame image, and be specific to the frame figure
The numerical value of picture.Different according to the type of characterization, the numerical value for being specific to the frame image is also corresponding different.If current frame image select with
A certain characterization type is corresponding, and is specific to the numerical value of the current frame image to characterize current frame image, then, prior frame figure
Also corresponding selection is corresponding with this kind characterization type for picture, and is specific to the numerical value of the prior frame image to characterize prior frame figure
Picture.
For example, current frame image selects its characteristic value of the corresponding brightness value as current frame image, correspondingly, prior frame
Image is also required to select its characteristic value of the corresponding brightness value as prior frame image.It similarly, can be by calculating present frame figure
The brightness value of picture and the brightness value of prior frame image, can be obtained characteristic value difference between the two, that is, between the two
Characteristic value similarity.
Also it should be further noted that on the one hand, the size of threshold value can be according to the selected characterization type of characteristic value
It is different and different.That is, for current frame image and prior frame image, select to use corresponding brightness value as
Characteristic value, then with the brightness value as the corresponding threshold size of characteristic value, and it is opposite as characteristic value with brightness value is not had to
The threshold size answered, will be different.On the other hand, the size of threshold value, also can difference according to the actual situation, and not
Together.
Further, it is emphasized that when the first frame image of acquisition is as current frame image, since first frame image does not have
Therefore prior frame image when getting first frame image, directly will carry out scene Recognition according to the data of first frame image.
In camera preview, the disclosure provide scene recognition method can by calculate current frame image characteristic value with
Characteristic value similarity between the characteristic value of prior frame image, and this feature value similarity is compared with threshold value, according to than
Compared with result determine whether to current frame image carry out scene Recognition.If not needing to carry out scene Recognition to current frame image,
Can be directly using the scene of previous frame image as the scene of current frame image, and the acquisition parameters of previous frame image is called to carry out
Current frame image shooting, to achieve the effect that reduce camera power consumption by the number for reducing scene Recognition.
In a kind of possible embodiment, judgment step S102 includes, according to comparative feature value similarity and threshold value, judgement
Whether scene Recognition step S103 is executed.If the characteristic value similarity of current frame image and prior frame image is greater than threshold value, recognize
It is larger for feature of image difference between the two, correspondingly, scene between the two is also different, therefore, scene will be executed
Identification step S103;If the characteristic value similarity of current frame image and prior frame image is less than or equal to threshold value, then it is assumed that the two
Between feature of image difference it is smaller, correspondingly, scene between the two is also approximately uniform, therefore, scene Recognition will not be executed
Step S103.
In a kind of possible embodiment, in order to improve the accuracy of scene Recognition, prior frame image can be present frame
The preceding first frame image of image, i.e. the previous frame image of current frame image.That is, by comparing current frame image and currently
Characteristic value similarity between the previous frame image of frame image, the size relation with threshold value, to determine whether executing scene Recognition
Step S103.
In order to improve the efficiency and accuracy of scene Recognition, in practical applications, can comparative feature value similarity with
On the basis of threshold value, further relatively current frame image acquisition the moment and it is previous execute scene Recognition step the execution moment it
Between time interval, the relationship between interval threshold.And it is based on this, judge whether to execute scene Recognition step S103.
Therefore, in a kind of possible embodiment, wherein obtaining step S101 further includes obtaining adopting for current frame image
Collect the moment, and according to the acquisition moment of current frame image and the previous execution moment for executing scene Recognition step, obtains the two
Between time interval.Judgment step S102 includes, according to comparative feature value similarity and threshold value and compared with time interval and
Every threshold value, judge whether to execute scene Recognition step S103.
Being described in detail below existing for above scheme may embodiment.
According to the judging result of the characteristic value similarity of current frame image and prior frame image, if the spy of current frame image
The similarity of value indicative and the characteristic value of prior frame image is smaller, then the scene to current frame image will be needed to be known again
Not.But if current frame image the acquisition moment not long ago, this time interval can be understood as between the very short time
Every, for example be the step of scene recognition method 100 has just executed scene Recognition within 400ms.So, in practical application
In, it could be theoretically argued that, although the scene of current frame image and prior frame image is different, it is held apart from upper primary
In the very short time interval of row scene Recognition step, no longer current frame image is re-recognized and more new scene, actually not
The experience of user can be had an impact.Therefore, in this case, in practical applications, can not have to current frame image weight
New identification and more new scene, in this way, the number that will reduce scene Recognition, thereby reduces the power consumption of device.
For above situation, the disclosure will also provide corresponding embodiment and be illustrated.
In a kind of possible embodiment, time interval threshold value includes first time interval threshold value;Scene recognition method is removed
Except including above-mentioned obtaining step S101, judgment step S102 and scene Recognition step S103, which is not described herein again, wherein
Obtaining step S101 and judgment step S102 also there are following features.
Obtaining step S101 further includes the acquisition moment for obtaining current frame image, that is, obtains and record and adopt each time
At the acquisition moment of the current frame image collected, T1 can be denoted as.
Judgment step S102 includes, and according to comparative feature value similarity and threshold value, judges whether to execute scene Recognition step
S103.If characteristic value similarity is greater than threshold value, i.e., it could be theoretically argued that the scene of current frame image and the scene of prior frame image have
Institute is different, then continues between the acquisition moment for judging current frame image and the previous execution moment for executing scene Recognition step S103
Time interval.
This execution moment for executing scene Recognition step S103 refers to that the last time nearest apart from current frame image executes field
At the time of scape identification step S103, T2 can be denoted as.Correspondingly, the acquisition moment T1 of current frame image and previous execution scene are known
Time interval between the execution moment T2 of other step S103 is T1-T2.
If time interval T1-T2 is greater than or equal to the first interval threshold, since current frame image and prior frame image pass through
Characteristic value similarity judges that scene between the two is different, and in addition, the acquisition moment T1 distance of current frame image is last
Execution scene Recognition step execute moment T2 between time interval T1-T2 it is not short enough, then, then will be directed to
First frame image executes scene Recognition step S103, that is, carries out scene Recognition to prior frame image and update;If time interval
T1-T2 does not execute scene Recognition step S103 then less than the first interval threshold.
It, can according to the actual situation it should be noted that first time interval threshold value, is a kind of very short time interval
Size that is different and adjusting the numerical value, first time interval threshold can be the numerical value within 400ms.
In the scene recognition method 100 that the disclosure provides, if the characteristic value of current frame image and prior frame image
Similarity between characteristic value is greater than threshold value, it is possible to think, the feature of image between prior frame image and current frame image
Difference is larger, correspondingly, the scene of the two is also different, that is to say, that the scene of prior frame image needs to re-start field
Scape identification.But in practical applications, by further judging the acquisition moment T1 of frame image in the ban and apart from current frame image
The recent difference T1-T2 executed between moment T2 for executing scene Recognition step S103, if T1-T2 is less than first
Between threshold value, then it is assumed that although the scene of current frame image and prior frame image is different, the short time, more new scene was not yet
It will affect the experience of user, and then reduce the number of scene Recognition, reduce the power consumption of device.
Similarly, if judged by characteristic value similarity, the scene of current frame image and the scene of prior frame image are obtained
Have essentially identical, is not need to re-recognize the scene of current frame image theoretically.But if in present frame figure
A period of time before the acquisition moment of picture, this time interval may be considered the time interval other than 2s, just executed
The step of scene Recognition of last time.So, in practical applications, although according to judging result, it is believed that current frame image with
First frame image has similar scene, still, due to not carrying out scene Recognition for a long time and updating, in order to guarantee reality
Effect, it is desired nonetheless to current frame image be carried out forcing scene Recognition and be updated, to further increase the accuracy of actual result.
Correspondingly, being directed to above situation, the disclosure will provide corresponding embodiment and be illustrated.
In a kind of possible embodiment, time interval threshold value includes the second time interval threshold value;Scene recognition method is removed
Except including above-mentioned obtaining step S101, judgment step S102 and scene Recognition step S103, which is not described herein again, wherein
Obtaining step S101 and judgment step S102 also there are following features.
Obtaining step S101 further includes the acquisition moment for obtaining current frame image, that is, obtains and record and adopt each time
At the acquisition moment of the current frame image collected, T1 can be denoted as.
Judgment step S102 includes, and according to comparative feature value similarity and threshold value, judges whether to execute scene Recognition step
S103.If characteristic value similarity is less than or equal to threshold value, i.e., it could be theoretically argued that the scene of current frame image and prior frame image
Scene is similar, then continue to judge current frame image the acquisition moment and the previous execution moment for executing scene Recognition step S103 it
Between time interval.
This execution moment for executing scene Recognition step S103 refers to that the last time nearest apart from current frame image executes field
At the time of scape identification step S103, T2 can be denoted as.Correspondingly, the acquisition moment T1 of current frame image and previous execution scene are known
Time interval between the execution moment T2 of other step S103 is T1-T2.
If time interval T1-T2 is greater than the second interval threshold, scene Recognition step S103 can be executed, that is, to first
Frame image carries out scene Recognition and updates;If time interval T1-T2 is less than or equal to the second interval threshold, due to current frame image
With prior frame image by characteristic value similarity, judge that scene between the two is similar, in addition, the acquisition of current frame image
Time interval T1-T2 between the execution moment T2 of last the executions scene Recognition step of moment T1 distance not long enough,
So, for current frame image, then scene Recognition step S103 is no longer executed.
It should be noted that the second time interval threshold value, is a kind of relatively long time interval, it can be according to practical feelings
The difference of condition and the size for adjusting the numerical value, the second time interval threshold can be the numerical value other than 2s.
In the scene recognition method 100 that the disclosure provides, if the characteristic value of current frame image and prior frame image
Similarity between characteristic value is less than or equal to threshold value, it is possible to think, the figure between prior frame image and current frame image
As feature difference it is little, correspondingly, the scene of the two it is also assumed that be it is essentially identical, theoretically, the scene of prior frame image
It does not need to re-start scene Recognition.But in practical applications, by further judging the acquisition moment T1 of frame image in the ban
Difference T1-T2 between the execution moment T2 of the last execution scene Recognition step S103 nearest apart from current frame image,
If T1-T2 is greater than second time threshold, then it is assumed that, although it could be theoretically argued that the scene base of current frame image and prior frame image
This is identical, still, due to holding for the nearest last execution scene Recognition step S103 of the acquisition moment distance of current frame image
The time interval at row moment is too long, in order to further ensure actual effect, then can force to carry out scene Recognition to current frame image
And update, to further increase the accuracy of actual result.
In practical applications, if the pixel of the image of frame to be identified is more, to each of picture pixel
Characteristic value is compared one by one, will increase corresponding operand, therefore, can choose several pixels in each frame image
Characteristic value is compared, and to reduce operand, further decreases corresponding power consumption.
In a kind of possible embodiment, obtaining step S101 further includes, by way of sampling to current frame image and
Prior frame image carries out characteristics extraction, and spy of the corresponding characteristic value that will acquire as current frame image and prior frame image
Value indicative.In this manner, to reduce the operand in characteristic value deterministic process, corresponding power consumption is reduced.
For example, a certain frame image to be identified can be carried out N equal part, the central pixel point in every equal part then can be chosen
Characteristic value, the characteristic value as the equal part.Correspondingly, this frame image to be identified also just obtains N number of characteristic value.Actually answering
In, can by the average value of the characteristic value similarity of N number of characteristic value of obtained current frame image and prior frame image, as
The characteristic value similarity of current frame image and prior frame image.That is, if characteristic value similarity is enabled to be characterized value difference value, that
, can be by the average value of the characteristic value difference of N number of characteristic value of obtained current frame image and prior frame image, as current
The characteristic value difference of frame image and prior frame image.
In a kind of possible embodiment, characteristic value includes the brightness value or frame image to be identified of frame image to be identified
Rgb value.Wherein, rgb value can be the weighted average of the pixel value in the channel R, can be the weighted average of the pixel value in the channel G
Value, can be the weighted average of the pixel value of channel B, can also be the weighted average of the pixel value in tri- channels R, G, B.
As shown in Fig. 2, before obtaining step S101, scene recognition method 100 is also wrapped in a kind of possible embodiment
It includes and obtains image step S104.
In obtaining image step S104, frame image to be identified can be obtained by image collecting device in real time;It is obtaining
In step, it can be based on frame image to be identified, obtain the characteristic value of frame image to be identified.
In practical applications, image collecting device can be independently of corresponding with the scene recognition method that the disclosure provides
Identification device except a kind of front-end acquisition device, known by the way that acquired image to be passed to the scene provided with the disclosure
The corresponding identification device of other method;It is also possible to be present in identification dress corresponding with the scene recognition method that the disclosure provides
The inside set constitutes an image capture module of the identification device.
As shown in figure 3, scene recognition method 100 further includes generating parameter step in a kind of possible embodiment
S105。
It, can be based on the scene Recognition obtained in scene Recognition step S103 as a result, raw in generating parameter step S105
At the acquisition parameters for carrying out image taking for image collecting device.For example, if scene Recognition the result is that the scene be light
Darker scene, then the characteristics of disclosure can be according to the scene of dark, generates the image taking parameter of large aperture, and
The acquisition parameters are passed into image collecting device.
Based on identical inventive concept, the embodiment of the present disclosure also provides a kind of scene Recognition device 200.Referring to fig. 4, scene
Identification device 200 includes obtaining module 201, and for obtaining the characteristic value of frame image to be identified, frame image to be identified includes current
Frame image and prior frame image, prior frame image are the preceding nth frame image of current frame image;Judgment module 202, it is special for comparing
Value indicative similarity and threshold value, wherein characteristic value similarity is between the characteristic value of current frame image and the characteristic value of prior frame image
Similarity;Scene Recognition module 203: for the comparison result according to characteristic value similarity and threshold value, it is based on current frame image
Data, carry out scene Recognition.
In one example, judgment module 202 is used for, and according to comparative feature value similarity and threshold value, judges whether to execute scene
Identification step.If characteristic value similarity is greater than threshold value, scene Recognition step is executed;If characteristic value similarity is less than or equal to threshold
Value, then do not execute scene Recognition step.
In one example, prior frame image can also be the preceding first frame image of current frame image, that is, current frame image
Previous frame image.
In one example, it obtains module 201 to be also used to, obtains the acquisition moment of current frame image, and according to current frame image
Acquisition the moment and it is previous execute scene Recognition step the execution moment, obtain current frame image the acquisition moment and previous execution
The time interval of scene Recognition step executed between the moment;Judgment module 202 is used for, according to comparative feature value similarity and threshold
Value and compare time interval and interval threshold, judges whether to execute the scene Recognition step;Scene Recognition module 203 is also
For obtaining and saving the execution moment for executing scene Recognition step.
In one example, time interval threshold value includes the first interval threshold;It obtains module 201 to be also used to, obtains present frame figure
The acquisition moment of picture;Judgment module 202 is used for, and according to comparative feature value similarity and threshold value, judges whether to execute scene Recognition
Step.If characteristic value similarity be greater than threshold value, judge current frame image the acquisition moment and previous execution scene Recognition step
Execute the moment between time interval, if time interval be greater than or equal to the first interval threshold, execute scene Recognition step;
If time interval less than the first interval threshold, does not execute scene Recognition step;Scene Recognition module 203 is also used to, and is obtained simultaneously
Save the execution moment for executing scene Recognition step.
In one example, time interval threshold value includes the second interval threshold;It obtains module 201 to be also used to, obtains present frame figure
The acquisition moment of picture;Judgment module 202 is used for, and according to comparative feature value similarity and threshold value, judges whether to execute scene Recognition
Step.If characteristic value similarity is less than or equal to threshold value, judge that the acquisition moment of current frame image and previous execution scene know
The time interval of other step executed between the moment executes scene Recognition step if time interval is greater than the second interval threshold;
If time interval is less than or equal to the second interval threshold, scene Recognition step is not executed;Scene Recognition module 203 is also used to,
It obtains and saves the execution moment for executing scene Recognition step.
In one example, it obtains module 201 to be also used to, current frame image and prior frame image be carried out by sample mode special
Value indicative is extracted, characteristic value of the characteristic value that will acquire as current frame image and prior frame image.
In one example, characteristic value includes the brightness value of frame image to be identified or the rgb value of frame image to be identified.
In one example, before obtaining module 201, capturing and recognition device 200 further includes obtaining image module 204, being used for
By image collecting device, frame image to be identified is obtained in real time;Obtain module 201 be used for, be based on frame image to be identified, obtain to
Identify the characteristic value of frame image.
In one example, capturing and recognition device 200 further includes generating parameter module 205, for based in scene Recognition step
The scene Recognition of acquisition is as a result, generate the acquisition parameters for carrying out image taking for image collecting device.
Fig. 5 shows a kind of electronic equipment 30 that an embodiment of the disclosure provides.As shown in figure 5, the disclosure
The a kind of electronic equipment 30 that one embodiment provides, wherein the electronic equipment 30 includes memory 310, processor 320, defeated
Enter/export (Input/Output, I/O) interface 330.Wherein, memory 310, for storing instruction.Processor 320, for adjusting
The instruction execution disclosure scene recognition method stored with memory 310.Wherein, processor 320 respectively with memory 310, I/O
Interface 330 connects, such as can be attached by bindiny mechanism's (not shown) of bus system and/or other forms.Memory
310 can be used for storing program and data, and the program including scene Recognition involved in the embodiment of the present disclosure, processor 320 passes through
Operation is stored in the program of memory 310 thereby executing the various function application and data processing of electronic equipment 30.
Processor 320 can use digital signal processor (Digital Signal in the embodiment of the present disclosure
Processing, DSP), field programmable gate array (Field-Programmable Gate Array, FPGA), programmable patrol
At least one of volume array (Programmable Logic Array, PLA) example, in hardware realizes, the processor 320
It can be central processing unit (Central Processing Unit, CPU) or there is data-handling capacity and/or instruction
The combination of one or more of the processing unit of other forms of executive capability.
Memory 310 in the embodiment of the present disclosure may include one or more computer program products, the computer
Program product may include various forms of computer readable storage mediums, such as volatile memory and/or non-volatile deposit
Reservoir.The volatile memory for example may include random access memory (Random Access Memory, RAM) and/
Or cache memory (cache) etc..The nonvolatile memory for example may include read-only memory (Read-Only
Memory, ROM), flash memory (Flash Memory), hard disk (Hard Disk Drive, HDD) or solid state hard disk
(Solid-State Drive, SSD) etc..
In the embodiment of the present disclosure, I/O interface 330 can be used for receiving input instruction (such as number or character information, and
Generate key signals input related with the user setting of electronic equipment 30 and function control etc.), it can also be output to the outside various
Information (for example, image or sound etc.).In the embodiment of the present disclosure I/O interface 330 may include physical keyboard, function button (such as
Volume control button, switch key etc.), mouse, operating stick, trace ball, microphone, one in loudspeaker and touch panel etc.
It is a or multiple.
In some embodiments, present disclose provides a kind of computer readable storage medium, the computer-readable storages
Media storage has computer executable instructions, and computer executable instructions when executed by the processor, execute described above appoint
Where method.
Although description operation in a particular order in the accompanying drawings should not be construed as requiring specific shown in
Sequence or serial order operate to execute these operations, or shown in requirement execution whole to obtain desired result.?
In specific environment, multitask and parallel processing be may be advantageous.
Disclosed method and device can be completed using standard programming technology, using rule-based logic or its
His logic realizes various method and steps.It should also be noted that herein and the terms used in the claims " device "
" module " is intended to include using the realization of a line or multirow software code and/or hardware realization and/or for receiving input
Equipment.
One or more combined individually or with other equipment can be used in any step, operation or program described herein
A hardware or software module are executed or are realized.In one embodiment, software module use includes comprising computer program
The computer program product of the computer-readable medium of code is realized, can be executed by computer processor any for executing
Or whole described step, operation or programs.
For the purpose of example and description, the preceding description of disclosure implementation is had been presented for.Preceding description is not poor
The disclosure is restricted to exact form disclosed by also not the really wanting of act property, according to the above instruction there is likely to be various modifications and
Modification, or various changes and modifications may be obtained from the practice of the disclosure.Select and describe these embodiments and be in order to
Illustrate the principle and its practical application of the disclosure, so that those skilled in the art can be to be suitable for the special-purpose conceived
Come in a variety of embodiments with various modifications and using the disclosure.
Claims (12)
1. a kind of scene recognition method, wherein the described method includes:
Obtaining step obtains the characteristic value of frame image to be identified, and the frame image to be identified includes current frame image and prior frame
Image, the prior frame image are the preceding nth frame image of the current frame image;
Judgment step judges whether to execute scene Recognition step, wherein the feature according to comparative feature value similarity and threshold value
The similarity being worth between the characteristic value that similarity is the current frame image and the characteristic value of the prior frame image;
The scene Recognition step includes being based on the current frame image, carries out scene Recognition.
2. according to the method described in claim 1, wherein, the judgment step includes:
It is executed according to the characteristic value similarity and the threshold value if the characteristic value similarity is greater than the threshold value
Scene Recognition step;If the characteristic value similarity is less than or equal to the threshold value, the scene Recognition step is not executed.
3. according to the method described in claim 1, wherein,
The prior frame image is the preceding first frame image of the current frame image.
4. according to the method described in claim 1, wherein,
The obtaining step further include: obtain the acquisition moment of the current frame image, and adopting according to the current frame image
Collect moment and the previous execution moment for executing scene Recognition step, obtain acquisition moment of the current frame image with it is described previous
Execute the time interval of scene Recognition step executed between the moment;
The judgment step include: according to the characteristic value similarity with the threshold value and compared with the time interval
With interval threshold, judge whether to execute the scene Recognition step;
The scene Recognition step further include: obtain and save the execution moment for executing the scene Recognition step.
5. according to the method described in claim 4, wherein,
The interval threshold includes the first interval threshold;
The judgment step includes: according to the characteristic value similarity and the threshold value, if the characteristic value similarity is big
In the threshold value, then the acquisition moment and the previous execution moment for executing scene Recognition step of the current frame image are judged
Between time interval,
If the time interval is greater than or equal to the first interval threshold, scene Recognition step is executed;
If the time interval is less than first interval threshold, scene Recognition step is not executed.
6. method according to claim 4 or 5, wherein
The interval threshold includes the second interval threshold;
The judgment step includes: according to the characteristic value similarity and the threshold value, if the characteristic value similarity is small
In or equal to the threshold value, then the acquisition moment of the current frame image and holding for the previous execution scene Recognition step are judged
Time interval between the row moment,
If the time interval is greater than the second interval threshold, scene Recognition step is executed;
If the time interval is less than or equal to second interval threshold, scene Recognition step is not executed.
7. according to the method described in claim 1, wherein, the obtaining step further include:
Characteristics extraction, the spy that will acquire are carried out to the current frame image and the prior frame image by sample mode
Characteristic value of the value indicative as the current frame image and the prior frame image.
8. according to the method described in claim 1, wherein,
The characteristic value includes the brightness value of the frame image to be identified or the rgb value of the frame image to be identified.
9. according to the method described in claim 1, wherein, the method also includes:
Parameter step is generated, image collector is used for as a result, generating based on the scene Recognition obtained in the scene Recognition step
Set the acquisition parameters for carrying out image taking.
10. a kind of scene Recognition device, wherein described device includes:
Obtain module, for obtaining the characteristic value of frame image to be identified, the frame image to be identified include current frame image and
First frame image, the prior frame image are the preceding nth frame image of the current frame image;
Judgment module is used for comparative feature value similarity and threshold value, wherein the characteristic value similarity is the current frame image
Characteristic value and the prior frame image characteristic value between similarity;
Scene Recognition module: for the comparison result according to the characteristic value similarity and the threshold value, it is based on the present frame
The data of image carry out scene Recognition.
11. a kind of electronic equipment, wherein the electronic equipment includes:
Memory, for storing instruction;And
Processor, scene Recognition side described in any one of instruction execution claim 1-9 for calling memory storage
Method.
12. a kind of computer readable storage medium, wherein
The computer-readable recording medium storage has computer executable instructions, and the computer executable instructions are by handling
When device executes, perform claim requires scene recognition method described in any one of 1-9.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910596008.1A CN110336943B (en) | 2019-07-03 | 2019-07-03 | Scene recognition method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910596008.1A CN110336943B (en) | 2019-07-03 | 2019-07-03 | Scene recognition method and device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110336943A true CN110336943A (en) | 2019-10-15 |
CN110336943B CN110336943B (en) | 2021-05-25 |
Family
ID=68144783
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910596008.1A Active CN110336943B (en) | 2019-07-03 | 2019-07-03 | Scene recognition method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110336943B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114038197A (en) * | 2021-11-24 | 2022-02-11 | 浙江大华技术股份有限公司 | Scene state determination method and device, storage medium and electronic device |
Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1787635A (en) * | 2004-12-08 | 2006-06-14 | Lg电子株式会社 | Apparatus and method for video searching in a mobile communications terminal |
US20130021434A1 (en) * | 2003-05-02 | 2013-01-24 | Grandeye Ltd. | Method and System of Simultaneously Displaying Multiple Views for Video Surveillance |
CN103763458A (en) * | 2014-01-20 | 2014-04-30 | 华为技术有限公司 | Scene change detection method and device |
CN105262954A (en) * | 2015-11-17 | 2016-01-20 | 腾讯科技(深圳)有限公司 | Method and device for triggering camera to perform automatic focusing |
CN105812647A (en) * | 2014-12-30 | 2016-07-27 | 联想(北京)有限公司 | Information processing method and electronic equipment |
US9536161B1 (en) * | 2014-06-17 | 2017-01-03 | Amazon Technologies, Inc. | Visual and audio recognition for scene change events |
CN107592468A (en) * | 2017-10-23 | 2018-01-16 | 维沃移动通信有限公司 | A kind of shooting parameter adjustment method and mobile terminal |
CN107635101A (en) * | 2017-10-27 | 2018-01-26 | 广东欧珀移动通信有限公司 | Image pickup method, device, storage medium and electronic equipment |
CN108804980A (en) * | 2017-04-28 | 2018-11-13 | 合信息技术(北京)有限公司 | Switching detection method of video scene and device |
CN109063694A (en) * | 2018-09-12 | 2018-12-21 | 北京科技大学 | A kind of video object detection recognition method |
CN109327655A (en) * | 2017-07-31 | 2019-02-12 | 展讯通信(上海)有限公司 | Continuous Atomatic focusing method and system |
CN109426832A (en) * | 2017-08-30 | 2019-03-05 | 湖南拓视觉信息技术有限公司 | Closed loop detection method, storage medium and electronic equipment in scene three-dimensional modeling |
-
2019
- 2019-07-03 CN CN201910596008.1A patent/CN110336943B/en active Active
Patent Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130021434A1 (en) * | 2003-05-02 | 2013-01-24 | Grandeye Ltd. | Method and System of Simultaneously Displaying Multiple Views for Video Surveillance |
CN1787635A (en) * | 2004-12-08 | 2006-06-14 | Lg电子株式会社 | Apparatus and method for video searching in a mobile communications terminal |
CN103763458A (en) * | 2014-01-20 | 2014-04-30 | 华为技术有限公司 | Scene change detection method and device |
US9536161B1 (en) * | 2014-06-17 | 2017-01-03 | Amazon Technologies, Inc. | Visual and audio recognition for scene change events |
CN105812647A (en) * | 2014-12-30 | 2016-07-27 | 联想(北京)有限公司 | Information processing method and electronic equipment |
CN105262954A (en) * | 2015-11-17 | 2016-01-20 | 腾讯科技(深圳)有限公司 | Method and device for triggering camera to perform automatic focusing |
CN108804980A (en) * | 2017-04-28 | 2018-11-13 | 合信息技术(北京)有限公司 | Switching detection method of video scene and device |
CN109327655A (en) * | 2017-07-31 | 2019-02-12 | 展讯通信(上海)有限公司 | Continuous Atomatic focusing method and system |
CN109426832A (en) * | 2017-08-30 | 2019-03-05 | 湖南拓视觉信息技术有限公司 | Closed loop detection method, storage medium and electronic equipment in scene three-dimensional modeling |
CN107592468A (en) * | 2017-10-23 | 2018-01-16 | 维沃移动通信有限公司 | A kind of shooting parameter adjustment method and mobile terminal |
CN107635101A (en) * | 2017-10-27 | 2018-01-26 | 广东欧珀移动通信有限公司 | Image pickup method, device, storage medium and electronic equipment |
CN109063694A (en) * | 2018-09-12 | 2018-12-21 | 北京科技大学 | A kind of video object detection recognition method |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114038197A (en) * | 2021-11-24 | 2022-02-11 | 浙江大华技术股份有限公司 | Scene state determination method and device, storage medium and electronic device |
Also Published As
Publication number | Publication date |
---|---|
CN110336943B (en) | 2021-05-25 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102096805B (en) | Apparatus and method for registering plurality of facial images for face recognition | |
CN108924420B (en) | Image shooting method, image shooting device, image shooting medium, electronic equipment and model training method | |
CN109951595A (en) | Intelligence adjusts method, apparatus, storage medium and the mobile terminal of screen intensity | |
US10419661B2 (en) | Shooting method and shooting device | |
CN110248098A (en) | Image processing method, device, storage medium and electronic equipment | |
CN109598250B (en) | Feature extraction method, device, electronic equipment and computer readable medium | |
CN108650457A (en) | Automatic photographing method, device, storage medium and mobile terminal | |
KR20210048272A (en) | Apparatus and method for automatically focusing the audio and the video | |
Yang et al. | Personalized exposure control using adaptive metering and reinforcement learning | |
CN110213479A (en) | A kind of video capture anti-fluttering method and device | |
CN106101540A (en) | Focusing determines method and device | |
CN109819176A (en) | A kind of image pickup method, system, device, electronic equipment and storage medium | |
CN104662889B (en) | Method and apparatus for being shot in portable terminal | |
US10769416B2 (en) | Image processing method, electronic device and storage medium | |
CN110035237B (en) | Image processing method, image processing device, storage medium and electronic equipment | |
CN107147851A (en) | Photo processing method, device, computer-readable recording medium and electronic equipment | |
CN110336943A (en) | A kind of scene recognition method and device | |
JP5862071B2 (en) | Image processing apparatus, imaging apparatus, and program | |
WO2024051697A1 (en) | Image fusion method and apparatus, electronic device, and storage medium | |
CN106210541A (en) | A kind of video generation method, device and mobile terminal | |
CN115862081A (en) | Image acquisition method, device, equipment and computer readable storage medium | |
CN104486553B (en) | A kind of distant view photograph image pickup method and terminal | |
WO2023001110A1 (en) | Neural network training method and apparatus, and electronic device | |
CN112019737A (en) | Method of operating an electronic device comprising an image sensor | |
CN108495038B (en) | Image processing method, image processing device, storage medium and electronic equipment |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |