CN109766815A - A kind of pair of object event carries out pre-warning system and method - Google Patents
A kind of pair of object event carries out pre-warning system and method Download PDFInfo
- Publication number
- CN109766815A CN109766815A CN201910004746.2A CN201910004746A CN109766815A CN 109766815 A CN109766815 A CN 109766815A CN 201910004746 A CN201910004746 A CN 201910004746A CN 109766815 A CN109766815 A CN 109766815A
- Authority
- CN
- China
- Prior art keywords
- resolution image
- image
- target area
- pixel
- resolution
- 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.)
- Pending
Links
Abstract
Present applicant proposes the method and systems that a kind of pair of object event carries out early warning, wherein this method comprises: obtaining first resolution image, the first resolution image includes target area, and the first resolution image is shot by the first equipment;Extract the corresponding feature vector of each pixel in the first resolution image;After the corresponding feature vector input Early-warning Model of each pixel, the output of the Early-warning Model is obtained as a result, the output result includes whether the target area occurs object event;And the output result of the Early-warning Model be target area occur object event when, issue warning information.
Description
Technical field
The present invention relates to satellite remote sensing field more particularly to a kind of object event early warning is carried out using remote sensing satellite image
System and method.
Background technique
With the rapid development of space technology, more and more artificial earth satellites are launched into space, are taken according to it
The equipment of load undertakes various tasks.Remote sensing satellite can cover its flight track by its photographing device carried
The ground areas of lid is taken pictures.This photographing device is other than it can receive visible light, moreover it is possible to receive such as infrared
The electromagnetic wave of wave band, far infrared band, so as to more clearly restore surface state.With the development of miniaturization satellite, more
It can be launched lift-off come more remote sensing satellites, operate in different tracks, numerous satellites after networking are covered
More and more ground regions.
In recent years, remote sensing satellite can also be used to be observed the object events such as the natural calamity of earth's surface generation.
For example, can be determined to the range that forest fire occurs from remote sensing images.But these applications are only to send out in disaster
The coverage of disaster is determined after life, is unable to respond when disaster occurs.
It would therefore be desirable to a kind of new carries out pre-warning system and method to object event, with being capable of basis in real time
The object events such as remote sensing images discovery natural calamity achieve the purpose that prevent and reduce natural disasters to make early warning in time.
Summary of the invention
The application be designed to provide it is a kind of new carry out pre-warning system and method using remote sensing satellite, to solve to divide
The problem of lower satellite of resolution can not identify such as fire, the object events such as vegetation pest and disease damage.
The first aspect of the application proposes a kind of pair of object event and carries out pre-warning system, may include: that image obtains
Unit, for obtaining first resolution image, the first resolution image includes target area, the first resolution image
It is shot by the first equipment;Feature extraction unit, for extracting the corresponding feature of each pixel in the first resolution image
Vector;Model arithmetic unit, for will be after the corresponding feature vector input Early-warning Model of each pixel, acquisition be described pre-
The output of alert model is as a result, the output result includes whether the target area occurs object event;And prewarning unit, it uses
When the output result of the Early-warning Model is that object event occurs for target area, warning information is issued.
In some embodiments, the system may further include: image call unit, for calling the second equipment to obtain
Second resolution image is obtained, the second resolution image includes the target area;And confirmation unit, for according to
Second resolution image confirms that object event occurs for the target area, wherein the unit pixel of the first resolution image
The corresponding geographic range of point is greater than the corresponding geographic range of unit pixel point of the second resolution image.
In some embodiments, the system may further include scheduling unit, and the second equipment of the calling obtains the
Two image in different resolution may include: that the determination unit determines that the output result includes that target may occur for the target area
Event;The scheduling unit calls stand-by equipment to shoot the second resolution image.
Two aspects of the application propose the method that a kind of pair of object event carries out early warning, may include: to obtain first to differentiate
Rate image, the first resolution image includes target area, and the first resolution image is shot by the first equipment;Extract institute
State the corresponding feature vector of each pixel in first resolution image;By the corresponding feature vector input of each pixel
After Early-warning Model, the output of the Early-warning Model is obtained as a result, the output result includes whether the target area occurs mesh
Mark event;And the output result of the Early-warning Model be target area occur object event when, issue warning information.
In some embodiments, the method may further include: call the second equipment to obtain second resolution image,
The second resolution image includes the target area;And the target area is confirmed according to the second resolution image
Object event occurs, wherein the corresponding geographic range of unit pixel point of the first resolution image is greater than described second minute
The corresponding geographic range of unit pixel point of resolution image.
In some embodiments, it may include: determining described defeated that the second equipment of the calling, which obtains second resolution image,
Result includes that object event may occur for the target area out;Stand-by equipment is called to shoot the second resolution image.
In some embodiments, the first resolution image can be satellite remote sensing images;The second resolution figure
As that can be the satellite remote sensing images for target area;The stand-by equipment can be in-orbit spacecraft.
In some embodiments, the corresponding feature vector of each pixel may include at least one in following characteristics
A: the parameter of at least one wave band, the wave band include the optional frequency section in spectrum;Weather parameters, reflection described first
The weather condition of image in different resolution target area when shooting;Terrain parameter reflects the topographic features of the target area;When
Field parameter reflects that at least one feature of the pixel changes with time trend;And spatial parameter, reflect pixel week
Enclose the feature situation of at least one pixel.
In some embodiments, the feature may further include vegetation index, and the vegetation index includes the pixel
The near-infrared parameter of point and the ratio of IR parameters.
In some embodiments, the Early-warning Model can be obtained by default training process, the default training process
It may include: to obtain multiple sample images, the resolution ratio of the multiple sample image is not less than second resolution, the multiple sample
The corresponding incident region of each image in this image, the object event occurs for each incident region;For described more
Each of a sample image image: corresponding institute when determining that at least one described object event is occurring according to the image
State the history image not less than first resolution in incident region;And using the multiple history image as described in sample training
Early-warning Model.
In some embodiments, at least one history image described in the determination may include: the thing identified in the image
Hair region and the shooting time for obtaining the image;According to the incident region and the shooting time, candidate time model is determined
It include the candidate history first resolution image in the incident region in enclosing, when the candidate time range includes the shooting
It carves;And at least one described history image is determined from the candidate history first resolution image.
In some embodiments, described to determine at least one described history from the candidate history first resolution image
Image includes: to be changed with time trend according at least one characteristic parameter in the candidate history first resolution image, really
Determine incident time range, the candidate time range includes the incident time range;Determine that the incident time range is corresponding
Candidate history first resolution image be at least one described history image.
In some embodiments, described using the multiple history image as Early-warning Model described in sample training, it can wrap
It includes: for each history image, extracting the training feature vector of each of which pixel, the feature class in the training feature vector
Type is identical as the characteristic type in described eigenvector;And using the training feature vector of each pixel as sample
Collection, is input in the Early-warning Model, is trained to the Early-warning Model.
The third aspect of the application proposes a kind of electronic equipment, may include memory, processor and is stored in described
On memory and the computer program that can run on the processor, which is characterized in that the processor executes the calculating
The step of method of early warning is carried out to object event described in the application first aspect is realized when machine program.
The geographic range that object event occurs first is determined by high-resolution satellite image, and then finds out corresponding geographical model
The image for enclosing low resolution satellite when lower object event is occurring, then using the satellite image of low resolution as training set
It goes to train Early-warning Model, this enables the Early-warning Model after training only can judge target area by the image of low resolution
Whether object event is occurred.Since the satellite image of low resolution opposite may be easier to obtain, while also can more timely obtain
It takes, early warning system or method that the application is proposed are had more preferable applicability and better instantaneity by this.
Other feature will be set forth in part in the description in the application.By the elaboration, make the following drawings and
The content of embodiment narration becomes apparent for those of ordinary skills.Inventive point in the application can pass through
Practice is sufficiently illustrated using method described in detailed example discussed below, means and combinations thereof.
Detailed description of the invention
Exemplary embodiment disclosed in this application is described in detail in the following drawings.Wherein identical appended drawing reference is in attached drawing
Several views in indicate similar structure.Those of ordinary skill in the art will be understood that these embodiments be non-limiting,
Exemplary embodiment, the purpose that attached drawing is merely to illustrate and describes, it is no intended to it limits the scope of the present disclosure, other modes
Embodiment may also similarly complete the intention of the invention in the application.It should be appreciated that the drawings are not drawn to scale.Wherein:
Fig. 1 shows the usage scenario figure according to shown in some embodiments of the present application.
Fig. 2 is the exemplary process diagram of one embodiment that early warning is carried out using remote sensing satellite in the application.
Fig. 3 is the exemplary process diagram of one embodiment of the training Early-warning Model in the application.
Fig. 4 is the exemplary process diagram of one embodiment of at least one history image described in determination in the application.
Fig. 5 is to determine at least one described history image from the candidate history first resolution image in the application
One embodiment exemplary process diagram.
Fig. 6 is the schematic diagram of situation of change of certain parameter value of certain pixel in the application within the scope of candidate time.
Fig. 7 is the schematic diagram of one embodiment of the object event early warning system in the application.
Specific embodiment
Following description provides the specific application scene of the application and requirements, it is therefore an objective to those skilled in the art be enable to make
It makes and using the content in the application.To those skilled in the art, to the various partial modifications of the disclosed embodiments
Be it will be apparent that and without departing from the spirit and scope of the disclosure, the General Principle that will can be defined here
Applied to other embodiments and application.Therefore, the embodiment the present disclosure is not limited to shown in, but it is consistent most wide with claim
Range.
Term used herein is only used for the purpose of description specific example embodiments, rather than restrictive.For example, unless
Context is expressly stated otherwise, used herein above, singular " one ", and " one " and " being somebody's turn to do " also may include plural form.
When used in this manual, term " including ", " including " and/or " containing " is meant that associated integer, step, behaviour
Make, element and/or component exist, but be not excluded for other one or more features, integer, step, operation, element, component and/or
Group presence or can be added in the system/method other features, integer, step, operation, element, component and/or.
In view of being described below, the operation of the related elements of these features of the disclosure and other features and structure and
The economy of combination and the manufacture of function and component may be significantly raising.With reference to attached drawing, all these formation disclosure
A part.It is to be expressly understood, however, that the purpose that attached drawing is merely to illustrate and describes, it is no intended to limit the disclosure
Range.
Fig. 1 shows the usage scenario figure according to shown in some embodiments of the present application.In Fig. 1, the first satellite 1 (under
Text is referred to as the first equipment) and the second satellite 2 (being hereafter referred to as the second equipment) be the remote sensing satellite, respective
Running track on around earth rotation.In Fig. 1 by taking two satellites as an example, two in the satellite network of multi-satellite composition are indicated
, any satellite in the satellite network can apply method for early warning described herein and system.Some
In embodiment, first satellite 1 can be low resolution satellite, and second satellite 2 can be high-resolution satellite, described
Low resolution and high-resolution only indicate the corresponding geographical model of unit pixel point in remote sensing images that first satellite 1 takes
It encloses and is greater than the corresponding geographic range of unit pixel point in the remote sensing images that second satellite 2 takes, without to the exhausted of satellite
Limitation is made to resolution ratio.For example, the corresponding geographic range of unit pixel point can be in the image that first satellite 1 is shot
1km*The geographic range of 1km, the corresponding geographic range of unit pixel point can be 10m in the image that second satellite 2 is shot*
The geographic range of 10m, then first satellite 1 is low resolution satellite relative to the second satellite 2, and the second satellite 2 is relative to the
One satellite 1 is high-resolution satellite.
First satellite 1 and the second satellite 2 can shoot the remote sensing images of ground region.Due to second satellite
2 resolution ratio is relatively high, opposite for the identification of the natural calamity on some ground first in corresponding high-definition picture
The low-resolution image that satellite 1 is shot can be relatively easy to.But due to some such as Satellite Networking modes and equipment performance itself
Etc. the reason of, second satellite 2 to the filming frequency on ground may first satellite 1 it is low.Such as described
One satellite 1 is different with the orbit altitude of the second satellite 2, quantity difference and second satellite 2 in the satellite network
High-resolution lens the shooting service life it is shorter etc..It therefore, if can be timely over the ground with higher first satellite 1 of filming frequency
The object event (such as described natural calamity, vegetation pest and disease damage or other similar event) in face carries out early warning, it will has preferably i.e.
Shi Xing.
In some embodiments, first satellite 1 and the second satellite 2 may operate in the Earth's orbit of different height
On, two satellites are different around the period of earth operation at this time.For example, the running track of first satellite 1 is opposite
Lower track can may enclose in one day around the earth 2 at this time;The running track of second satellite 2 can be for relatively
High track can may enclose in one day around the earth one at this time.It is described for the geographic area that two satellites can cover
It can pass through twice in first satellite 1 one days, and be taken pictures twice, can be passed through once in second satellite 2 one days, and
It takes pictures primary.
In some embodiments, quantity of first satellite 1 in the satellite network can be defended more than described second
Quantity of the star 2 in the satellite network.For example two the first satellites 1 and second satellite 2 are arranged on same track, then institute
The frequency of taking pictures for stating two the first satellites 1 can take pictures for second satellite 2 twice of frequency.
In some embodiments, the service life for the high-resolution camera arrangement that second satellite 2 carries is compared to described the
The low resolution camera arrangement that one satellite 1 carries may be shorter, or due to other reasons cause to need after its shooting is primary every
A period of time could shoot again, this, which is equally possible, causes the filming frequency of second satellite 2 to be likely less than described first
Satellite 1.
Whether the system or method that the application proposes can obtain target area according to the image that first satellite 1 is shot
The relatively reliable judgement that object event occurs, to make early warning in time.It, can also be in institute meanwhile if conditions permit
The second satellite 2 is stated by whether making confirmation judgement to the object event when target area.
Fig. 2 is the exemplary process diagram of one embodiment that early warning is carried out using remote sensing satellite in the application.The process
It mainly include tentatively judging whether target area occurs object event from the image that first satellite 1 is shot, and pass through institute
The image for stating the shooting of the second satellite 2 further confirms that.
In 210, the available first resolution image of processing equipment, the first resolution image includes target area
Domain, the first resolution image are shot by the first equipment.What available first satellite 1 of processing equipment was shot includes target area
The first resolution image in domain.The processing equipment can be the electronic equipment of any ground-based electronic equipment or Seeds of First Post-flight.When
When the processing equipment is ground-based electronic equipment, it can be communicated with first satellite 1 and/or second satellite 2,
Obtain the image of its shooting.It, can be in satellite sheet when the processing equipment is the processing equipment that first satellite 1 carries
Ground executes method for early warning described herein.
First satellite 1 can periodically shoot the ground region of its track covering.Such as described first
Satellite 1 can shoot the ground region immediately below it every 10 minutes, and the ground region can be described at this time
Target area.In some embodiments, it is also possible to which the topographic(al) data of combined ground determines the target area.Such as one
It include the first resolution image of half land and half ocean, it is believed that the land area is the target area
Domain.In some embodiments, the identification of the target area can make different judgements according to the type of the object event.Than
When such as the object event is forest fire, then the target area can only include wood land.First satellite 1 can also
With targetedly region is shot on a surface target.For example, when the first satellite 1 flies over the target area overhead,
Once shot.First satellite 1 can be sent to the processing equipment after taking the first resolution image.
In some embodiments, it is (or anti-that the photographing device that first satellite 1 carries can capture ground target transmitting
Penetrate) multiwave electromagnetic wave.The multiwave electromagnetic wave can be optional frequency section spectrally, such as visible light wave
Section, near infrared band, far infrared band etc..It may include the multiwave data on the first resolution image, and
The data (such as at the time of shooting etc.) of some other reaction first resolution image attributes.
In 220, the processing equipment can extract the corresponding feature of each pixel in the first resolution image
Vector.It may include a series of feature of pixels in described eigenvector.The feature may include at least one wave
Parameter, weather parameters, terrain parameter, time domain parameter, spatial parameter, the vegetation index etc. of section.A part in the feature is (such as
Parameter, spatial parameter, vegetation index of at least one wave band etc.) it can be mentioned from the image data of the first resolution image
It takes, another part can be in conjunction with the time of the shooting of the first resolution image and the position of the target area from data
It is transferred in library.
The parameter of at least one wave band can refer to the received different-waveband of photographing device of first satellite 1
Embodiment of the electromagnetic wave in the first resolution image.For example, any one pixel in the first resolution image
It may include visible optical parameter, near-infrared parameter, indicate the electromagnetic wave of the corresponding geographic range transmitting (or reflection) of the pixel
In visible light wave range and near infrared band part captured by the photographing device of first satellite 1 after the parameter value that is converted to.
For certain object event, when occurring, one or more wave bands in the electromagnetic wave of target area transmitting may be sent out
It is raw abnormal, to be reacted in the parameter of the pixel of the first resolution image.For example, in the case of fire, target area
The infrared band electromagnetic wave enhancing of domain transmitting, then correspond to the infrared of the pixel of the target area in the first resolution image
The parameter value of wave band may more usually generate variation.
When the weather parameters can be the reflection first resolution image taking, the target area weather condition
Parameter.For all pixels point in first resolution image, corresponding weather parameters can be identical.Such as institute
State first resolution image capture moment, the weather of the target area be it is sunny, then it is each in the first resolution image
The characteristic parameter of pixel can be sunny.The processing equipment can be according to the shooting time of the first resolution image
And the geographical location of the target area therein, searching database is to obtain the weather parameters.The database can be with
Including multiple target areas multiple moment Weather information.When being identified to certain object events, the shadow of weather conditions
Ringing may be larger.For example, judge whether occur fire when, if the weather condition of target area be it is rainy, fire occurs
A possibility that it is lower.
The terrain parameter can reflect the topographic features of the target area.The topographic features may include mountainous region,
Grassland, wetland etc..It is similar with the weather parameters, for all pixels point in first resolution image, can correspond to same
A terrain parameter.It is similar, the processing equipment can be retrieved according to the geographical location of the target area database with
Obtain the terrain parameter.When identifying to certain object events, the influence of landform may be larger.For example, being in judgement
When no generation vegetation pest and disease damage, if the landform of target area is grassland, a possibility that vegetation pest and disease damage occurs, is lower.
At least one feature that the time domain parameter can reflect the pixel changes with time trend.The processing is set
Standby available other moment shooting closed on the first resolution image capture moment includes the target area
Other first resolution images.Described other first resolution images can be to be shot by first satellite 1, is also possible to
It is shot by other satellites.For example, the shooting time of the first resolution image can be T=5.Then the processing equipment can
To obtain other first resolution images including the target area of T=1~4 and the moment of T=6~10.This series of drawing
As in, some feature (such as visible light, thermal infrared etc.) of the corresponding pixel with geographic area is in the time range
Variation tendency can be used as the time domain parameter.For some kinds of object event, this time domain parameter may be with mesh
Whether mark event has stronger relevance.For example, when the object event is fire, the electricity of the target area transmitting
Thermal infrared part in magnetic wave may enhance rapidly when fire just occurs, and as intensity of a fire decrease reduces gradually.This heat
IR parameters change over time characteristic may with fire there are stronger positive correlations.
The spatial parameter can reflect the feature situation of at least one pixel around the pixel.The feature can be with
Including the multiple wave band parameter, the vegetation index or it is other similar can be from the number of the first resolution image zooming-out
According to.By taking infrared band parameter as an example, the corresponding infrared band parameter of the pixel itself can be X, around several pixels
The mean value (being also possible to maximum value or minimum value) of infrared band parameter may be Y.If X corresponds to infrared when fire occurs
Band strength, but Y corresponds to infrared band intensity when not having fire, then and a energy expression can in the corresponding region of the pixel
The event, such as the burning behavior of human controllable etc. of similar fire but non-fire can have occurred.At this point, several pixels around described
The mean value Y of the infrared band parameter of point can be used as a feature of the pixel.
The vegetation index can reflect the vegetation of the pixel, for example the vegetation index of the pixel is higher
When, it can indicate that the coverage rate of vegetation in the corresponding geographic range of the pixel is higher, it is denseer.In some embodiments, institute
The near-infrared parameter of the pixel and the ratio of IR parameters can be represented as by stating vegetation index.The object event is to occur
When vegetation pest and disease damage, the vegetation index can reflect the degree of pest and disease damage generation.
In 230, the processing equipment can by after the corresponding feature vector input Early-warning Model of each pixel,
The output of the Early-warning Model is obtained as a result, the output result includes whether the target area occurs object event.It is described
Early-warning Model can be trained model in advance, and store computer program in the storage device, and the processing is set
It is standby to transfer the Early-warning Model at any time.The function of the Early-warning Model can be, and differentiate once inputting described first to it
The feature vector of each pixel in rate image, can export about the target area in the first resolution image whether
The judging result of object event occurs.Training process about the Early-warning Model is see Fig. 3~5 and its associated description.
In 240, when the output result of the Early-warning Model is that object event occurs for target area, the processing equipment can
To issue warning information.The processing equipment can lead to by the way that display equipment, loudspeaker apparatus etc. issues warning information instruction
Know that staff there may be object event.It, can be with when the processing equipment is the processor that first satellite 1 carries
Equipment sends the warning information to the ground.
In 250, the processing equipment can call the second equipment to obtain second resolution image, the second resolution
Image includes the target area.What available second satellite 2 of the processing equipment was shot includes the of the target area
Two image in different resolution.The resolution ratio of the second resolution image is more described, and first resolution image is higher, can be used to confirm that
Whether the object event occurs.For the object event shorter similar to duration such as fire, if second satellite 2
It shoots too late at the time of the second resolution image, then can be omitted this step.For the target thing of vegetation pest and disease damage etc
Part, pest and disease damage are time-continuing processes, can further be shot with second satellite 2 the second resolution image into
Row confirmation.Second satellite 2 can take pictures to the target area when it flies over the target area.If described second
The track of satellite 2 cannot cover the target area, and the processing equipment can call spare capture apparatus, such as described in scheduling
Second satellite 2 carries out becoming rail change rail, and the overhead for making it run to the target area is taken pictures.
In 260, the processing equipment can confirm that mesh occurs for the target area according to the second resolution image
Mark event.Since the resolution ratio of the second resolution image is higher, thus the processing equipment can relatively easily (such as
Pass through trained image recognition model) tell whether object event occurs from the second resolution image.
Fig. 3 is the exemplary process diagram of one embodiment of the training Early-warning Model in the application.The process mainly includes
The specific location that object event occurs is determined according to the high-definition picture that a large amount of second satellite 2 is shot, then according to
Specific location retrieves the low-resolution image of the shooting of the first satellite 1 when the object event is occurring, with these low point
Resolution image is trained the Early-warning Model as training sample, so that the Early-warning Model can pass through low resolution
Rate image judges whether that object event occurs.
In 310, the available multiple sample images of processing equipment, the resolution ratio of the multiple sample image is not less than the
Two resolution ratio, the corresponding incident region of each image in the multiple sample image, described in each incident region occurs
Object event.The sample image may include multiple second resolution images that second satellite 2 is shot.Here processing
Equipment can be the equipment with data-handling capacity different from processing equipment described in Fig. 2.The second resolution figure
The high-definition picture shot as can be second satellite 2, the processing equipment or manually can relatively easily from
It tells thereon and the object event whether occurred.The region that the object event occurred can be the incident region,
Indicate once there is object event to occur in the region, but the time occurred can not determine.For example, in citing before,
The filming frequency of second satellite 2 may be lower, for same ground region, may could shoot within 1 day or more days once,
For the fire time, it fiery may extinguish when shooting, the range burnt can only have been taken, the model burnt here
It encloses as the incident region.
In 320, for each of the multiple sample image image: the processing equipment can be according to the image
Determine the history not less than first resolution in corresponding incident region when at least one described object event is occurring
Image.At least one each history image may include at least one the history first resolution shot by first satellite 1
Image.Each second resolution image corresponds to its incident region, and the object event period is occurring in the incident region,
May there is more than one first satellite 1 to be taken pictures more than once to the incident region, i.e., it is described at least one
History first resolution image.At least one described history first resolution image can store in the database, the processing
Equipment can obtain from the database, and specific acquisition methods are see Fig. 4 and Fig. 5 and its associated description.
In 330, the processing equipment can be using the multiple history image as Early-warning Model described in sample training.It is right
In each history first resolution image, the processing equipment can extract the wherein corresponding feature vector of each pixel,
And using feature vector as the training sample of the Early-warning Model.Here be used to the feature vector feature quantity that includes of training with
The feature quantity that type can include with the feature vector described in being applied when Early-warning Model progress object event judgement in Fig. 2
It is identical as type.The Early-warning Model may include general image segmentation algorithm, as FPN (characteristic pattern pyramid network,
Feature pyramid networks) model.
Fig. 4 is the exemplary process diagram of one embodiment of at least one history image described in determination in the application.It should
Process mainly includes according to the shooting time of the second resolution image, and institute may be occurred by estimating in conjunction with some other factor
The time range of object event is stated, and selects at least one described history first in the corresponding image of this time range and differentiates
Rate image.
In 410, the processing equipment can identify the incident region in the second resolution image and obtain the image
Shooting time.Since the resolution ratio of the second resolution image is higher, the processing equipment can be by some general
Image recognition algorithm (such as deep learning) identifies the incident region in image.It simultaneously can also in the second resolution image
To include the data of its shooting time, can be obtained together by the processing equipment.The mesh is occurring for the incident region
The time of mark event will not be later than the shooting time of the second resolution image the latest.
In 420, the processing equipment can determine candidate time according to the incident region and the shooting time
It include the candidate history first resolution image in the incident region in range, when the candidate time range includes the shooting
It carves.The processing equipment can determine the candidate time range first.The candidate time range can indicate the target
Event may the occurent period in the incident region.
In some embodiments, the processing equipment can determine the time based on the shooting period of second satellite 2
Select time range.For example second satellite 2 can be two days to the shooting period of areal, if before two days, it is described
Second satellite 2 does not have to find the trace of the object event in the image of the shooting in the incident region, then can incite somebody to action
The candidate time range is determined as the past time interval a few days ago of shooting time of the second resolution image.
In some embodiments, the processing equipment can also determine the time range based on some third party's data.
For example, third party's data can be the approximate time of the generation of object event described in news report.Such as certain day news
The object event is reported, then the candidate time range can be determined as to the zero point of this day to the second resolution figure
The shooting time of picture.
After determining the candidate moment, the processing equipment can be retrieved from database based on the candidate moment
The history first resolution figure including the incident region shot within the scope of the candidate time by first satellite 1 out
As the candidate history first resolution image.
In 430, the processing equipment can determine described at least one from the candidate history first resolution image
A history image.In candidate's history first resolution image, it may include some object events and do not occur also, or
Image through having crossed, so needing further to screen the candidate history first resolution image, to obtain
State at least one described history first resolution image when object event is occurring.The detailed description of the screening is see figure
5 and its associated description.
Fig. 5 is to determine at least one described history image from the candidate history first resolution image in the application
One embodiment exemplary process diagram.The process mainly includes according to some in candidate history first resolution image
Or the variation of multiple parameters, the period that the object event is occurring is judged, so that it is determined that it is corresponding to go out the period
History first resolution image.
In 510, the processing equipment can be according at least one feature in the candidate history first resolution image
Parameter changes with time trend, determines that incident time range, the candidate time range include the incident time range.Institute
Stating incident time range can refer to that the occurent time range of the object event, the time range can be the candidate
The a subset of time range.For the object event, it is reflected in the pixel on the first resolution image
One or more features may generate significant change in the incident time range, then when generating this part of significant change
Between section can be the incident time range.For example, in schematic diagram shown in Fig. 6, horizontal axis indicates to wait by taking fire as an example
The time span of time range is selected, the longitudinal axis indicates the parameter value of some feature (such as thermal infrared parameter), and curve can be in reference axis
Reflect the intensity of a fire from the process started to extinguishing.The period of t1 to t2 corresponding parameter value can be for the fire in figure
The period of generation, i.e. the incident time range.
In 520, the processing equipment can determine the corresponding candidate history first resolution of the incident time range
Image is at least one described history image.After determining the incident time range, the processor can be by the candidate
In history first resolution image, image of the shooting time in the incident time range is as at least one described history figure
Picture.
Fig. 7 is the schematic diagram of one embodiment of the object event early warning system in the application.The object event early warning
System 700 may include image acquisition unit 710, feature extraction unit 720, model arithmetic unit 730, prewarning unit 740, figure
As call unit 750, confirmation unit 760 and scheduling unit 770.
Described image acquiring unit 710 can be used for obtaining first resolution image, and the first resolution image includes
Target area, the first resolution image are shot by the first equipment.The feature extraction unit 720 can be used for extracting described
The corresponding feature vector of each pixel in first resolution image.The model arithmetic unit 730 can be used for will be described every
After the corresponding feature vector input Early-warning Model of a pixel, the output of the Early-warning Model is obtained as a result, the output result
Whether object event occurs including the target area.The prewarning unit 740 can be used for the output knot of the Early-warning Model
When fruit is that object event occurs for target area, warning information is issued.
Described image call unit 750 can be used for calling the second equipment to obtain second resolution image, and described second point
Resolution image includes the target area.The confirmation unit 760 can be used for confirming institute according to the second resolution image
It states target area and object event occurs, wherein the corresponding geographic range of unit pixel point of the first resolution image is greater than
The corresponding geographic range of unit pixel point of the second resolution image.
The confirmation unit 760 determines that the output result includes institute after object event may occur for the target area
Stating scheduling unit 770 can call stand-by equipment to shoot the second resolution image.
In conclusion after reading this detailed disclosures, it will be understood by those skilled in the art that aforementioned detailed disclosure
Content can be only presented in an illustrative manner, and can not be restrictive.Although not explicitly described or shown herein, this field skill
Art personnel are understood that improve and modify it is intended to include the various reasonable changes to embodiment.These change, improve and
It modifies and is intended to be proposed by the disclosure, and in the spirit and scope of the exemplary embodiment of the disclosure.
In addition, certain terms in the application have been used for describing implementation of the disclosure example.For example, " one embodiment ", "
Embodiment " and/or " some embodiments " means to combine the special characteristic of embodiment description, structure or characteristic may include
In at least one embodiment of the disclosure.Therefore, can emphasize and it is to be understood that in the various pieces of this specification to "
Embodiment " or " one embodiment " or " alternate embodiment " two or more references be not necessarily all referring to identical implementation
Example.In addition, special characteristic, structure or characteristic can be appropriately combined in one or more other embodiments of the present disclosure.
It should be appreciated that in the foregoing description of embodiment of the disclosure, in order to help to understand a feature, originally for simplification
Disclosed purpose, the application sometimes combine various features in single embodiment, attached drawing or its description.Alternatively, the application is again
Be by various characteristic dispersions in multiple the embodiment of the present invention.However, this be not to say that the combination of these features be it is necessary,
Those skilled in the art are entirely possible to come out a portion feature extraction as individual when reading the application
Embodiment understands.That is, embodiment in the application it can be appreciated that multiple secondary embodiments integration.And it is each
The content of secondary embodiment is also to set up when being less than individually all features of aforementioned open embodiment.
In some embodiments, the quantity or property for certain embodiments of the application to be described and claimed as are expressed
The number of matter is interpreted as in some cases through term " about ", " approximation " or " substantially " modification.For example, unless otherwise saying
Bright, otherwise " about ", " approximation " or " substantially " can indicate ± 20% variation of the value of its description.Therefore, in some embodiments
In, the numerical parameter listed in written description and the appended claims is approximation, can be tried according to specific embodiment
Scheme the required property obtained and changes.In some embodiments, numerical parameter should be according to the quantity of the effective digital of report simultaneously
It is explained by the common rounding-off technology of application.Although illustrating that some embodiments of the application list broad range of numerical value
Range and parameter are approximations, but numerical value reported as precisely as possible is all listed in specific embodiment.
Herein cited each patent, patent application, the publication and other materials of patent application, such as article, books,
Specification, publication, file, article etc. can be incorporated herein by reference.Full content for all purposes, in addition to
Its relevant any prosecution file history, may or conflicting any identical or any possibility inconsistent with this document
On any identical prosecution file history of the restrictive influence of the widest range of claim.Now or later and this document
It is associated.For example, if in description, definition and/or the use of term associated with any included material and this
The relevant term of document, description, definition and/or between there are it is any inconsistent or conflict when, be using the term in this document
It is quasi-.
Finally, it is to be understood that the embodiment of application disclosed herein is the explanation to the principle of the embodiment of the application.
Other modified embodiments are also within the scope of application.Therefore, herein disclosed embodiment it is merely exemplary rather than
Limitation.Those skilled in the art can take alternative configuration according to the embodiment in the application to realize the invention in the application.
Therefore, embodiments herein is not limited to which embodiment accurately described in application.
Claims (17)
1. a kind of pair of object event carries out pre-warning system characterized by comprising
Image acquisition unit, for obtaining first resolution image, the first resolution image includes target area, and described
One image in different resolution is shot by the first equipment;
Feature extraction unit, for extracting the corresponding feature vector of each pixel in the first resolution image;
Model arithmetic unit, for will be after the corresponding feature vector input Early-warning Model of each pixel, acquisition be described pre-
The output of alert model is as a result, the output result includes whether the target area occurs object event;And
Prewarning unit issues warning information when the output result for the Early-warning Model is that object event occurs for target area.
2. system according to claim 1, which is characterized in that further comprise:
Image call unit, for calling the second equipment to obtain second resolution image, the second resolution image includes institute
State target area;And
Confirmation unit, for confirming that object event occurs for the target area according to the second resolution image,
Wherein, the corresponding geographic range of unit pixel point of the first resolution image is greater than the second resolution image
The corresponding geographic range of unit pixel point.
3. system according to claim 2, which is characterized in that the system further comprises scheduling unit, the calling
Second equipment obtains second resolution image
The confirmation unit determines that the output result includes that object event may occur for the target area;
The scheduling unit calls stand-by equipment to shoot the second resolution image.
4. according to claim 1 or system described in 2 or 3, which is characterized in that the first resolution image is that satellite is distant
Feel image;The second resolution image is the satellite remote sensing images for target area;The stand-by equipment is in-orbit space flight
Device.
5. system according to claim 1, which is characterized in that the corresponding feature vector of each pixel includes following
At least one of feature:
The parameter of at least one wave band, the wave band include the optional frequency section in spectrum;
Weather parameters reflects the weather condition of the target area when first resolution image taking;
Terrain parameter reflects the topographic features of the target area;
Time domain parameter reflects that at least one feature of the pixel changes with time trend;And
Spatial parameter reflects the feature situation of at least one pixel around the pixel.
6. according to the method described in claim 5, it is characterized in that, the feature further comprises vegetation index, the vegetation
Index includes the near-infrared parameter of the pixel and the ratio of IR parameters.
7. the method that a kind of pair of object event carries out early warning characterized by comprising
First resolution image is obtained, the first resolution image includes target area, and the first resolution image is by the
The shooting of one equipment;
Extract the corresponding feature vector of each pixel in the first resolution image;
After the corresponding feature vector input Early-warning Model of each pixel, obtain the output of the Early-warning Model as a result,
The output result includes whether the target area occurs object event;And
When the output result of the Early-warning Model is that object event occurs for target area, warning information is issued.
8. the method according to the description of claim 7 is characterized in that the method further includes:
The second equipment is called to obtain second resolution image, the second resolution image includes the target area;And
Confirm that object event occurs for the target area according to the second resolution image,
Wherein, the corresponding geographic range of unit pixel point of the first resolution image is greater than the second resolution image
The corresponding geographic range of unit pixel point.
9. according to the method described in claim 8, it is characterized in that, the second equipment of the calling obtains second resolution image packet
It includes:
Determine that the output result includes that object event may occur for the target area;
Stand-by equipment is called to shoot the second resolution image.
10. according to method described in claim 7 or 8 or 9, which is characterized in that the first resolution image is satellite
Remote sensing images;The second resolution image is the satellite remote sensing images for target area;The stand-by equipment is in-orbit boat
Its device.
11. the method according to the description of claim 7 is characterized in that the corresponding feature vector of each pixel include with
At least one of lower feature:
The parameter of at least one wave band, the wave band include the optional frequency section in spectrum;
Weather parameters reflects the weather condition of the target area when first resolution image taking;
Terrain parameter reflects the topographic features of the target area;
Time domain parameter reflects that at least one feature of the pixel changes with time trend;And
Spatial parameter reflects the feature situation of at least one pixel around the pixel.
12. according to the method for claim 11, which is characterized in that the feature further comprises vegetation index, the plant
It include the near-infrared parameter of the pixel and the ratio of IR parameters by index.
13. the method according to the description of claim 7 is characterized in that the Early-warning Model passes through default training process acquisition, institute
Stating default training process includes:
Multiple sample images are obtained, the resolution ratio of the multiple sample image is not less than second resolution, the multiple sample graph
The corresponding incident region of each image as in, the object event occurs for each incident region;
For each of the multiple sample image image:
The corresponding incident region not less than the when determining that at least one described object event is occurring according to the image
The history image of one resolution ratio;And
Using the multiple history image as Early-warning Model described in sample training.
14. according to the method for claim 13, which is characterized in that at least one history image described in the determination includes:
It identifies the incident region in the image and obtains the shooting time of the image;
According to the incident region and the shooting time, the candidate within the scope of candidate time including the incident region is determined
History first resolution image, the candidate time range include the shooting time;And
At least one described history image is determined from the candidate history first resolution image.
15. according to the method for claim 14, which is characterized in that described from the candidate history first resolution image
Determine that at least one described history image includes:
It is changed with time trend according at least one characteristic parameter in the candidate history first resolution image, determines incident
Time range, the candidate time range include the incident time range;
Determine that the corresponding candidate history first resolution image of the incident time range is at least one described history image.
16. according to the method for claim 15, which is characterized in that described to make the multiple history first resolution image
For Early-warning Model described in sample training, comprising:
For each history first resolution image, extract the training feature vector of each of which pixel, the training characteristics to
Characteristic type in amount is identical as the characteristic type in described eigenvector;And
It using the training feature vector of each pixel as sample set, is input in the Early-warning Model, to the early warning
Model is trained.
17. a kind of electronic equipment, including memory, processor and it is stored on the memory and can transports on the processor
Capable computer program, which is characterized in that the processor is realized when executing the computer program as in claim 7 to 16
The step of described in any item pairs of object events carry out the method for early warning.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910004746.2A CN109766815A (en) | 2019-01-03 | 2019-01-03 | A kind of pair of object event carries out pre-warning system and method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910004746.2A CN109766815A (en) | 2019-01-03 | 2019-01-03 | A kind of pair of object event carries out pre-warning system and method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109766815A true CN109766815A (en) | 2019-05-17 |
Family
ID=66453498
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910004746.2A Pending CN109766815A (en) | 2019-01-03 | 2019-01-03 | A kind of pair of object event carries out pre-warning system and method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109766815A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110517435A (en) * | 2019-09-08 | 2019-11-29 | 天津大学 | The portable instant fire prevention early warning of one kind and Information Collecting & Processing early warning system and method |
CN110991333A (en) * | 2019-11-30 | 2020-04-10 | 内蒙古蒙草生命共同体大数据有限公司 | Aboveground biomass early warning method |
CN110991332A (en) * | 2019-11-30 | 2020-04-10 | 内蒙古蒙草生命共同体大数据有限公司 | Vegetation index early warning method |
CN112035679A (en) * | 2020-09-10 | 2020-12-04 | 广东新禾道信息科技有限公司 | Data processing method and device for remote sensing monitoring natural disasters |
CN115657035A (en) * | 2022-12-29 | 2023-01-31 | 北京九天微星科技发展有限公司 | Polarization synthetic aperture radar imaging method and device based on inter-satellite cooperation |
CN116206215A (en) * | 2023-03-17 | 2023-06-02 | 银河航天(北京)网络技术有限公司 | Forest land state monitoring method, forest land state monitoring device and storage medium |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105137421A (en) * | 2015-06-25 | 2015-12-09 | 苏州途视电子科技有限公司 | Photoelectric composite low-altitude early warning detection system |
CN105809372A (en) * | 2016-04-20 | 2016-07-27 | 国家电网公司 | Natural disaster risk monitoring system based on satellite remote sensing image |
CN106842350A (en) * | 2016-12-26 | 2017-06-13 | 首都师范大学 | Combine moving target detecting system and detection method with platform different resolution sensor |
CN106915462A (en) * | 2017-02-14 | 2017-07-04 | 福建兴宇信息科技有限公司 | Forestry pests & diseases intelligent identifying system based on multi-source image information |
CN108288344A (en) * | 2017-12-26 | 2018-07-17 | 李文清 | A kind of efficient forest fire early-warning system |
CN109086685A (en) * | 2018-07-11 | 2018-12-25 | 国家林业局森林病虫害防治总站 | Forestry biological hazards monitoring method and system based on satellite remote sensing images |
-
2019
- 2019-01-03 CN CN201910004746.2A patent/CN109766815A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105137421A (en) * | 2015-06-25 | 2015-12-09 | 苏州途视电子科技有限公司 | Photoelectric composite low-altitude early warning detection system |
CN105809372A (en) * | 2016-04-20 | 2016-07-27 | 国家电网公司 | Natural disaster risk monitoring system based on satellite remote sensing image |
CN106842350A (en) * | 2016-12-26 | 2017-06-13 | 首都师范大学 | Combine moving target detecting system and detection method with platform different resolution sensor |
CN106915462A (en) * | 2017-02-14 | 2017-07-04 | 福建兴宇信息科技有限公司 | Forestry pests & diseases intelligent identifying system based on multi-source image information |
CN108288344A (en) * | 2017-12-26 | 2018-07-17 | 李文清 | A kind of efficient forest fire early-warning system |
CN109086685A (en) * | 2018-07-11 | 2018-12-25 | 国家林业局森林病虫害防治总站 | Forestry biological hazards monitoring method and system based on satellite remote sensing images |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110517435A (en) * | 2019-09-08 | 2019-11-29 | 天津大学 | The portable instant fire prevention early warning of one kind and Information Collecting & Processing early warning system and method |
CN110991333A (en) * | 2019-11-30 | 2020-04-10 | 内蒙古蒙草生命共同体大数据有限公司 | Aboveground biomass early warning method |
CN110991332A (en) * | 2019-11-30 | 2020-04-10 | 内蒙古蒙草生命共同体大数据有限公司 | Vegetation index early warning method |
CN112035679A (en) * | 2020-09-10 | 2020-12-04 | 广东新禾道信息科技有限公司 | Data processing method and device for remote sensing monitoring natural disasters |
CN115657035A (en) * | 2022-12-29 | 2023-01-31 | 北京九天微星科技发展有限公司 | Polarization synthetic aperture radar imaging method and device based on inter-satellite cooperation |
CN116206215A (en) * | 2023-03-17 | 2023-06-02 | 银河航天(北京)网络技术有限公司 | Forest land state monitoring method, forest land state monitoring device and storage medium |
CN116206215B (en) * | 2023-03-17 | 2023-09-29 | 银河航天(北京)网络技术有限公司 | Forest land state monitoring method, forest land state monitoring device and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109766815A (en) | A kind of pair of object event carries out pre-warning system and method | |
Barclay et al. | A revised exoplanet yield from the transiting exoplanet survey satellite (TESS) | |
US8705887B2 (en) | Method and apparatus for filling in or replacing image pixel data | |
Goward et al. | Historical record of Landsat global coverage | |
CN1650622B (en) | Systems and methods for digitally re-mastering or otherwise modifying motion pictures or other image sequences data | |
Xu et al. | Arid: A new dataset for recognizing action in the dark | |
CN108717582A (en) | Forest fire prediction technique, device, computer equipment and storage medium | |
Stipaničev et al. | Advanced automatic wildfire surveillance and monitoring network | |
Mizunuma et al. | The comparison of several colour indices for the photographic recording of canopy phenology of Fagus crenata Blume in eastern Japan | |
da Luz et al. | The forestwatchers: a citizen cyberscience project for deforestation monitoring in the tropics | |
Brivio et al. | Mapping burned surfaces in Sub-Saharan Africa based on multi-temporal neural classification | |
CN102197606B (en) | Method and system for acquiring and repatriating satellite data | |
Coulter et al. | Automated detection of people and vehicles in natural environments using high temporal resolution airborne remote sensing | |
De Vries et al. | Results with a multispectral autonomous wildfire detection system | |
CN116152164A (en) | Night multispectral low-light remote sensing image brightness anomaly detection method and device | |
EP3921234B1 (en) | Computer systems for acquiring satellite images while taking into account meteorological uncertainty and corresponding satellite mission plan calculator | |
CN102667853A (en) | Filter setup learning for binary sensor | |
Matallana-Puerto et al. | Rarity up in the mountain: Ecological niche modeling, phenology, and reproductive biology of the most commercialized Masdevallia species | |
Gombe et al. | Characterization of urban heat island (UHI) changes from MODIS times series using principal component analysis (PCA): Case of Dar es Salaam City Tanzania | |
Campos | Image Manipulation and Classification: An Application to Fire Detection | |
George | Remote sensing of Earth resources: emerging opportunities for developing countries | |
Steinbring et al. | Radio Galaxies at z= 1.1 to 3.8: Adaptive Optics Imaging and Archival Hubble Space Telescope Data | |
Moeller | Low-light Camera field evaluation for underwater nighttime operations | |
CN109547804A (en) | Live broadcast service optimization method and device | |
Bonazountas | High resolution satellite imagery and Earth observation data policy |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190517 |