CN109145751A - Page turning detection method and device - Google Patents
Page turning detection method and device Download PDFInfo
- Publication number
- CN109145751A CN109145751A CN201810813117.XA CN201810813117A CN109145751A CN 109145751 A CN109145751 A CN 109145751A CN 201810813117 A CN201810813117 A CN 201810813117A CN 109145751 A CN109145751 A CN 109145751A
- Authority
- CN
- China
- Prior art keywords
- image
- detection
- detection image
- wire
- vegetarian refreshments
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
- G06V10/235—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition based on user input or interaction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Psychiatry (AREA)
- Social Psychology (AREA)
- Human Computer Interaction (AREA)
- Image Analysis (AREA)
Abstract
The embodiment of the present invention provides a kind of page turning detection method and device, belongs to computer application technology.This method comprises: extracting the characteristics of image of each detection image in multiple detection images during obtaining multiple detection images to detect to entity book to be detected；Any N number of detection image continuous for acquisition sequence in multiple detection images, if meeting preset condition between the characteristics of image of the last one detection image and the characteristics of image of other detection images in any N number of detection image, it is determined that the last one detection image is to shoot to obtain in the case where entity book to be detected is in page turning stable state；Wherein, after page turning stable state refers to the page turning movement completed to entity book to be detected, stable state locating for entity book to be detected.Due to requiring no matched process one by one, for any detection image, only the corresponding page turning state of the detection image need to can determine according to several detection images before the detection image, to improve detection efficiency.
Description
Technical field
The present embodiments relate to computer application technologies, more particularly, to a kind of page turning detection method and dress
It sets.
Background technique
With information-based and intelligentized development, the intelligent arrangement for reading for reading entity book is more more and more universal.Wherein,
User is when reading entity book using intelligent arrangement for reading, usually by the camera on intelligent arrangement for reading according to predeterminated frequency
Entity book is shot, to obtain content pages image to be detected according to the image that shooting obtains.Wherein, since camera may
User can be taken just in the image of page turning, to need to obtain the content pages image after the completion of user's page turning movement.In correlation
In technology, usually successively the image that shooting obtains is matched one by one with the entity book content pages image prestored, until inspection
The shooting image that the entity book content pages image for measuring and prestoring matches, then using the shooting image as content pages to be detected
Image.Due to needing to match the obtained image of shooting with determination content pages image to be detected one by one, thus efficiency compared with
It is low.
Summary of the invention
To solve the above-mentioned problems, the embodiment of the present invention provides one kind and overcomes the above problem or at least be partially solved
State the page turning detection method and device of problem.
According to a first aspect of the embodiments of the present invention, a kind of page turning detection method is provided, this method comprises:
During obtaining multiple detection images to detect to entity book to be detected, extract in multiple detection images
The characteristics of image of each detection image；
Any N number of detection image continuous for acquisition sequence in multiple detection images, if in any N number of detection image most
Meet preset condition between the characteristics of image of the latter detection image and the characteristics of image of other detection images, it is determined that last
A detection image is to shoot to obtain in the case where entity book to be detected is in page turning stable state；Wherein, N is the integer greater than 1, is turned over
After page stable state refers to the page turning movement completed to entity book to be detected, stable state locating for entity book to be detected.
Method provided in an embodiment of the present invention, by obtaining multiple detection images to detect to entity book to be detected
During, extract the characteristics of image of each detection image in multiple detection images.For acquisition sequence in multiple detection images
Continuous any N number of detection image, if the characteristics of image of the last one detection image and other detections in any N number of detection image
Meet preset condition between the characteristics of image of image, it is determined that the last one detection image is to be in page turning in entity book to be detected
Shooting obtains under stable state.It, only need to be according to the inspection for any detection image due to requiring no matched process one by one
Several detection images before altimetric image can determine the corresponding page turning state of the detection image, to improve detection efficiency.
According to a second aspect of the embodiments of the present invention, a kind of page turning detection device is provided, which includes:
Extraction module, for extracting during obtaining multiple detection images to detect to entity book to be detected
The characteristics of image of each detection image in multiple detection images；
First determining module, for any N number of detection image continuous for acquisition sequence in multiple detection images, if appointing
Meet between the characteristics of image of the last one detection image and the characteristics of image of other detection images in N number of detection image of anticipating default
Condition, it is determined that the last one detection image is to shoot to obtain in the case where entity book to be detected is in page turning stable state；Wherein,
N is the integer greater than 1, after page turning stable state refers to the page turning movement completed to entity book to be detected, entity book institute to be detected
The stable state at place.
According to a third aspect of the embodiments of the present invention, a kind of electronic equipment is provided, comprising:
At least one processor；And
At least one processor being connect with processor communication, in which:
Memory is stored with the program instruction that can be executed by processor, and the instruction of processor caller is able to carry out first party
Page turning detection method provided by any possible implementation in the various possible implementations in face.
According to the fourth aspect of the invention, a kind of non-transient computer readable storage medium, non-transient computer are provided
Readable storage medium storing program for executing stores computer instruction, and computer instruction makes the various possible implementations of computer execution first aspect
In page turning detection method provided by any possible implementation.
It should be understood that above general description and following detailed description be it is exemplary and explanatory, can not
Limit the embodiment of the present invention.
Detailed description of the invention
Fig. 1 is a kind of flow diagram of page turning detection method of the embodiment of the present invention；
Fig. 2 is a kind of flow diagram of page turning detection method of the embodiment of the present invention；
Fig. 3 is a kind of flow diagram of page turning detection method of the embodiment of the present invention；
Fig. 4 is a kind of flow diagram of page turning detection method of the embodiment of the present invention；
Fig. 5 is a kind of block diagram of page turning detection device of the embodiment of the present invention；
Fig. 6 is the block diagram of a kind of electronic equipment of the embodiment of the present invention.
Specific embodiment
With reference to the accompanying drawings and examples, the specific embodiment of the embodiment of the present invention is described in further detail.With
Lower embodiment is not limited to the range of the embodiment of the present invention for illustrating the embodiment of the present invention.
With information-based and intelligentized development, the intelligent arrangement for reading for reading entity book is more more and more universal.Intelligence
Arrangement for reading is mainly the explanation recording and the cover image and content pages figure of entity book by the way that some entity books are stored in advance
Picture is intelligently read to realize.User is when reading entity book using intelligent arrangement for reading, usually by intelligent arrangement for reading
Camera shoot entity book according to predeterminated frequency, to obtain content pages image to be detected according to the obtained image of shooting.
In shooting process, user may stir entity book to smart machine, and the image obtained so as to cause shooting may be comprising using
The gesture motion of entity book is stirred at family, namely can take user just in the image of page turning.And just the image of page turning can be due to
Content is not clear enough, so as to cause that can not carry out content detection.
In the related art, usually successively will the obtained image of shooting carried out with the entity book content pages image prestored by
One matching, until the shooting image that the entity book content pages image that detects and prestore matches, then using the shooting image as
Content pages image to be detected.Due to needing the image for obtaining shooting to be matched one by one with determination content pages figure to be detected
Picture, so that efficiency is lower.For example, since user's page turning movement needs the time in sustained segment, and camera can continual shooting
Image, to need just detect and entity book content pages until to act corresponding shooting image equal after it fails to match for page turning
The shooting image that image matches.Therefore, overall process expends overlong time and efficiency is lower.
For the demand, the embodiment of the invention provides a kind of page turning detection methods.It should be noted that the present invention is real
Applying the entity book that example is related to can be to draw the different reading matter types such as sheet, magazine and magazine, and the embodiment of the present invention does not make this to have
Body limits.It in addition, the corresponding executing subject of this method can be the intelligent arrangement for reading with read function, or is solution decoupling
Read function, only has the smart machine of entity book detection function, and the embodiment of the present invention is also not especially limited this.In order to just
In explanation, the embodiment of the present invention detects page turning provided in an embodiment of the present invention so that executing subject is intelligent arrangement for reading as an example
Method is illustrated.Referring to Fig. 1, this method comprises:
101, during obtaining multiple detection images to detect to entity book to be detected, multiple detection figures are extracted
The characteristics of image of each detection image as in.
Before executing 101, entity book to be detected first can will be turned to cover by user, to carry out cover inspection by camera
Any survey, to determine the entity book for this entity book.After determining which this entity book the entity book is, user can turn to entity book
Content pages, and read by intelligent arrangement for reading.Camera on intelligent arrangement for reading is according to predeterminated frequency to entity book
Captured in real-time is carried out, multiple detection images is obtained and carries out content detection.
For any detection image, the corresponding characteristics of image of the detection image is for determining the corresponding page turning of the detection image
State.Wherein, the corresponding characteristics of image of the detection image is extracted based on the detection image itself, and page turning state may include
Page turning carries out state and page turning stable state.After page turning stable state refers to the page turning movement completed to entity book to be detected, to
Detect stable state locating for entity book.When page turning progress state refers to that the movement of the page turning to entity book to be detected is carrying out,
State is stirred locating for entity book to be detected.
102, any N number of detection image continuous for acquisition sequence in multiple detection images, if any N number of detection image
In meet preset condition between the characteristics of image of the last one detection image and the characteristics of image of other detection images, it is determined that most
The latter detection image is to shoot to obtain in the case where entity book to be detected is in page turning stable state.
Wherein, N is the integer greater than 1, after page turning stable state refers to the page turning movement completed to entity book to be detected, to
Detect stable state locating for entity book.Any N number of detection image continuous for acquisition sequence in multiple detection images, can incite somebody to action
The characteristics of image (namely n-th detection image) of the last one detection image and the detection of front N-1 are schemed in N number of detection image
The characteristics of image of picture is compared, to determine whether to meet preset condition.If meeting preset condition, it can determine that n-th detects
Image is to shoot to obtain in the case where entity book to be detected is in page turning stable state.
Since camera filming frequency is usually very fast, when page turning acts progress and after the completion of page turning acts, camera
The multiple images after the completion of the multiple detection images and page turning movement when page turning movement carries out can be taken.For example, if turning over
Camera has taken M detection image when page movement carries out, and after the completion of page turning movement, camera has taken Z detection
Image.According to above-mentioned judgment mode, however, it is determined that Z detection image is clapped in the case where entity book to be detected is in page turning stable state
It takes the photograph, then it can be using first detection image in Z detection image as the detection pair of content detection during actual implementation
As perhaps an optional detection image is schemed as the test object of content detection or by Z detection from Z detection image
Test object of the last one detection image as content detection as in, or by next detection image of Z detection image
As the test object of content detection, the present invention is not especially limit this.Wherein, the value of above-mentioned Z can basis
Demand is configured, and such as value 2 to 6, the embodiment of the present invention is also not especially limited this.
It should be noted that in view of during actual implementation, according to above-mentioned judgment mode, it is possible that continuous multiple
Detection image is determined to be what the shooting in the case where entity book to be detected is in page turning carry out state obtained, but user turns over the process of one page
It there is no fear of continuing so long, to be determined to be at entity book to be detected if there is continuous Y detection image
Shooting obtains under page turning stable state, and Y is greater than preset threshold, then can directly examine Y+1 detection image as content
The test object of survey delays the long time to avoid because of lasting judgement error.
Method provided in an embodiment of the present invention, by obtaining multiple detection images to detect to entity book to be detected
During, extract the characteristics of image of each detection image in multiple detection images.For acquisition sequence in multiple detection images
Continuous any N number of detection image, if the characteristics of image of the last one detection image and other detections in any N number of detection image
Meet preset condition between the characteristics of image of image, it is determined that the last one detection image is to be in page turning in entity book to be detected
Shooting obtains under stable state.It, only need to be according to the inspection for any detection image due to requiring no matched process one by one
Several detection images before altimetric image can determine the corresponding page turning state of the detection image, to improve detection efficiency.
Content based on the above embodiment, as a kind of alternative embodiment, characteristics of image is Hash difference value, and N is greater than 2
Integer, preset condition is that the corresponding average Hamming distance of the last one image is not more than the first preset threshold or the last one figure
As corresponding maximum Hamming distance is not more than the second preset threshold；Correspondingly, determining that the last one detection image is to be checked
It surveys entity book be under page turning stable state before shooting obtains, average Hamming distance and maximum Hamming distance can also be calculated
From the present invention is not especially limit this.Referring to fig. 2, including but not limited to:
201, the Kazakhstan of each detection image in the Hash difference value and other detection images of the last one image is calculated separately
Difference between uncommon difference value.
In order to make it easy to understand, the Hash matrix of N number of detection image can carry out table with p1, p2 and p3 respectively so that N is 3 as an example
Show.Calculate the Hamming distance ham2 between the Hamming distance ham1 and p3 and p2 between p3 and p1.Wherein, Hash matrix can
Think 0,1 two values matrix of 9*8, Hamming distance is the number of elements that position is identical but numerical value is different in two matrixes.For example, if
A is (1,0,0,1,1,0,1), and B is (0,1,1,0,0,1,1), then the Hamming distance between AB is 6.It should be noted that practical
N can be with value 3 to 6 in implementation process, and the present invention is not especially limit this.
202, the average value for calculating all differences using average value as average Hamming distance, and is chosen from all differences
Maximum value, using maximum value as maximum Hamming distance.
Specifically, since the quantity of all differences is 2, so that the average value of all differences is (ham1+ham2)/2,
As be averaged Hamming distance.And the maximum value of all differences is represented by max (ham1, ham2), is also maximum Hamming distance.
It should be noted that can be based on when calculating the corresponding average Hamming distance of the Z detection image and maximum Hamming distance every time
(Z-N-1) a detection image to the detection image that Z detection image namely total quantity are N characteristics of image, to calculate the
The corresponding average Hamming distance of Z detection image and maximum Hamming distance.
As can be seen from the above embodiments, page turning state further includes that page turning carries out state other than including page turning stable state.
Correspondingly, if the corresponding average Hamming distance of the Z detection image is greater than pair of the first preset threshold and the Z detection image
The maximum Hamming distance answered is greater than the second preset threshold, then can determine that the Z detection image is to be in turn in entity book to be detected
Shooting obtains under page carry out state.
It should be noted that by the content in above-mentioned 201 it is found that Hamming distance is mensurable will in such a way that character is replaced
One character string is converted into required minimum replacement number when another character string, and average Hamming distance and maximum Hamming distance
From being in accordance with what the difference of Hash difference value between Z detection image and other detection images of front was calculated.Cause
This, average Hamming distance and maximum Hamming distance are bigger, then illustrate between Z detection image and other detection images of front
Difference it is also bigger, namely more can illustrate that the Z detection image is turned over corresponding be different of other detection images of front
Page status.And Z detection image is that shooting obtains after the completion of page turning movement, to work as the corresponding of the Z detection image
Average Hamming distance is greater than the first preset threshold and the corresponding maximum Hamming distance of the Z detection image is greater than the second default threshold
When value, then it can determine that the Z detection image is to shoot to obtain in the case where entity book to be detected is in page turning carry out state.Conversely,
It then can determine that the Z detection image is to shoot to obtain in the case where entity book to be detected is in page turning stable state.
Method provided in an embodiment of the present invention, by calculating the corresponding average Hamming distance of the last one image and most big fellow
Prescribed distance, and judge respectively the corresponding average Hamming distance of the last one image whether be greater than the first preset threshold or the last one
Whether the corresponding maximum Hamming distance of image is greater than the second preset threshold, and determines that the last one image is corresponding according to judging result
Page turning state.Due to can be according to the variation of Hash difference value between detection image, to determine the corresponding page turning shape of detection image
State, so that detection efficiency and accuracy rate can be improved.
Content based on the above embodiment, as a kind of alternative embodiment, the characteristics of image of each detection image is each
The profile line segment for including in detection image, N 2；Accordingly in any N number of detection image previous detection image and
The last one detection image, preset condition are in the profile line segment for including and the last one detection image in previous detection image
It is matched between the profile line segment for including.
Since N is 2, so that N number of detection image includes two detection images actually, respectively previous detection image and most
The latter detection image.Before the profile line segment for including in determining detection image, two-value processing first can be made to detection image.Tool
Body, image can be will test by Adaptive Thresholding and be converted to bianry image, the present invention is not especially limit this.
After it will test image and become bianry image, it may be determined that the profile line segment for including in bianry image.It needs to illustrate
It is, since detection image is to be shot and obtained to entity book to be detected, consequently, it is possible to photographing the side of entity book cover
The edge contour of edge profile or entity book content pages.It correspondingly, may include the contour line of edge contour in detection image
Section.After making two-value processing to detection image, these edge contours can be highlighted more.It is corresponding to be embodied as these edge contours
Gray value of the pixel in bianry image be mostly designated value, namely be 255, or be 0.Also need to illustrate
It is, since edge contour may be the edge contour of entity book cover, it is also possible to it is the edge contour of entity book content pages, thus
The profile line segment for including in detection image might have multiple.
Since entity book to be detected is when being in page turning stable state, the wheel that includes in the detection image that is continuously shot
Profile section is no variation in, hence for the previous detection image and the last one detection image being continuously shot,
The profile line segment that i.e. the two includes should match.On the contrary, if the profile line segment for including in previous detection image and last
It is matched between the profile line segment for including in one detection image, then can determine that the last one detection image is in entity book to be detected
Shooting obtains under page turning stable state.Based on the above principles, if the profile line segment for including in previous detection image with
It is mismatched between the profile line segment for including in the last one detection image, then can determine that the last one detection image is to be detected
Entity book is in what shooting under page turning carry out state obtained.
Method provided in an embodiment of the present invention passes through the profile line segment for judging to include in previous detection image and last
Whether matched between the profile line segment for including in a detection image, and determines that the last one image is corresponding according to judging result and turn over
Page status.Due to that can be turned over according to the matching status for the profile line segment for including between detection image with determining that detection image is corresponding
Page status, so that detection efficiency and accuracy rate can be improved.
Content based on the above embodiment, as a kind of alternative embodiment, determine the last one detection image be to
Detection entity book be under page turning stable state shoot obtain before, may further determine that the profile line segment for including in detection image.
The embodiment of the present invention does not limit the mode for the profile line segment for including in determining detection image specifically, including but not limited to: right
Any detection image in any N number of detection image, determines the contour pixel point set in any detection image, to each round
Contour pixel point in wide pixel collection is fitted, and obtains the profile line segment for including in any detection image.
By the content of above-described embodiment it is found that the profile line segment for including in detection image might have it is multiple.Therefore, for
Any detection image in any N number of detection image can first determine the corresponding wire-frame image of each contour line section in the detection image
Vegetarian refreshments set.By being fitted to the contour pixel point in each profile pixel collection, obtain include in the detection image
Profile line segment.Wherein, wire-frame image vegetarian refreshments is after making two-value processing to the detection image, and gray value is the pixel of designated value.
The definition of designated value can refer to the content of above-described embodiment, and details are not described herein again.In addition, the mode of fitting can be linear fit
Or approximately linear fitting, the present invention is not especially limit this.
Content based on the above embodiment, as a kind of alternative embodiment, the embodiment of the present invention is not to any detection of determination
The mode of contour pixel point set in image specifically limits, including but not limited to: to the pixel in any detection image
It is retrieved line by line, for any wire-frame image vegetarian refreshments retrieved, if the total quantity of current outline pixel collection is 0 or does not deposit
In the contour pixel point set for meeting the first specified requirements, and exists and meet the second specified requirements between any wire-frame image vegetarian refreshments
Specified wire-frame image vegetarian refreshments, then establish new contour pixel point set, and by any wire-frame image vegetarian refreshments and specified contour pixel
Point is added to new contour pixel point set, if the total quantity of current outline pixel collection is 0 or refers to there is no meeting first
The contour pixel point set of fixed condition, and there is no the specified profiles for meeting the second specified requirements between any wire-frame image vegetarian refreshments
Any wire-frame image vegetarian refreshments is then added to wire-frame image vegetarian refreshments and eliminates set, if there is currently meet the first specified requirements by pixel
Contour pixel point set, then by any wire-frame image vegetarian refreshments be added to meet the first specified requirements contour pixel point set, weight
Above-mentioned retrieval and adding procedure are executed again, until having retrieved all pixels point in any detection image；Wherein, profile is specified
Pixel is that next wire-frame image vegetarian refreshments that any wire-frame image vegetarian refreshments retrieves below or wire-frame image vegetarian refreshments are eliminated in set
Wire-frame image vegetarian refreshments.
Content based on the above embodiment, as a kind of alternative embodiment, the first specified requirements is contour pixel point set
Inside have that contour pixel is selected in the specified range that any contour pixel selects surrounding or there are wire-frame image vegetarian refreshments and any profile
The distance between pixel is less than third predetermined threshold value；Second specified requirements is specified wire-frame image vegetarian refreshments wire-frame image vegetarian refreshments in office
In the specified range of surrounding or the distance between any wire-frame image vegetarian refreshments and specified wire-frame image vegetarian refreshments are less than third predetermined threshold value.
Wherein, for wire-frame image vegetarian refreshments A, the specified range around wire-frame image vegetarian refreshments A is referred to: being with wire-frame image vegetarian refreshments A
Make a circle in center week pixel.For example, being the then finger around the wire-frame image vegetarian refreshments for (5,5) with the coordinate of wire-frame image vegetarian refreshments A
Determine range be coordinate be (4,6), (5,6), (6,6), (4,5), (6,5), (4,4), (5,4) and (6,4) this 8 pixels.
If the coordinate of wire-frame image vegetarian refreshments B is any one coordinate in above-mentioned 8 coordinates, it can determine wire-frame image vegetarian refreshments B in wire-frame image
In specified range around vegetarian refreshments A.It should be noted that above-described embodiment be related to the first preset threshold, the second default threshold
The value of value and third predetermined threshold value can be configured according to the actual situation, and the present invention is not especially limit this.
In order to make it easy to understand, including 5*5 picture with the detection image for any detection image in N number of detection image
For vegetarian refreshments, the process for determining the contour pixel point set in the detection image is illustrated.It is somebody's turn to do it should be noted that determining
The contour pixel point for constituting differently contoured line segment as polymerize by the process of the contour pixel point set in detection image
Process.Wherein, the mode retrieved line by line to the pixel in the detection image can be opened from first, upper left corner pixel
Begin, from left to right retrieval downwards line by line.Alternatively, can also be since the pixel of first, the lower left corner, from left to right line by line upwards
Retrieval.It is, of course, also possible to which the embodiment of the present invention is to this using other retrieval modes line by line or different retrieval modes by column
It is not especially limited.
Retrieved for retrieval downwards for first line by line since the pixel of first, the upper left corner from left to right
Wire-frame image vegetarian refreshments (the wire-frame image vegetarian refreshments is not necessarily first pixel, and the definition of wire-frame image vegetarian refreshments can refer to above-mentioned reality
Apply the content of example), due to being 0 there is currently no the total quantity of contour pixel point set namely current outline pixel collection, from
And it can determine current with the presence or absence of the specified wire-frame image vegetarian refreshments for meeting the second specified requirements between any wire-frame image vegetarian refreshments.
Due to specify wire-frame image vegetarian refreshments be first wire-frame image vegetarian refreshments retrieve below next wire-frame image vegetarian refreshments (namely
Second wire-frame image vegetarian refreshments) or wire-frame image vegetarian refreshments eliminate set in wire-frame image vegetarian refreshments, and current outline pixel eliminate collection
It is combined into sky, so that specified wire-frame image vegetarian refreshments is second wire-frame image vegetarian refreshments.Then, can determine whether second wire-frame image vegetarian refreshments whether
In specified range around first wire-frame image vegetarian refreshments or between second wire-frame image vegetarian refreshments and first wire-frame image vegetarian refreshments
Whether distance, which is less than third predetermined threshold value, can determine second wire-frame image vegetarian refreshments and first if meeting one of condition
Meet the second specified requirements between wire-frame image vegetarian refreshments, so as to create new contour pixel point set, and by first wire-frame image
Vegetarian refreshments and second wire-frame image vegetarian refreshments are added to new contour pixel point set.
If two conditions be not satisfied namely second wire-frame image vegetarian refreshments and first wire-frame image vegetarian refreshments between be unsatisfactory for
Two specified requirements, and current outline pixel eliminate collection be combined into sky, thus can determine there is currently no with first contour pixel
Meet the specified wire-frame image vegetarian refreshments of the second specified requirements between selecting, so as to which first wire-frame image vegetarian refreshments is added to contour pixel
Point eliminates set.
After the above process, it can continue to retrieve subsequent pixel.For the first situation in the above process,
First wire-frame image vegetarian refreshments and second wire-frame image vegetarian refreshments are added to new contour pixel point set.At this point, due to second
A wire-frame image vegetarian refreshments has been retrieved and has been added into the contour pixel point set of foundation, thus in the first scenario,
Next wire-frame image vegetarian refreshments being retrieved is third wire-frame image vegetarian refreshments.Since the total quantity of current outline pixel collection is
1 without being 0, to can determine whether the contour pixel point set that currently whether there is the first specified requirements of satisfaction, namely before judgement
Whether the contour pixel point set of creation meets the first specified requirements.
Specifically, it can determine whether in the contour pixel point set with the presence or absence of wire-frame image vegetarian refreshments in third wire-frame image vegetarian refreshments week
In the specified range enclosed or there are the distance between wire-frame image vegetarian refreshments and third wire-frame image vegetarian refreshments to be less than third predetermined threshold value.
If meeting one of condition, that is, it can determine that the contour pixel point set meets the first specified requirements, so as to take turns third
Wide pixel is added to the contour pixel point set.
However, can determine whether to exist and be taken turns with third if the contour pixel point set is unsatisfactory for the first specified requirements
Meet the specified wire-frame image vegetarian refreshments of the second specified requirements between wide pixel.Since specified wire-frame image vegetarian refreshments is third wire-frame image
The next wire-frame image vegetarian refreshments (namely the 4th wire-frame image vegetarian refreshments) or wire-frame image vegetarian refreshments that vegetarian refreshments retrieves below eliminate set
In wire-frame image vegetarian refreshments, and current outline pixel eliminates set or for sky, so that specified wire-frame image vegetarian refreshments is the 4th wheel
Wide pixel.Then, can determine whether the 4th wire-frame image vegetarian refreshments whether in the specified range around third wire-frame image vegetarian refreshments or
Whether the distance between the 4th wire-frame image vegetarian refreshments of person and third wire-frame image vegetarian refreshments are less than third predetermined threshold value, if meeting wherein
One condition can determine and meets the second specified requirements between the 4th wire-frame image vegetarian refreshments and third wire-frame image vegetarian refreshments, thus
New contour pixel point set can be created, and the 4th wire-frame image vegetarian refreshments and third wire-frame image vegetarian refreshments are added to new profile
Pixel collection.
If two conditions be not satisfied namely the 4th wire-frame image vegetarian refreshments and third wire-frame image vegetarian refreshments between be unsatisfactory for
Two specified requirements, and current outline pixel eliminate collection be combined into sky, thus can determine there is currently no with third contour pixel
Meet the specified wire-frame image vegetarian refreshments of the second specified requirements between selecting, so as to which third wire-frame image vegetarian refreshments is added to contour pixel
Point eliminates set.
For the second situation in the above process, first wire-frame image vegetarian refreshments is also added to wire-frame image vegetarian refreshments and is eliminated
Set.At this point, since second wire-frame image vegetarian refreshments does not find ownership also, thus in the latter case, it is next to be retrieved
Wire-frame image vegetarian refreshments be still second wire-frame image vegetarian refreshments.Since the total quantity of current outline pixel collection is still 0, thus
It can determine current with the presence or absence of the specified wire-frame image vegetarian refreshments for meeting the second specified requirements between second wire-frame image vegetarian refreshments.
Due to specify wire-frame image vegetarian refreshments be second wire-frame image vegetarian refreshments retrieve below next wire-frame image vegetarian refreshments (namely
Third wire-frame image vegetarian refreshments) or wire-frame image vegetarian refreshments eliminate set in wire-frame image vegetarian refreshments, and wire-frame image vegetarian refreshments eliminate set in
Only one wire-frame image vegetarian refreshments namely first wire-frame image vegetarian refreshments, and first wire-frame image vegetarian refreshments and second wire-frame image vegetarian refreshments
Before it is determined that mistake, is unsatisfactory for the second specified requirements between the two, so that specified wire-frame image vegetarian refreshments is third wheel
Wide pixel.Then, can determine whether third wire-frame image vegetarian refreshments whether in the specified range around second wire-frame image vegetarian refreshments or
Whether the distance between person's third wire-frame image vegetarian refreshments and second wire-frame image vegetarian refreshments are less than third predetermined threshold value, if meeting wherein
One condition can determine and meets the second specified requirements between third wire-frame image vegetarian refreshments and second wire-frame image vegetarian refreshments, thus
New contour pixel point set can be created, and second wire-frame image vegetarian refreshments and third wire-frame image vegetarian refreshments are added to new profile
Pixel collection.
If two conditions be not satisfied namely third wire-frame image vegetarian refreshments and second wire-frame image vegetarian refreshments between be unsatisfactory for
Two specified requirements, and between the first wire-frame image vegetarian refreshments and second wire-frame image vegetarian refreshments in the superseded set of wire-frame image vegetarian refreshments also not
Meet the second specified requirements, to can determine that there is currently no meet the second specified requirements between second wire-frame image vegetarian refreshments
Specified wire-frame image vegetarian refreshments eliminates set so as to which second wire-frame image vegetarian refreshments is also added to wire-frame image vegetarian refreshments.At this point, wire-frame image
Vegetarian refreshments eliminates set, and there are two wire-frame image vegetarian refreshments, respectively first wire-frame image vegetarian refreshments and second wire-frame image vegetarian refreshments.
Subsequent process can and so on, until all pixels point in the detection image has been retrieved, so that the inspection can be obtained
Contour pixel point set in altimetric image.It should be noted that during retrieval is with addition, if being eliminated from wire-frame image vegetarian refreshments
A wire-frame image vegetarian refreshments is had selected in set and is added in a certain contour pixel point set, then eliminates set in wire-frame image vegetarian refreshments
In need to delete the wire-frame image vegetarian refreshments.The all pixels point in the detection image is being retrieved, can be obtained in the detection image
Contour pixel point set.By the above process, the contour pixel point set in N number of detection image in each detection image can be obtained
It closes.
Method provided in an embodiment of the present invention constitutes different wheels due to that can will test in such a way that iteration judges in image
The contour pixel point of profile section is polymerize, to obtain the contour pixel point set in detection image, to be based on contour pixel
When point set is fitted profile line segment, the accuracy rate of fitting can be improved.
Content based on the above embodiment, as a kind of alternative embodiment, determine the last one detection image be to
Detection entity book be under page turning stable state shoot obtain before, also can determine whether the profile for including in previous detection image
Whether matched between the profile line segment for including in line segment and the last one detection image.Referring to Fig. 3, the embodiment of the present invention is not to sentencing
Disconnected whether matched mode specifically limits between the two, including but not limited to:
If 301, the contour line segment number for including in previous detection image and the wheel for including in the last one detection image
Profile segment number is identical, then judges the profile line segment for including in previous detection image and include in the last one detection image
Whether matching is corresponded between profile line segment.
Due to the profile for including in the contour line segment number that includes in previous detection image and the last one detection image
Line segment quantity is identical, to can be in advance the profile line segment for including and last in previous detection image before executing 301
Corresponding relationship is established between the profile line segment for including in a detection image, such as according to the contour line for including in previous detection image
The position for the profile line segment for including in the position of section and the last one detection image, establishes corresponding relationship between the two, this
Inventive embodiments are not especially limited this.
In addition, the profile line segment for including in judging previous detection image and the wheel for including in the last one detection image
When whether corresponding matching between profile section, whether the shape that matching also refers to two profile line segments is approximate, Huo Zhe
Whether position is similar under same coordinate system, and the present invention is not especially limit this.
It should be noted that if in the contour line segment number for including in previous detection image and the last one detection image
The contour line segment number for including is not identical, then can determine the profile line segment for including in previous detection image and the last one detection
It is mismatched between the profile line segment for including in image.By the content of above-described embodiment it is found that can determine the last one detection in turn
Image is to shoot to obtain in the case where entity book to be detected is in page turning carry out state.
If 302, the profile line segment for including in previous detection image and the contour line for including in the last one detection image
Matching is corresponded between section, it is determined that the profile line segment for including in previous detection image and wrap in the last one detection image
It is matched between the profile line segment contained.
It should be noted that if the profile line segment for including in previous detection image and in the last one detection image including
Profile line segment between not correspond matching, then can determine the profile line segment for including in previous detection image and last
It is mismatched between the profile line segment for including in a detection image.Accordingly, it can be determined that the last one detection image is in reality to be detected
Body book is in what shooting under page turning carry out state obtained.
Method provided in an embodiment of the present invention, since the matching degree of profile line segment between two detection images is able to reflect
To judge the page turning state of detection image with this Detection accuracy can be improved in consistent degree between two detection images out.
Content based on the above embodiment, as a kind of alternative embodiment, the embodiment of the present invention is not to judging previous inspection
Whether correspond and matches between the profile line segment for including in altimetric image and the profile line segment for including in the last one detection image
Mode specifically limit.Referring to fig. 4, including but not limited to:
3011, according to the size order of the starting point abscissa of profile line segment, to the wheel for including in previous detection image
Profile section is ranked up number, and is ranked up number, previous inspection to the profile line segment for including in the last one detection image
Between altimetric image and the last one detection image coordinate system to establish mode identical.
It should be noted that the mode of sequencing numbers can also mode from small to large.In addition, the foundation of sequencing numbers is also
It can be the end point of starting point ordinate perhaps the end point abscissa of profile line segment or profile line segment of contour line section
Ordinate, the present invention is not especially limit this.
3012, for numbering identical any two contour line between previous detection image and the last one detection image
Section, if between any two profile line segment in the angle on same coordinate axis in the first preset range and on same coordinate axis
Intercept difference in the second preset range, it is determined that any two profile line segment is mutually matched.
For example, if there is 5 profile line segments in previous detection image, be followed successively by respectively after sequencing numbers A1, A2,
A3, A4 and A5.And also have 5 profile line segments in the last one detection image, be followed successively by respectively after sequencing numbers B1,
B2, B3, B4 and B5.Wherein, A1 is identical as B1 number, subsequent and so on.Then, it can determine that between A1 and B1 in identical seat
Whether whether the intercept difference in the first preset range and on same coordinate axis is in the second preset range for angle on parameter.
If meeting two conditions, that is, it can determine that A1 is mutually matched with B1.Similarly, A2 and B2, subsequent and so on can be subsequently determined that.
It should be noted that the first preset range can be to be positive and negative between 5 degree, the second preset range can be positive and negative 15
The distance of pixel, the present invention is not especially limit this.
If 3013, numbering identical profile line segment between previous detection image and the last one detection image mutually
Match, it is determined that between the profile line segment for including in previous detection image and the profile line segment for including in the last one detection image
Correspond matching.
For example, in conjunction with above-mentioned example, if A1 is mutually matched with B2, A3 with B3, A4 with B4 and A5 with B5 with B1, A2,
It then can determine between the profile line segment for including in previous detection image and the profile line segment for including in the last one detection image
Correspond matching.
It is mismatched it should be noted that if there is the identical profile line segment of a pair of of number, then can determine true previous inspection
It not corresponds and matches between the profile line segment for including in altimetric image and the profile line segment for including in the last one detection image,
And then can determine include in the profile line segment for including in previous detection image and the last one detection image profile line segment it
Between mismatch.
It should be noted that above-mentioned all alternative embodiments, can form optional implementation of the invention using any combination
Example, this is no longer going to repeat them.
Content based on the above embodiment, the embodiment of the invention provides a kind of page turning detection device, the device is for holding
Page turning detection method in row above method embodiment.Referring to Fig. 5, which includes:
Extraction module 501, for mentioning during obtaining multiple detection images to detect to entity book to be detected
Take the characteristics of image of each detection image in multiple detection images；
First determining module 502 is used for any N number of detection image continuous for acquisition sequence in multiple detection images,
If meeting between the characteristics of image of the last one detection image and the characteristics of image of other detection images in any N number of detection image
Preset condition, it is determined that the last one detection image is to shoot to obtain in the case where entity book to be detected is in page turning stable state；
Wherein, N is the integer greater than 1, after page turning stable state refers to the page turning movement completed to entity book to be detected, entity to be detected
Stable state locating for book.
As a kind of alternative embodiment, characteristics of image is Hash difference value, and N is the integer greater than 2, and preset condition is last
The corresponding average Hamming distance of one image maximum Hamming distance corresponding no more than the first preset threshold or the last one image
No more than the second preset threshold；Correspondingly, the device further include:
First computing module, for calculate separately the Hash difference value of the last one image with it is each in other detection images
Difference between the Hash difference value of detection image；
Second computing module, for calculating the average value of all differences, using average value as average Hamming distance, and from institute
Have and choose maximum value in difference, using maximum value as maximum Hamming distance.
As a kind of alternative embodiment, the characteristics of image of each detection image is the contour line for including in each detection image
Section, N 2；Accordingly for the previous detection image and the last one detection image in any N number of detection image, item is preset
Part is to match between the profile line segment for including and the profile line segment for including in the last one detection image in previous detection image.
As a kind of alternative embodiment, the device further include:
Second determining module, for determining any detection image for any detection image in any N number of detection image
Interior contour pixel point set；
Fitting module obtains any detection for being fitted to the contour pixel point in each profile pixel collection
The profile line segment for including in image；Wherein, wire-frame image vegetarian refreshments is after making two-value processing to any detection image, and gray value is specified
The pixel of value.
As a kind of alternative embodiment, fitting module, for being retrieved line by line to the pixel in any detection image,
For any wire-frame image vegetarian refreshments retrieved, if the total quantity of current outline pixel collection is 0 or refers to there is no meeting first
The contour pixel point set of fixed condition, and exist and the specified wire-frame image that meets the second specified requirements between any wire-frame image vegetarian refreshments
Vegetarian refreshments then establishes new contour pixel point set, and any wire-frame image vegetarian refreshments and specified wire-frame image vegetarian refreshments is added to new
Contour pixel point set, if the total quantity of current outline pixel collection is 0 or there is no the profiles for meeting the first specified requirements
Pixel collection, and there is no the specified wire-frame image vegetarian refreshments for meeting the second specified requirements between any wire-frame image vegetarian refreshments, then will
Any wire-frame image vegetarian refreshments is added to wire-frame image vegetarian refreshments and eliminates set, if there is currently the wire-frame image vegetarian refreshments for meeting the first specified requirements
Any wire-frame image vegetarian refreshments is then added to the contour pixel point set for meeting the first specified requirements, repeats above-mentioned inspection by set
Rope and adding procedure, until having retrieved all pixels point in any detection image；Wherein, it is any for specifying wire-frame image vegetarian refreshments
The next wire-frame image vegetarian refreshments or wire-frame image vegetarian refreshments that wire-frame image vegetarian refreshments retrieves below eliminate the wire-frame image vegetarian refreshments in set.
As a kind of alternative embodiment, the first specified requirements is that there are contour pixel points any in contour pixel point set
In specified range around wire-frame image vegetarian refreshments or there are the distance between wire-frame image vegetarian refreshments and any wire-frame image vegetarian refreshments less than the
Three preset thresholds；Second specified requirements is in the specified range specified around wire-frame image vegetarian refreshments wire-frame image vegetarian refreshments in office or to appoint
The distance between one wire-frame image vegetarian refreshments and specified wire-frame image vegetarian refreshments are less than third predetermined threshold value.
As a kind of alternative embodiment, the device further include:
Judgment module, for when in the contour line segment number and the last one detection image for including in previous detection image
When the contour line segment number for including is identical, then the profile line segment for including in previous detection image and the last one detection figure are judged
Matching whether is corresponded between the profile line segment for including as in；
Third determining module, for when in the profile line segment and the last one detection image for including in previous detection image
Between the profile line segment for including correspond matching when, it is determined that the profile line segment for including in previous detection image and last
It is matched between the profile line segment for including in a detection image.
As a kind of alternative embodiment, judgment module, for the size order of the starting point abscissa according to profile line segment,
Number is ranked up to the profile line segment for including in previous detection image, and to the profile for including in the last one detection image
Line segment is ranked up number, between previous detection image and the last one detection image coordinate system to establish mode identical；It is right
Identical any two profile line segment is numbered between previous detection image and the last one detection image, if any two wheel
Intercept difference between profile section in the angle on same coordinate axis in the first preset range and on same coordinate axis is
In two preset ranges, it is determined that any two profile line segment is mutually matched；If previous detection image and the last one detection are schemed
It numbers identical profile line segment as between and mutually matches, it is determined that the profile line segment that includes in previous detection image and last
Matching is corresponded between the profile line segment for including in one detection image.
Device provided in an embodiment of the present invention, by obtaining multiple detection images to detect to entity book to be detected
During, extract the characteristics of image of each detection image in multiple detection images.For acquisition sequence in multiple detection images
Continuous any N number of detection image, if the characteristics of image of the last one detection image and other detections in any N number of detection image
Meet preset condition between the characteristics of image of image, it is determined that the last one detection image is to be in page turning in entity book to be detected
Shooting obtains under stable state.It, only need to be according to the inspection for any detection image due to requiring no matched process one by one
Several detection images before altimetric image can determine the corresponding page turning state of the detection image, to improve detection efficiency.
Secondly, by calculating the corresponding average Hamming distance of the last one image and maximum Hamming distance, and judge respectively
Whether the corresponding average Hamming distance of the last one image is greater than the first preset threshold or the corresponding most big fellow of the last one image
Whether prescribed distance is greater than the second preset threshold, and determines the corresponding page turning state of the last one image according to judging result.Due to
It can be according to the variation of Hash difference value between detection image, to determine the corresponding page turning state of detection image, so that inspection can be improved
Survey efficiency and accuracy rate.
In addition, by judging the profile line segment for including in previous detection image and including in the last one detection image
Whether matched between profile line segment, and determines the corresponding page turning state of the last one image according to judging result.Due to can basis
The matching status for the profile line segment for including between detection image, to determine the corresponding page turning state of detection image, to can be improved
Detection efficiency and accuracy rate.
Finally, due to which the wire-frame image vegetarian refreshments for constituting differently contoured line segment in image can be will test in such a way that iteration judges
Polymerize, to obtain the contour pixel point set in detection image, thus based on contour pixel point set to profile line segment into
When row fitting, the accuracy rate of fitting can be improved.
The embodiment of the invention provides a kind of electronic equipment.Referring to Fig. 6, which includes: processor (processor)
601, memory (memory) 602 and bus 603；
Wherein, processor 601 and memory 602 complete mutual communication by bus 603 respectively；
Processor 601 is used to call the program instruction in memory 602, to execute the inspection of page turning provided by above-described embodiment
Survey method, for example, during obtaining multiple detection images to detect to entity book to be detected, extract multiple inspections
The characteristics of image of each detection image in altimetric image；Any N number of detection figure continuous for acquisition sequence in multiple detection images
Picture, if in any N number of detection image between the characteristics of image of the last one detection image and the characteristics of image of other detection images
Meet preset condition, it is determined that the last one detection image is to shoot to obtain in the case where entity book to be detected is in page turning stable state
's；Wherein, N is the integer greater than 1, to be detected after page turning stable state refers to the page turning movement completed to entity book to be detected
Stable state locating for entity book.
The embodiment of the present invention provides a kind of non-transient computer readable storage medium, the non-transient computer readable storage medium
Matter stores computer instruction, which makes computer execute page turning detection method provided by above-described embodiment, such as
It include: to be extracted every in multiple detection images during obtaining multiple detection images to detect to entity book to be detected
The characteristics of image of one detection image；Any N number of detection image continuous for acquisition sequence in multiple detection images, if any N
Meet default item in a detection image between the characteristics of image of the last one detection image and the characteristics of image of other detection images
Part, it is determined that the last one detection image is to shoot to obtain in the case where entity book to be detected is in page turning stable state；Wherein, N
For the integer greater than 1, after page turning stable state refers to the page turning movement completed to entity book to be detected, locating for entity book to be detected
Stable state.
Those of ordinary skill in the art will appreciate that: realize that all or part of the steps of above method embodiment can pass through
The relevant hardware of program instruction is completed, and program above-mentioned can be stored in a computer readable storage medium, the program
When being executed, step including the steps of the foregoing method embodiments is executed；And storage medium above-mentioned includes: ROM, RAM, magnetic disk or light
The various media that can store program code such as disk.
The embodiments such as electronic equipment described above are only schematical, wherein unit as illustrated by the separation member
It may or may not be physically separated, component shown as a unit may or may not be physics list
Member, it can it is in one place, or may be distributed over multiple network units.It can be selected according to the actual needs
In some or all of the modules achieve the purpose of the solution of this embodiment.Those of ordinary skill in the art are not paying creativeness
Labour in the case where, it can understand and implement.
Through the above description of the embodiments, those skilled in the art can be understood that each embodiment can
It realizes by means of software and necessary general hardware platform, naturally it is also possible to pass through hardware.Based on this understanding, on
Stating technical solution, substantially the part that contributes to existing technology can be embodied in the form of software products in other words, should
Computer software product may be stored in a computer readable storage medium, such as ROM/RAM, magnetic disk, CD, including several fingers
It enables and using so that a computer equipment (can be personal computer, server or the network equipment etc.) executes each implementation
Certain Part Methods of example or embodiment.Finally, the present processes are only preferable embodiment, it is not intended to limit this
The protection scope of inventive embodiments.All spirit in the embodiment of the present invention within principle, equally replace by made any modification
It changes, improve, should be included within the protection scope of the embodiment of the present invention.
Claims (11)
1. a kind of page turning detection method characterized by comprising
During obtaining multiple detection images to detect to entity book to be detected, extract in the multiple detection image
The characteristics of image of each detection image；
Any N number of detection image continuous for acquisition sequence in the multiple detection image, if any N number of detection image
In meet preset condition between the characteristics of image of the last one detection image and the characteristics of image of other detection images, it is determined that institute
Stating the last one detection image is to shoot to obtain in the case where the entity book to be detected is in page turning stable state；Wherein, the N
It is described to be checked after the page turning stable state refers to the page turning movement completed to the entity book to be detected for the integer greater than 1
Survey stable state locating for entity book.
2. the method according to claim 1, wherein described image feature be Hash difference value, the N be greater than
2 integer, the preset condition be the corresponding average Hamming distance of the last one described image no more than the first preset threshold or
The corresponding maximum Hamming distance of the last one described image is not more than the second preset threshold；Correspondingly, the determination is described last
One detection image be in the case where the entity book to be detected is in page turning stable state shooting obtain before, further includes:
Calculate separately the Kazakhstan of each detection image in the Hash difference value and other detection images of the last one image
Difference between uncommon difference value；
The average value for calculating all differences using the average value as the average Hamming distance, and is chosen from all differences
Maximum value, using the maximum value as the maximum Hamming distance.
3. the method according to claim 1, wherein the characteristics of image of each detection image is each detection image
In include profile line segment, the N be 2；Accordingly in any N number of detection image previous detection image and
The last one described detection image, the preset condition be the previous detection image in include profile line segment and it is described most
It is matched between the profile line segment for including in the latter detection image.
4. according to the method described in claim 3, it is characterized in that, the last one detection image described in the determination is described
Entity book to be detected is under page turning stable state before shooting obtains, further includes:
For any detection image in any N number of detection image, the contour pixel in any detection image is determined
Point set is fitted the contour pixel point in each profile pixel collection, obtain include in any detection image
Profile line segment；Wherein, the wire-frame image vegetarian refreshments is after making two-value processing to any detection image, and gray value is designated value
Pixel.
5. according to the method described in claim 4, it is characterized in that, contour pixel in the determination any detection image
Point set, comprising:
Pixel in any detection image is retrieved line by line, for any wire-frame image vegetarian refreshments retrieved, if working as
The total quantity of front profile pixel collection be 0 or there is no meet the first specified requirements contour pixel point set, and exist with
The specified wire-frame image vegetarian refreshments for meeting the second specified requirements between any wire-frame image vegetarian refreshments, then establish new contour pixel point set
It closes, and any wire-frame image vegetarian refreshments and the specified wire-frame image vegetarian refreshments is added to the new contour pixel point set,
If the total quantity of current outline pixel collection is 0 or there is no the contour pixel point set for meeting first specified requirements,
And there is no the specified wire-frame image vegetarian refreshments for meeting second specified requirements between any wire-frame image vegetarian refreshments, then it will be described
Any wire-frame image vegetarian refreshments is added to wire-frame image vegetarian refreshments and eliminates set, if there is currently the wire-frame images for meeting first specified requirements
Any wire-frame image vegetarian refreshments is then added to the contour pixel point set for meeting first specified requirements by vegetarian refreshments set, weight
Above-mentioned retrieval and adding procedure are executed again, until having retrieved all pixels point in any detection image；
Wherein, the specified wire-frame image vegetarian refreshments be next wire-frame image vegetarian refreshments for retrieving below of any wire-frame image vegetarian refreshments or
Wire-frame image vegetarian refreshments described in person eliminates the wire-frame image vegetarian refreshments in set.
6. according to the method described in claim 5, it is characterized in that, first specified requirements is contour pixel point set memory
Wire-frame image vegetarian refreshments in the specified range around any wire-frame image vegetarian refreshments or there are wire-frame image vegetarian refreshments with it is described any
The distance between wire-frame image vegetarian refreshments is less than the third predetermined threshold value；Second specified requirements is the specified wire-frame image vegetarian refreshments
In the specified range around any wire-frame image vegetarian refreshments or any wire-frame image vegetarian refreshments and the specified contour pixel
The distance between point is less than third predetermined threshold value.
7. method according to any one of claim 3 to 6, which is characterized in that the last one is detected described in the determination
Image be in the case where the entity book to be detected is in page turning stable state shooting obtain before, further includes:
If the contour line segment number for including in the previous detection image and the wheel for including in the last one described detection image
Profile segment number is identical, then judges the profile line segment for including in the previous detection image and the last one described detection image
In include profile line segment between whether correspond matching；
If the profile line segment for including in the previous detection image and the contour line for including in the last one described detection image
Matching is corresponded between section, it is determined that the profile line segment for including in the previous detection image and the last one described detection
It is matched between the profile line segment for including in image.
8. the method according to the description of claim 7 is characterized in that the wheel for including in the judgement previous detection image
Whether correspond and match between the profile line segment for including in profile section and the last one described detection image, comprising:
According to the size order of the starting point abscissa of profile line segment, to the profile line segment for including in the previous detection image
It is ranked up number, and number is ranked up to the profile line segment for including in the last one described detection image, it is described previous
Between detection image and the last one described detection image coordinate system to establish mode identical；
For numbering identical any two contour line between the previous detection image and the last one described detection image
Section, if between any two profile line segment in the angle on same coordinate axis in the first preset range and in same coordinate
Intercept difference on axis is in the second preset range, it is determined that any two profile line segment is mutually matched；
If numbering identical profile line segment between the previous detection image and the last one described detection image mutually
Match, it is determined that the profile line segment for including in the previous detection image and the profile for including in the last one described detection image
Matching is corresponded between line segment.
9. a kind of page turning detection device characterized by comprising
Extraction module, for during obtaining multiple detection images to detect to entity book to be detected, described in extraction
The characteristics of image of each detection image in multiple detection images；
First determining module, for any N number of detection image continuous for acquisition sequence in the multiple detection image, if institute
It states and meets between the characteristics of image of the last one detection image and the characteristics of image of other detection images in any N number of detection image
Preset condition, it is determined that the last one described detection image is shot in the case where the entity book to be detected is in page turning stable state
It obtains；Wherein, the N is the integer greater than 1, and the page turning stable state refers to that the entity book to be detected is turned in completion
After page movement, stable state locating for the entity book to be detected.
10. a kind of electronic equipment characterized by comprising
At least one processor；And
At least one processor being connect with the processor communication, in which:
The memory is stored with the program instruction that can be executed by the processor, and the processor calls described program to instruct energy
Enough execute method as described in any of the claims 1 to 8.
11. a kind of non-transient computer readable storage medium, which is characterized in that the non-transient computer readable storage medium is deposited
Computer instruction is stored up, the computer instruction makes the computer execute method as described in any of the claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810813117.XA CN109145751A (en) | 2018-07-23 | 2018-07-23 | Page turning detection method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810813117.XA CN109145751A (en) | 2018-07-23 | 2018-07-23 | Page turning detection method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109145751A true CN109145751A (en) | 2019-01-04 |
Family
ID=64801490
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810813117.XA Pending CN109145751A (en) | 2018-07-23 | 2018-07-23 | Page turning detection method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109145751A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110443224A (en) * | 2019-08-14 | 2019-11-12 | 广东小天才科技有限公司 | Page turning detection method, device, electronic equipment and storage medium |
| CN110516634A (en) * | 2019-08-30 | 2019-11-29 | 安徽淘云科技有限公司 | A kind of page turning detection method, device, equipment and storage medium |
| CN111639713A (en) * | 2020-06-01 | 2020-09-08 | 广东小天才科技有限公司 | Page turning detection method and device, electronic equipment and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012167110A2 (en) * | 2011-06-02 | 2012-12-06 | Honda Motor Co., Ltd. | Target recognition and localization methods using a laser sensor for wheeled mobile robots |
| CN104574403A (en) * | 2015-01-12 | 2015-04-29 | 飞天诚信科技股份有限公司 | Intelligent cutting method |
| CN107734254A (en) * | 2017-10-14 | 2018-02-23 | 上海瞬动科技有限公司合肥分公司 | A kind of unmanned plane is selected a good opportunity photographic method automatically |
| CN108229406A (en) * | 2018-01-11 | 2018-06-29 | 海信集团有限公司 | A kind of method for detecting lane lines, device and terminal |
-
2018
- 2018-07-23 CN CN201810813117.XA patent/CN109145751A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012167110A2 (en) * | 2011-06-02 | 2012-12-06 | Honda Motor Co., Ltd. | Target recognition and localization methods using a laser sensor for wheeled mobile robots |
| CN104574403A (en) * | 2015-01-12 | 2015-04-29 | 飞天诚信科技股份有限公司 | Intelligent cutting method |
| CN107734254A (en) * | 2017-10-14 | 2018-02-23 | 上海瞬动科技有限公司合肥分公司 | A kind of unmanned plane is selected a good opportunity photographic method automatically |
| CN108229406A (en) * | 2018-01-11 | 2018-06-29 | 海信集团有限公司 | A kind of method for detecting lane lines, device and terminal |
Non-Patent Citations (2)
| Title |
|---|
| 刘凯伦 等: "基于图像感知哈希的运动目标跟踪", 《电脑知识与技术》 * |
| 童礼华 等: "建筑轮廓的车载和航空LiDAR数据配准", 《测绘学报》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110443224A (en) * | 2019-08-14 | 2019-11-12 | 广东小天才科技有限公司 | Page turning detection method, device, electronic equipment and storage medium |
| CN110516634A (en) * | 2019-08-30 | 2019-11-29 | 安徽淘云科技有限公司 | A kind of page turning detection method, device, equipment and storage medium |
| CN111639713A (en) * | 2020-06-01 | 2020-09-08 | 广东小天才科技有限公司 | Page turning detection method and device, electronic equipment and storage medium |
| CN111639713B (en) * | 2020-06-01 | 2023-09-08 | 广东小天才科技有限公司 | Page turning detection method and device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108665508B (en) | Instant positioning and map construction method, device and storage medium | |
| CN104850858B (en) | A kind of injection-molded item defects detection recognition methods | |
| CN107953329B (en) | Object recognition and attitude estimation method and device and mechanical arm grabbing system | |
| CN109145751A (en) | Page turning detection method and device | |
| CN110473254A (en) | A kind of position and orientation estimation method and device based on deep neural network | |
| WO2022033076A1 (en) | Target detection method and apparatus, device, storage medium, and program product | |
| CN111126412B (en) | Image key point detection method based on characteristic pyramid network | |
| CN109033989B (en) | Target identification method and device based on three-dimensional point cloud and storage medium | |
| CN107423306B (en) | Image retrieval method and device | |
| CN105453132B (en) | The information processing equipment and image processing method of real-time image processing | |
| CN112132279B (en) | Convolutional neural network model compression method, device, equipment and storage medium | |
| CN103744974B (en) | Method and device for selecting local interest points | |
| CN103578093A (en) | Image registration method and device and augmented reality system | |
| CN113132717A (en) | Data processing method, terminal and server | |
| CN112307984B (en) | Safety helmet detection method and device based on neural network | |
| CN109034002A (en) | Entity book detection method and device | |
| CN108052918A (en) | A kind of person's handwriting Compare System and method | |
| CN109117748A (en) | Page turning detection method and device | |
| JP7336653B2 (en) | Indoor positioning method using deep learning | |
| CN108563769A (en) | Image display method, device, equipment and the storage medium of mobile terminal | |
| CN111633337B (en) | Reflection eliminating method and device for laser welding seam measurement | |
| CN110728222B (en) | Pose estimation method for target object in mechanical arm grabbing system | |
| CN113469994A (en) | Pantograph detection method, pantograph detection device, electronic apparatus, and storage medium | |
| US11127158B2 (en) | Image indexing and retrieval using local image patches for object three-dimensional pose estimation | |
| JP6717769B2 (en) | Information processing device and program |
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 | ||
| CB02 | Change of applicant information |
Address after: 230031 China (Anhui) pilot Free Trade Zone, Hefei, Anhui province 6 / F and 23 / F, scientific research building, building 2, zone a, China sound Valley, No. 3333 Xiyou Road, high tech Zone, Hefei Applicant after: Anhui taoyun Technology Co.,Ltd. Address before: 230031 9th floor, building 1, tianyuandike science and Technology Park, 66 Qianshui East Road, high tech Zone, Hefei City, Anhui Province Applicant before: ANHUI TAOYUN TECHNOLOGY Co.,Ltd. |
|
| CB02 | Change of applicant information |