CN106780595A - A kind of recognition methods of panoramic pictures and device - Google Patents
A kind of recognition methods of panoramic pictures and device Download PDFInfo
- Publication number
- CN106780595A CN106780595A CN201611163170.7A CN201611163170A CN106780595A CN 106780595 A CN106780595 A CN 106780595A CN 201611163170 A CN201611163170 A CN 201611163170A CN 106780595 A CN106780595 A CN 106780595A
- Authority
- CN
- China
- Prior art keywords
- picture
- pixel
- color value
- identified
- rows
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
Abstract
Recognition methods and device the invention discloses a kind of panoramic pictures.The method includes：Picture to be identified is loaded into internal memory；The similarity of the color value of the symmetrical pixel of relative picture medium line of the right half part of the similarity of color value and the left-half of bottom C rows and bottom C rows of the left side C row pixel symmetrical with the relative picture medium line of the right half part of the similarity, the left-half of top C rows and top C rows of the color value of the symmetrical pixel of relative picture medium line of right side C row of picture to be identified is calculated respectively；If these three similarities are both greater than default similarity threshold Y, judge that picture to be identified is panoramic pictures.It can be seen that, after the present invention is loaded into internal memory to a pictures of user's selection, the type of the similarity-rough set automatic identification picture of the color value of the pixel according to picture, and then according to the corresponding picture scan tool of type selecting of picture, save the operation that user manually selects, it is simple to operate, strengthen Consumer's Experience.
Description
Technical field
The present invention relates to picture recognition technical field, the recognition methods of more particularly to a kind of panoramic pictures and device.
Background technology
With reaching its maturity for virtual reality technology (Virtual Reality, VR), resource under VR patterns is applied not
It is disconnected to emerge in large numbers, it is especially suitable for the various types of pictures watched under VR patterns, such as 2D pictures, panoramic pictures etc..When by VR
When pattern watches various types of pictures, the type of picture is different, and the picture scan tool of selection is also differed.
In the prior art, because cannot automatic identification picture type, select picture scan tool when, typically exist
Corresponding prompting frame is shown to user in display screen, to point out user to manually select corresponding picture scan tool, is caused in VR
Cumbersome, reduction Consumer's Experience during picture is checked under pattern.
The content of the invention
In view of prior art under VR patterns cannot automatic identification picture type, cumbersome when checking picture, drop
The problem of low Consumer's Experience, it is proposed that a kind of recognition methods of panoramic pictures of the invention and device, to solve or at least portion
Solve the above problems with dividing.
According to an aspect of the invention, there is provided a kind of recognition methods of panoramic pictures, methods described includes：
Picture to be identified is loaded into internal memory；
The left side C row of the picture to be identified pixel symmetrical with the relative picture medium line that right side C is arranged is calculated respectively
The similarity of color value, the left-half of the top C rows picture symmetrical with the relative picture medium line of the right half part of top C rows
The relative picture medium line of the similarity of the color value of vegetarian refreshments and the left-half of bottom C rows and the right half part of bottom C rows
The similarity of the color value of symmetrical pixel；Wherein C values are more than or equal to 1, less than picture height H and less than picture width
The natural number of W；
If these three similarities are both greater than default similarity threshold Y, judge that the picture to be identified is panoramic pictures,
Otherwise judge that the picture to be identified is not panoramic pictures.
According to another aspect of the present invention, there is provided a kind of identifying device of panoramic pictures, described device includes：
Picture loading unit, for picture to be identified to be loaded into internal memory；
First similarity calculated, the left side C for calculating the picture to be identified arranges the relative picture arranged with right side C
The similarity of the color value of the symmetrical pixel of medium line；
Second similarity calculated, left-half and top C rows for calculating the top C rows of the picture to be identified
Right half part the symmetrical pixel of relative picture medium line color value similarity；
Third phase seemingly spends computing unit, left-half and bottom C rows for calculating the bottom C rows of the picture to be identified
Right half part the symmetrical pixel of relative picture medium line color value similarity；
Panoramic pictures judging unit, if for first similarity calculated, the second similarity calculated and
Three similarities that three similarity calculateds are calculated are both greater than default similarity threshold Y, then judge described to be identified
Picture is panoramic pictures, otherwise judges that the picture to be identified is not panoramic pictures.
In sum, the technical scheme is that picture to be identified is loaded into internal memory, figure to be identified is calculated respectively
The left side C of piece arranges the similarity of the color value of the pixel symmetrical with the relative picture medium line of right side C row, a left side for top C rows
The similarity of the color value of the half part pixel symmetrical with the relative picture medium line of the right half part of top C rows and bottom
The color value of the left-half of the portion's C rows pixel symmetrical with the relative picture medium line of the right half part of bottom C rows it is similar
Degree, if these three similarities are all higher than default similarity threshold Y, judges that the picture to be identified is panoramic pictures, above-mentioned
Any one or the similarity of two are not more than Y, all judge that the picture to be identified is not panoramic pictures.By skill of the invention
Art scheme, after being loaded into the pictures that user selects in internal memory, can be according to the similarity of the color value of the pixel of picture
Compare the type of automatic identification picture, and then according to the corresponding picture scan tool of type selecting of picture, save user manual
The operation of selection, it is simple to operate, strengthen Consumer's Experience.
Brief description of the drawings
A kind of schematic flow sheet of the recognition methods of panoramic pictures that Fig. 1 is provided for one embodiment of the invention；
Fig. 2 chooses schematic diagram for a kind of pixel of panoramic pictures identification that one embodiment of the invention is provided；
A kind of schematic flow sheet of the recognition methods of panoramic pictures that Fig. 3 is provided for another embodiment of the present invention；
A kind of schematic device of the recognition methods of panoramic pictures that Fig. 4 is provided for one embodiment of the invention.
Specific embodiment
Mentality of designing of the invention is：In view of prior art under VR patterns cannot automatic identification picture type, check
It is cumbersome during picture, the problem of Consumer's Experience is reduced, it is also contemplated that during panoramic pictures are checked in VR models,
Be the equal of to be attached on a sphere, to constitute a real panoramic scene, therefore the left and right sides of panoramic pictures, top
The color of the pixel of left-half and right half part, the left-half of bottom and right half part is close.Basis of the present invention
The These characteristics of panoramic pictures, obtain the left and right sides, top, the color value of the pixel of bottom of picture to be identified respectively, point
Not Ji Suan the left and right sides, the left-half at top and right half part, the left-half of bottom and right half part pixel face
After the similarity of colour, compared with predetermined threshold value, and then automatic decision goes out whether picture to be identified is panoramic pictures, saves use
The operation that family manually selects, it is simple to operate, strengthen Consumer's Experience.To make the object, technical solutions and advantages of the present invention more clear
Chu, below in conjunction with accompanying drawing, embodiment of the present invention is described further in detail.
A kind of schematic flow sheet of the recognition methods of panoramic pictures that Fig. 1 is provided for one embodiment of the invention.This implementation
The height of picture to be identified is H in example, and width is W.As shown in figure 1, the method includes：
Step S110, picture to be identified is loaded into internal memory, and the reading of picture is carried out so as to program.
Step S120, calculates the left side C row of the picture to be identified picture symmetrical with the relative picture medium line that right side C is arranged respectively
The similarity of the color value of vegetarian refreshments, the left-half of top C rows are symmetrical with the relative picture medium line of the right half part of top C rows
Pixel the similarity of color value and the left-half of bottom C rows with the relative picture of the right half part of bottom C rows
The similarity of the color value of the symmetrical pixel of top-stitching；Wherein C values are more than or equal to 1, less than picture height H and less than picture
The natural number of width W.
Fig. 2 chooses schematic diagram for a kind of pixel of panoramic pictures identification that one embodiment of the invention is provided.Such as Fig. 2 institutes
Show, obtain left side C row L (C*H) of picture to be identified, right side C row R (C*H), top C rows UP (C*W), bottom C row DOWN (C*
The color value of each pixel W)；Wherein C values are oneself more than or equal to 1, less than picture height H and less than picture width W
So count.
As shown in Fig. 2 calculating left side C row L (C*H) of picture to be identified and the pixel color value on right side C row R (C*H)
Similarity；Calculate the pixel face of the left-half UP_L (C*W) and the right half part UP_R (C*W) of top C rows of top C rows
The similarity of colour；And calculate the left-half DOWN_L (C*W) of the bottom C rows and right half part DOWN_R (C* of bottom C rows
The similarity of pixel color value W).
Here similarity refers to the pixel color value identical pixel of correspondence position between the two parts being compared
The number of point accounts for the ratio of the pixel total number of the appropriate section of acquisition.For example, obtaining the row of picture left side 1 to be identified and right side
The color value of each pixel of 1 row, the number of total pixel is 1*H, calculate picture to be identified left side 1 arrange L (1*H) with
When the similarity of the pixel color value on row R of right side 1 (1*H), the number of pixel color value identical pixel is obtained
It is a (a≤H), then similarity is exactly a/H.
It should be noted that when judging the number of pixel color value identical pixel, judgement compares
Whether the size of the difference of the color value of correspondence position identical pixel symmetrical relative to center line in two parts is less than pre-
If colour-difference threshold value, be color identical pixel less than the pixel of default colour-difference threshold value.For example, obtaining to be identified
Picture left side 1 arranges L (1*H) and right side 1 arrange R (1*H) each pixel color value, the color value of the pixel in left side is from upper
Be successively under L1, L2 ..., LH, the color value of the pixel on right side be successively from top to bottom R1, R2 ..., RH.Judge
When be the difference, L2 and the R2 that judge L1 and R1 difference, by that analogy, after H pixel is judged, statistical color value
The number of identical pixel.
Step S130, if these three similarities are both greater than default similarity threshold Y, judges that picture to be identified is panorama
Picture, otherwise judges that picture to be identified is not panoramic pictures.
If it should be noted that picture to be identified is panoramic pictures, it is necessary to meet above three similarity all greater than pre-
If similarity threshold Y.If any one or any two are unsatisfactory for, this waits that it is not just panoramic pictures by picture to be.
It can be seen that, by technical scheme, after being loaded into internal memory to the pictures that user selects, can be according to the figure
The type of the similarity-rough set automatic identification picture of the color value of the pixel of piece, so it is corresponding according to the type selecting of picture
Picture scan tool, saves the operation that user manually selects, simple to operate, strengthens Consumer's Experience.
In one embodiment of the invention, the left side C row of the calculating picture to be identified in step S120 and right side C row
The similarity of the color value of the pixel symmetrical with respect to picture medium line includes：
(1) pixel total number is calculated according to C row * picture heights H.For example, what is calculated is that left side 2 arranges, right side 2 arranges
Each pixel color value similarity, then the number of total pixel is 2*H.
(2) color value and relative picture medium line pair of each pixel of the left side C row of picture to be identified are judged successively
Whether the difference of the color value of each pixel of the right side C row of the picture to be identified for claiming is less than default colour-difference threshold X, if
Certain two differences of the color value of pixel symmetrical with respect to picture medium line are less than default colour-difference threshold X, then judge this
Two pixels are same pixel point, and count the number of same pixel point.
For example, what is obtained is that left side 2 arranges L (2*H), the color value of each pixel on row R of right side 2 (2*H).The picture in left side
The pixel color value of the first row of vegetarian refreshments be designated as L11, L12 ..., L1H, the pixel color value of secondary series be designated as L21,
L22、...、L2H；The pixel color value of the first row of the pixel on right side be designated as R11, R12 ..., R1H, the picture of secondary series
Vegetarian refreshments color value be designated as R21, R22 ..., R2H.Wherein, it is symmetrical with respect to picture medium line (line of symmetry) to be：L11 and R11,
L12 and R12 ..., L1H and R1H；And L21 and R21, L22 and R22 ..., L2H and R2H.Above-mentioned symmetrical two are judged successively
Whether the difference between pixel value is less than default colour-difference threshold X, if its difference is less than default colour-difference threshold X, sentences
The two pixels break for same pixel point, and counts the number of same pixel point；If its difference is not less than default colour-difference
Threshold X, then be the pixel for differing.
(3) calculate the ratio of number and the pixel total number of same pixel point, as picture to be identified left side C row with
The similarity of the pixel color value of right side C row.
In above-mentioned example, if the number of the same pixel point of statistics is b (b≤2*H), the number b and picture of same pixel point
The ratio b/ (2*H) of vegetarian refreshments total number 2*H is exactly the phase of the C row in left side with the pixel color value of right side C row of picture to be identified
Like degree.
It should be noted that for the center line in the present embodiment, if W cannot be divided exactly by 2, center line is self-defined
Picture line of symmetry；If W can not be divided exactly by 2, center line is the middle column of whole pixels.
In one embodiment of the invention, the left-half of the top C rows of the calculating picture to be identified in step S130
The similarity of the color value of the pixel symmetrical with the relative picture medium line of the right half part of top C rows includes：
(1) pixel total number is calculated according to C row * picture width W/2.W in the present embodiment can be divided exactly by 2.Example
Such as, calculating be top 3 rows each pixel color value similarity, then the number of total pixel is that 3* (W/2) is individual.
(2) each pixel color value and relative picture of the left-half of the top C rows of picture to be identified are judged successively
Whether the difference of each pixel color value of the right half part of the top C rows of the symmetrical picture to be identified of medium line is less than default
Colour-difference threshold X, if certain symmetrical two difference of the color value of pixel of relative picture medium line are less than default color
Difference threshold X, then judge that the two pixels are same pixel point, and count the number of same pixel point.
For example, what is obtained is the color value of each pixel of 3 row UP (3*W) of top.From left to right, the left side at top
Point pixel the first row pixel color value be designated as UP_L11, UP_L12 ..., UP_L1m, the pixel face of the second row
Colour be designated as UP_L21, UP_L22 ..., UP_L2m, the pixel color value of the third line be designated as UP_L31, UP_L32 ...,
UP_L3m；From right to left, the pixel color value of the first row of the pixel of the right half part at top is designated as UP_R11, UP_
R12 ..., UP_R1m, the pixel color value of the second row be designated as UP_R21, UP_R22 ..., UP_R2m, the pixel of the third line
Point color value be designated as UP_R31, UP_R32 ..., UP_R3m.Wherein, it is symmetrical with respect to picture medium line (line of symmetry) to be：UP_
L11 and UP_R11, UP_L12 and UP_R12 ..., UP_L1m and UP_R1m；UP_L21 and UP_R21, UP_L22 and UP_
R22 ..., UP_L2m and UP_R2m；And UP_L31 and UP_R31, UP_L32 and UP_R32 ..., UP_L3m and UP_R3m.
Wherein, m is equal to W/2.Whether the difference between above-mentioned two symmetrical pixel values is judged successively less than default colour-difference threshold X,
If its difference is less than default colour-difference threshold X, judge that the two pixels are same pixel point, and count same pixel point
Number；It is the pixel for differing if its difference is not less than default colour-difference threshold X.
(3) ratio of number and the pixel total number of same pixel point is calculated, as the top C rows of picture to be identified
The similarity of the pixel color value of the right half part of left-half and top C rows.
In above-mentioned example, if the number of the same pixel point of statistics is d (d≤3* (W/2)), the number d of same pixel point
Left-half and top with the top C rows that the ratio d/ (3* (W/2)) of pixel total number 3* (W/2) is exactly picture to be identified
The similarity of the pixel color value of the right half part of C rows.
It should be noted that whether the width W of the picture to be identified in the present invention is not limited to be divided exactly by 2, of the invention
In embodiment, if W cannot be divided exactly by 2, can be using the middle column of pixel as center line, it is ensured that the left-half at top and
Right half part is symmetrical with respect to center line.
In one embodiment of the invention, the left-half of the bottom C rows of the calculating picture to be identified in step S130
The similarity of the color value of the pixel symmetrical with the relative picture medium line of the right half part of bottom C rows includes：
(1) pixel total number is calculated according to C row * picture width W/2.W in the present embodiment can be divided exactly by 2.Example
Such as, calculating be the row of bottom 3 each pixel color value similarity, then the number of total pixel is that 3* (W/2) is individual.
(2) each pixel color value and relative picture of the left-half of the bottom C rows of picture to be identified are judged successively
Whether the difference of each pixel color value of the right half part of the bottom C rows of the symmetrical picture to be identified of medium line is less than default
Colour-difference threshold X, if certain symmetrical two difference of the color value of pixel of relative picture medium line are less than default color
Difference threshold X, then judge that the two pixels are same pixel point, and count the number of same pixel point.
For example, what is obtained is the color value of each pixel of 3 row DOWN (3*W) of top.From left to right, left the half of top
The pixel color value of the first row of partial pixel be designated as DOWN_L11, DOWN_L12 ..., DOWN_L1m, the second row
Pixel color value be designated as DOWN_L21, DOWN_L22 ..., DOWN_L2m, the pixel color value of the third line is designated as DOWN_
L31、DOWN_L32、...、DOWN_L3m；From right to left, the pixel color of the first row of the pixel of the right half part at top
Value be designated as DOWN_R11, DOWN_R12 ..., DOWN_R1m, the pixel color value of the second row is designated as DOWN_R21, DOWN_
R22 ..., DOWN_R2m, the pixel color value of the third line be designated as DOWN_R31, DOWN_R32 ..., DOWN_R3m.Wherein,
Symmetrical with respect to picture medium line (line of symmetry) is：DOWN_L11 and DOWN_R11, DOWN_L12 and DOWN_R12 ...,
DOWN_L1m and DOWN_R1m；DOWN_L21 and DOWN_R21, DOWN_L22 and DOWN_R22 ..., DOWN_L2m and DOWN_
R2m；And DOWN_L31 and DOWN_R31, DOWN_L32 and DOWN_R32 ..., DOWN_L3m and DOWN_R3m.Wherein, m etc.
In W/2.Whether the difference between above-mentioned two symmetrical pixel values is judged successively less than default colour-difference threshold X, if its difference
Less than default colour-difference threshold X, then judge that the two pixels are same pixel point, and count the number of same pixel point；
It is the pixel for differing if its difference is not less than default colour-difference threshold X.
(3) ratio of number and the pixel total number of same pixel point is calculated, as the bottom C rows of picture to be identified
The similarity of the pixel color value of the right half part of left-half and bottom C rows.
In above-mentioned example, if the number of the same pixel point of statistics is c (c≤3* (W/2)), the number c of same pixel point
Left-half and top with the bottom C rows that the ratio c/ (3* (W/2)) of pixel total number 3* (W/2) is exactly picture to be identified
The similarity of the pixel color value of the right half part of C rows.
It should be noted that whether the width W of the picture to be identified in the present invention is not limited to be divided exactly by 2, of the invention
In embodiment, if W cannot be divided exactly by 2, can be using the middle column of pixel as center line, it is ensured that the left-half of bottom and
Right half part is symmetrical with respect to center line.
In one embodiment of the invention, similarity threshold Y and colour-difference threshold X are preset based on experience value, and by C
Value is equal to 1.
A kind of schematic flow sheet of the recognition methods of panoramic pictures that Fig. 3 is provided for another embodiment of the present invention.Such as Fig. 3
Shown, the recognition methods of the panoramic pictures includes：
Step S310, internal memory is loaded into by picture to be identified.
Step S320, judges whether the left side C row of picture to be identified are big with the similarity of the pixel color value that right side C is arranged
In default similarity threshold Y.If being judged as YES, step S330 is carried out；If being judged as NO, judge that picture to be identified is not
Panoramic pictures.
Step S330, judges that the left-half of top C rows is similar to the pixel color value of the right half part of top C rows
Whether degree is more than default similarity threshold Y.If being judged as YES, step S340 is carried out；If being judged as NO, judge to be identified
Picture is not panoramic pictures.
Step S340, judges that the left-half of bottom C rows is similar to the pixel color value of the right half part of bottom C rows
Whether degree is more than default similarity threshold Y.If being judged as YES, judge that picture to be identified is panoramic pictures；If being judged as NO,
Then judge that picture to be identified is not panoramic pictures.
In the present embodiment, first determine whether that the left side C row of picture to be identified are similar to the pixel color value that right side C is arranged
Degree；When being judged as YES, judge that the left-half of top C rows is similar to the pixel color value of the right half part of top C rows
Degree；When being judged as YES, the left-half of bottom C rows and the phase of the pixel color value of the right half part of bottom C rows are finally judged
Like degree.It is to be judged as NO that there is any one step centre, it is determined that picture to be identified is not panoramic pictures.
It should be noted that the left and right sides for picture to be identified, top and bottom in technical scheme
What the judgement order of the similarity of pixel was not restricted, can be judged successively, and the sequencing for judging successively is not limited；
Also can be judged simultaneously.
A kind of schematic device of the recognition methods of panoramic pictures that Fig. 4 is provided for one embodiment of the invention.Such as Fig. 4 institutes
Show, 400 devices of the recognition methods of the panoramic pictures include：
Picture loading unit 410, for picture to be identified to be loaded into internal memory.
First similarity calculated 420, the left side C for calculating picture to be identified arranges the relative picture arranged with right side C
The similarity of the color value of the symmetrical pixel of medium line.
Second similarity calculated 430, left-half and top C rows for calculating the top C rows of picture to be identified
Right half part the symmetrical pixel of relative picture medium line color value similarity.
Third phase is like degree computing unit 440, left-half and bottom C rows for calculating the bottom C rows of picture to be identified
Right half part the symmetrical pixel of relative picture medium line color value similarity.
Panoramic pictures judging unit 450, if for the first similarity calculated, the second similarity calculated and the 3rd
Three similarities that similarity calculated is calculated are both greater than default similarity threshold Y, then judge that picture to be identified is
Panoramic pictures, otherwise judge that picture to be identified is not panoramic pictures.
In one embodiment of the invention, the first similarity calculated 420 specifically for,
Pixel total number is calculated according to C row * picture heights H；
Judge that each pixel color value of the left side C row of picture to be identified is symmetrical with relative picture medium line successively to treat
Recognize whether the difference of each pixel color value of the right side C row of picture is less than default colour-difference threshold X, if relative picture
Certain symmetrical two difference of the color value of pixel of medium line are less than default colour-difference threshold X, then judge the two pixels
Point is same pixel point, and counts the number of same pixel point；
The number of same pixel point and the ratio of pixel total number are calculated, as left side C row and the right side of picture to be identified
The similarity of the pixel color value of side C row.
In one embodiment of the invention, the second similarity calculated 430 specifically for,
Pixel total number is calculated according to C row * picture width W/2；
Judge successively in the middle of each pixel color value and relative picture of the left-half of the top C rows of picture to be identified
Whether the difference of each pixel color value of the right half part of the top C rows of the symmetrical picture to be identified of line is less than default face
Color difference threshold X, if certain symmetrical two difference of the color value of pixel of relative picture medium line are less than default color difference limen
Value X, then judge that the two pixels are same pixel point, and count the number of same pixel point；
The number of same pixel point and the ratio of pixel total number are calculated, as a left side for the top C rows of picture to be identified
The similarity of the pixel color value of the right half part of half part and top C rows.
In one embodiment of the invention, third phase like spend computing unit 440 specifically for,
Pixel total number is calculated according to C row * picture width W/2；
Judge successively in the middle of each pixel color value and relative picture of the left-half of the bottom C rows of picture to be identified
Whether the difference of each pixel color value of the right half part of the bottom C rows of the symmetrical picture to be identified of line is less than default face
Color difference threshold X, if certain symmetrical two difference of the color value of pixel of relative picture medium line are less than default color difference limen
Value X, then judge that the two pixels are same pixel point, and count the number of same pixel point；
The number of same pixel point and the ratio of pixel total number are calculated, as a left side for the bottom C rows of picture to be identified
The similarity of the pixel color value of the right half part of half part and bottom C rows.
In one embodiment of the invention, similarity threshold Y and colour-difference threshold X are preset based on experience value, and C takes
Value is equal to 1.
It should be noted that each embodiment and the method shown in Fig. 1 of device shown in method and Fig. 4 shown in Fig. 3
Each embodiment correspondence is identical, has been described in detail above, will not be repeated here.
In sum, the technical scheme is that picture to be identified is loaded into internal memory, figure to be identified is calculated respectively
The left side C of piece arranges the similarity of the color value of the pixel symmetrical with the relative picture medium line of right side C row, a left side for top C rows
The similarity of the color value of the half part pixel symmetrical with the relative picture medium line of the right half part of top C rows and bottom
The color value of the left-half of the portion's C rows pixel symmetrical with the relative picture medium line of the right half part of bottom C rows it is similar
Degree, if these three similarities are all higher than default similarity threshold Y, judges that the picture to be identified is panoramic pictures, above-mentioned
Any one or the similarity of two are not more than Y, all judge that the picture to be identified is not panoramic pictures.By skill of the invention
Art scheme, after being loaded into internal memory to the pictures that user selects, can be according to the similarity of the color value of the pixel of picture ratio
Compared with the type of automatic identification picture, and then according to the corresponding picture scan tool of type selecting of picture, save user and select manually
The operation selected, it is simple to operate, strengthen Consumer's Experience.
The above, specific embodiment only of the invention, under above-mentioned teaching of the invention, those skilled in the art
Other improvement or deformation can be carried out on the basis of above-described embodiment.It will be understood by those skilled in the art that above-mentioned tool
The purpose of the present invention is simply preferably explained in body description, and protection scope of the present invention should be defined by scope of the claims.
Claims (10)
1. a kind of recognition methods of panoramic pictures, it is characterised in that methods described includes：
Picture to be identified is loaded into internal memory；
The left side C for calculating the picture to be identified respectively arranges the face of the pixel symmetrical with the relative picture medium line that right side C is arranged
The similarity of colour, the left-half of top C the rows pixel symmetrical with the relative picture medium line of the right half part of top C rows
The similarity of color value and the left-half of bottom C rows it is symmetrical with the relative picture medium line of the right half part of bottom C rows
Pixel color value similarity；Wherein C values are more than or equal to 1, less than picture height H and less than picture width W
Natural number；
If these three similarities are both greater than default similarity threshold Y, judge that the picture to be identified is panoramic pictures, otherwise
Judge that the picture to be identified is not panoramic pictures.
2. the method for claim 1, it is characterised in that left side C row and the right side C of the calculating picture to be identified
The similarity of the color value of the symmetrical pixel of relative picture medium line of row includes：
Pixel total number is calculated according to C row * picture heights H；
Judge that the color value of each pixel of the left side C row of the picture to be identified is symmetrical with relative picture medium line successively
Whether the difference of the color value of each pixel of the right side C row of the picture to be identified is less than default colour-difference threshold X, if
Certain two differences of the color value of pixel symmetrical with respect to picture medium line are less than default colour-difference threshold X, then judge this
Two pixels are same pixel point, and count the number of same pixel point；
The number of the same pixel point and the ratio of the pixel total number are calculated, as the left side of the picture to be identified
C arranges the similarity with the pixel color value of right side C row.
3. the method for claim 1, it is characterised in that left the half of the top C rows of the calculating picture to be identified
The similarity of the part color value symmetrical with the relative picture medium line of the pixel of the right half part of top C rows includes：
Pixel total number is calculated according to C row * picture width W/2；
Judge successively in the middle of each pixel color value and relative picture of the left-half of the top C rows of the picture to be identified
Whether the difference of each pixel color value of the right half part of the top C rows of the symmetrical picture described to be identified of line is less than default
Colour-difference threshold X, if certain symmetrical two difference of the color value of pixel of relative picture medium line are less than default color
Difference threshold X, then judge that the two pixels are same pixel point, and count the number of same pixel point；
The number of the same pixel point and the ratio of the pixel total number are calculated, as the top of the picture to be identified
The similarity of the pixel color value of the left-half of C rows and the right half part of top C rows.
4. the method for claim 1, it is characterised in that left the half of the bottom C rows of the calculating picture to be identified
The similarity of the color value of the part pixel symmetrical with the relative picture medium line of the right half part of bottom C rows includes：
Pixel total number is calculated according to C row * picture width W/2；
Judge successively in the middle of each pixel color value and relative picture of the left-half of the bottom C rows of the picture to be identified
Whether the difference of each pixel color value of the right half part of the bottom C rows of the symmetrical picture described to be identified of line is less than default
Colour-difference threshold X, if certain symmetrical two difference of the color value of pixel of relative picture medium line are less than default color
Difference threshold X, then judge that the two pixels are same pixel point, and count the number of same pixel point；
The number of the same pixel point and the ratio of the pixel total number are calculated, as the bottom of the picture to be identified
The similarity of the pixel color value of the left-half of C rows and the right half part of bottom C rows.
5. the method as described in claim any one of 2-4, it is characterised in that preset the similarity threshold Y based on experience value
With the colour-difference threshold X, and by C values be equal to 1.
6. a kind of identifying device of panoramic pictures, it is characterised in that described device includes：
Picture loading unit, for picture to be identified to be loaded into internal memory；
First similarity calculated, the left side C for calculating the picture to be identified is arranged in the middle of the relative picture arranged with right side C
The similarity of the color value of the symmetrical pixel of line；
Second similarity calculated, for calculating the left-half of the top C rows of the picture to be identified and the right side of top C rows
The similarity of the color value of the symmetrical pixel of the relative picture medium line of half part；
Third phase seemingly spends computing unit, for calculating the left-half of the bottom C rows of the picture to be identified and the right side of bottom C rows
The similarity of the color value of the symmetrical pixel of the relative picture medium line of half part；
Panoramic pictures judging unit, if for first similarity calculated, the second similarity calculated and third phase
Three similarities being calculated like degree computing unit are both greater than default similarity threshold Y, then judge the picture to be identified
It is panoramic pictures, otherwise judges that the picture to be identified is not panoramic pictures.
7. device as claimed in claim 6, it is characterised in that first similarity calculated specifically for,
Pixel total number is calculated according to C row * picture heights H；
Each pixel color value of the left side C row of the picture to be identified institute symmetrical with relative picture medium line is judged successively
Whether the difference of each pixel color value of the right side C row of picture to be identified is stated less than default colour-difference threshold X, if relatively
Certain symmetrical two difference of the color value of pixel of picture medium line are less than default colour-difference threshold X, then judge the two
Pixel is same pixel point, and counts the number of same pixel point；
The number of the same pixel point and the ratio of the pixel total number are calculated, as the left side of the picture to be identified
C arranges the similarity with the pixel color value of right side C row.
8. device as claimed in claim 6, it is characterised in that second similarity calculated specifically for,
Pixel total number is calculated according to C row * picture width W/2；
Judge successively in the middle of each pixel color value and relative picture of the left-half of the top C rows of the picture to be identified
Whether the difference of each pixel color value of the right half part of the top C rows of the symmetrical picture described to be identified of line is less than default
Colour-difference threshold X, if certain symmetrical two difference of the color value of pixel of relative picture medium line are less than default color
Difference threshold X, then judge that the two pixels are same pixel point, and count the number of same pixel point；
The number of the same pixel point and the ratio of the pixel total number are calculated, as the top of the picture to be identified
The similarity of the pixel color value of the left-half of C rows and the right half part of top C rows.
9. device as claimed in claim 6, it is characterised in that the third phase like spend computing unit specifically for,
Pixel total number is calculated according to C row * picture width W/2；
Judge successively in the middle of each pixel color value and relative picture of the left-half of the bottom C rows of the picture to be identified
Whether the difference of each pixel color value of the right half part of the bottom C rows of the symmetrical picture described to be identified of line is less than default
Colour-difference threshold X, if certain symmetrical two difference of the color value of pixel of relative picture medium line are less than default color
Difference threshold X, then judge that the two pixels are same pixel point, and count the number of same pixel point；
The number of the same pixel point and the ratio of the pixel total number are calculated, as the bottom of the picture to be identified
The similarity of the pixel color value of the left-half of C rows and the right half part of bottom C rows.
10. the device as described in claim any one of 6-9, it is characterised in that
The similarity threshold Y and the colour-difference threshold X are preset based on experience value, and C values are equal to 1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611163170.7A CN106780595A (en) | 2016-12-15 | 2016-12-15 | A kind of recognition methods of panoramic pictures and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611163170.7A CN106780595A (en) | 2016-12-15 | 2016-12-15 | A kind of recognition methods of panoramic pictures and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106780595A true CN106780595A (en) | 2017-05-31 |
Family
ID=58892785
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611163170.7A Pending CN106780595A (en) | 2016-12-15 | 2016-12-15 | A kind of recognition methods of panoramic pictures and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106780595A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107197326A (en) * | 2017-06-21 | 2017-09-22 | 歌尔科技有限公司 | Video frequency identifying method and device |
| CN107330944A (en) * | 2017-06-27 | 2017-11-07 | 深圳市冠旭电子股份有限公司 | A kind of recognition methods of panoramic picture, device, terminal device and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103927719A (en) * | 2014-04-04 | 2014-07-16 | 北京金山网络科技有限公司 | Picture processing method and device |
| CN106101743A (en) * | 2016-08-23 | 2016-11-09 | 广东欧珀移动通信有限公司 | Panoramic video recognition methods and device |
| CN106162146A (en) * | 2016-07-29 | 2016-11-23 | 暴风集团股份有限公司 | Automatically identify and the method and system of playing panoramic video |
-
2016
- 2016-12-15 CN CN201611163170.7A patent/CN106780595A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103927719A (en) * | 2014-04-04 | 2014-07-16 | 北京金山网络科技有限公司 | Picture processing method and device |
| CN106162146A (en) * | 2016-07-29 | 2016-11-23 | 暴风集团股份有限公司 | Automatically identify and the method and system of playing panoramic video |
| CN106101743A (en) * | 2016-08-23 | 2016-11-09 | 广东欧珀移动通信有限公司 | Panoramic video recognition methods and device |
Non-Patent Citations (1)
| Title |
|---|
| 董绍克: "《汉语方言词汇比较研究》", 31 March 2013, 北京：商务印书馆 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107197326A (en) * | 2017-06-21 | 2017-09-22 | 歌尔科技有限公司 | Video frequency identifying method and device |
| CN107197326B (en) * | 2017-06-21 | 2019-12-03 | 歌尔科技有限公司 | Video frequency identifying method and device |
| CN107330944A (en) * | 2017-06-27 | 2017-11-07 | 深圳市冠旭电子股份有限公司 | A kind of recognition methods of panoramic picture, device, terminal device and storage medium |
| CN107330944B (en) * | 2017-06-27 | 2020-01-21 | 深圳市冠旭电子股份有限公司 | Panoramic image identification method and device, terminal equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104486552B (en) | A kind of method and electronic equipment obtaining image | |
| CN103632352B (en) | Method for time domain noise reduction of noise image and related device | |
| US7931602B2 (en) | Gaze guidance degree calculation system, gaze guidance degree calculation program, storage medium, and gaze guidance degree calculation method | |
| CN104902088B (en) | Mobile terminal adjusts the method and device of screen intensity | |
| CN109257582A (en) | A kind of bearing calibration of projection device and device | |
| US9406140B2 (en) | Method and apparatus for generating depth information | |
| CN110517246A (en) | A kind of image processing method, device, electronic equipment and storage medium | |
| CN103902875B (en) | The generation method and device and verification method and device of identifying code picture | |
| US20130300828A1 (en) | Image processing device, non-transitory computer readable recording medium, and image processing method | |
| CN104994374B (en) | Method for automatically detecting three-dimensional format of video | |
| CN111062331B (en) | Image mosaic detection method and device, electronic equipment and storage medium | |
| CN106534683A (en) | Method and system for replacing background images of smart camera in real time | |
| CN102271262A (en) | Multithread-based video processing method for 3D (Three-Dimensional) display | |
| CN107798704A (en) | A kind of realtime graphic stacking method and device for augmented reality | |
| CN105243662B (en) | The determination method and terminal device of a kind of terminal position | |
| CN106651879A (en) | Method and system for extracting nail image | |
| CN107464251A (en) | The black edge detection method and device of a kind of image | |
| CN106780595A (en) | A kind of recognition methods of panoramic pictures and device | |
| CN105468617A (en) | Method and device for regulating webpage content color | |
| CN110458790A (en) | A kind of image detecting method, device and computer storage medium | |
| CN106815598B (en) | 360-degree panoramic picture identification method and device | |
| CN105872516A (en) | Method and device for obtaining parallax parameters of three-dimensional film source | |
| CN106548114A (en) | Image processing method and device | |
| CN107909579A (en) | Product profile extraction method in vision-based detection | |
| CN108156504B (en) | Video display method and device |
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 |
Application publication date: 20170531 |
|
| RJ01 | Rejection of invention patent application after publication |