CN108921779A - A kind of tripleplane's mapping method and device - Google Patents
A kind of tripleplane's mapping method and device Download PDFInfo
- Publication number
- CN108921779A CN108921779A CN201810771103.6A CN201810771103A CN108921779A CN 108921779 A CN108921779 A CN 108921779A CN 201810771103 A CN201810771103 A CN 201810771103A CN 108921779 A CN108921779 A CN 108921779A
- Authority
- CN
- China
- Prior art keywords
- coordinate
- source images
- targeted graphical
- point
- graphical region
- 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
- 238000000034 method Methods 0.000 title claims abstract description 33
- 238000013507 mapping Methods 0.000 title claims abstract description 24
- 238000012545 processing Methods 0.000 claims abstract description 27
- 239000011159 matrix material Substances 0.000 claims description 12
- 238000006243 chemical reaction Methods 0.000 claims description 6
- 230000003139 buffering effect Effects 0.000 claims description 2
- 239000002699 waste material Substances 0.000 abstract description 8
- 238000010586 diagram Methods 0.000 description 9
- 230000006870 function Effects 0.000 description 4
- 230000009466 transformation Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 230000005055 memory storage Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Classifications
-
- G06T3/08—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T1/00—General purpose image data processing
- G06T1/0007—Image acquisition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T1/00—General purpose image data processing
- G06T1/60—Memory management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformation in the plane of the image
- G06T3/40—Scaling the whole image or part thereof
- G06T3/4023—Decimation- or insertion-based scaling, e.g. pixel or line decimation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformation in the plane of the image
- G06T3/40—Scaling the whole image or part thereof
- G06T3/4092—Image resolution transcoding, e.g. client/server architecture
Abstract
The present invention relates to technical field of data processing, discloses a kind of tripleplane's mapping method and device, this method include：Targeted graphical region is obtained, the coordinate of any point in targeted graphical region is obtained by point by point scanning；The coordinate of corresponding source images is calculated according to the coordinate of any point in targeted graphical region；The coordinate of source images is saved to cache memory；The coordinate of source images is obtained from cache memory, and the coordinate of source images is converted to the physical address of source images；Interpolation processing is carried out according to the coordinate of physical address and source images, obtains interpolating pixel；It is exported according to the displaing coordinate of current output image；Source images coordinate is cached by cache memory CACHE to cache on piece, and main process modules are executed parallel, improve processing speed；Accomplish the real-time processing that the image three-dimensional for the above resolution ratio of full HD converts, avoids the waste for writing region of memory bring memory bandwidth again.
Description
Technical field
The present invention relates to technical field of data processing more particularly to a kind of tripleplane's mapping methods and device.
Background technique
One feature of tripleplane's mapping of image is the point coordinate in source images corresponding to the consecutive points of target area
It changes greatly, the feature is especially apparent when especially projection angle is big, the position of required data in memory when causing to handle
It changes greatly, for common bus structures, burst transfer characteristic, which is unable to get, to be made full use of, high for more common large capacity
Fast memory storage SDRAM, it is low that address discontinuously will also result in its access efficiency.And can not when anamorphose angle is larger
Guarantee that the real-time of the above image in different resolution deformation process of full HD is both needed to write back SDRAM, cause system bandwidth into one
Step waste.
Summary of the invention
It is a primary object of the present invention to propose a kind of tripleplane's mapping method and device, pass through cache memory
CACHE caches to cache source images coordinate on piece, and main process modules are executed parallel, improve processing speed；
Accomplish the real-time processing that the image three-dimensional for the above resolution ratio of full HD converts, avoids and write region of memory again and bring
Memory bandwidth waste.
To achieve the above object, a kind of tripleplane's mapping method provided by the invention, including：
Targeted graphical region is obtained, the coordinate of any point in the targeted graphical region is obtained by point by point scanning；
The coordinate of corresponding source images is calculated according to the coordinate of any point in the targeted graphical region；
The coordinate of the source images is saved to cache memory；
The coordinate of the source images is obtained from the cache memory, and the coordinate of the source images is converted into source
The physical address of image；
Interpolation processing is carried out according to the coordinate of the physical address and the source images, obtains interpolating pixel；
It is exported according to the displaing coordinate of current output image.
Optionally, the acquisition targeted graphical region obtains any point in the targeted graphical region by point by point scanning
Coordinate include：
Targeted graphical region is obtained, the profile in the targeted graphical region is calculated；
Point by point scanning is carried out according to preset spacing to the image in the profile；
Obtain the coordinate of any point in the targeted graphical region.
Optionally, the coordinate according to any point in the targeted graphical region calculates the coordinate packet of corresponding source images
It includes：
Obtain the apex coordinate in the targeted graphical region and the apex coordinate of the source images；
The coordinate of any point and the coordinate of corresponding source images in the targeted graphical region are calculated according to the apex coordinate
Between coefficient matrix；
The coordinate of source images is calculated according to the following formula：
U=(m11*x+m12*y+m13)/(m31*x+m32*y+m33)
V=(m21*x+m22*y+m23)/(m31*x+m32*y+m33)
Wherein, (u, v) be source images coordinate, (x, y) be targeted graphical region coordinate, m11, m12, m13, m21,
M22, m23, m31, m32, m33 are coefficient matrix.
Optionally, the coordinate that the source images are obtained from the cache memory, and by the source images
The physical address that coordinate is converted to source images includes：
Judge that the cache memory whether there is the coordinate of required source images, if so, directly acquiring described
The coordinate of source images；
Otherwise, update it is one whole data cached in the cache memory, from update it is described it is data cached in
Obtain the coordinate of the source images；
The coordinate of the source images is converted to the physical address of source images.
Optionally, the displaing coordinate according to current output image, which export, includes：
Judge that the displaing coordinate of current output image whether in the targeted graphical region, is inserted if so, output is described
It is worth pixel；
Otherwise, background color is exported.
As another aspect of the present invention, a kind of tripleplane's mapping device provided, including：
Module is obtained to obtain by point by point scanning any in the targeted graphical region for obtaining targeted graphical region
The coordinate of point；
Computing module calculates the coordinate of corresponding source images for the coordinate according to any point in the targeted graphical region；
Preserving module, for saving the coordinate of the source images to cache memory；
Conversion module, for obtaining the coordinate of the source images from the cache memory, and by the source images
Coordinate be converted to the physical address of source images；
Interpolating module obtains interpolation for carrying out interpolation processing according to the coordinate of the physical address and the source images
Pixel；
Output module, for being exported according to the displaing coordinate of current output image.
Optionally, the acquisition module includes：
Profile acquiring unit calculates the profile in the targeted graphical region for obtaining targeted graphical region；
Scanning element, for carrying out point by point scanning according to preset spacing to the image in the profile；
Coordinate acquiring unit, for obtaining the coordinate of any point in the targeted graphical region.
Optionally, the computing module includes：
Vertex acquiring unit, the vertex for the apex coordinate and the source images that obtain the targeted graphical region are sat
Mark；
Coefficient calculation unit, for according to the apex coordinate calculate in the targeted graphical region coordinate of any point with
Coefficient matrix between the coordinate of corresponding source images；
The coordinate of source images is calculated according to the following formula：
U=(m11*x+m12*y+m13)/(m31*x+m32*y+m33)
V=(m21*x+m22*y+m23)/(m31*x+m32*y+m33)
Wherein, (u, v) be source images coordinate, (x, y) be targeted graphical region coordinate, m11, m12, m13, m21,
M22, m23, m31, m32, m33 are coefficient matrix.
Optionally, the conversion module includes：
Judge that the cache memory whether there is the coordinate of required source images, if so, directly acquiring described
The coordinate of source images；
Otherwise, update it is one whole data cached in the cache memory, from update it is described it is data cached in
Obtain the coordinate of the source images；
The coordinate of the source images is converted to the physical address of source images.
Optionally, the output module includes：
Judge that the displaing coordinate of current output image whether in the targeted graphical region, is inserted if so, output is described
It is worth pixel；
Otherwise, background color is exported.
A kind of tripleplane's mapping method proposed by the present invention and device, this method include：Targeted graphical region is obtained, is led to
It crosses point by point scanning and obtains the coordinate of any point in the targeted graphical region；According to the seat of any point in the targeted graphical region
Mark calculates the coordinate of corresponding source images；The coordinate of the source images is saved to cache memory；From the speed buffering
Memory obtains the coordinate of the source images, and the coordinate of the source images is converted to the physical address of source images；According to institute
The coordinate for stating physical address and the source images carries out interpolation processing, obtains interpolating pixel；According to the display of current output image
Coordinate is exported；It caches source images coordinate by cache memory CACHE to cache on piece, so that mainly handling mould
Block can execute parallel, improve processing speed；Accomplish the real-time place that the image three-dimensional for the above resolution ratio of full HD converts
Reason, avoids the waste for writing region of memory bring memory bandwidth again.
Detailed description of the invention
Fig. 1 is a kind of flow chart for tripleplane's mapping method that the embodiment of the present invention one provides；
Fig. 2 is a kind of schematic diagram for tripleplane's mapping function that the embodiment of the present invention one provides；
Fig. 3 is the method flow diagram of step S10 in Fig. 1；
Fig. 4 is the scanning schematic diagram that the embodiment of the present invention one provides；
Fig. 5 is the structural schematic diagram for the cache memory CACHE that the embodiment of the present invention one provides；
Fig. 6 is a kind of exemplary block diagram of tripleplane's mapping device provided by Embodiment 2 of the present invention；
Fig. 7 is the exemplary block diagram that module is obtained in Fig. 6；
Fig. 8 is the exemplary block diagram of computing module in Fig. 6.
The embodiments will be further described with reference to the accompanying drawings for the realization, the function and the advantages of the object of the present invention.
Specific embodiment
It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, it is not intended to limit the present invention.
Embodiment one
As shown in Figure 1, in the present embodiment, a kind of tripleplane's mapping method, including：
S10, targeted graphical region is obtained, the coordinate of any point in the targeted graphical region is obtained by point by point scanning；
S20, the coordinate that corresponding source images are calculated according to the coordinate of any point in the targeted graphical region；
S30, the coordinate of the source images is saved to cache memory；
S40, the coordinate that the source images are obtained from the cache memory, and the coordinate of the source images is converted
For the physical address of source images；
S50, interpolation processing is carried out according to the coordinate of the physical address and the source images, obtains interpolating pixel；
S60, it is exported according to the displaing coordinate of current output image.
In the present embodiment, source images coordinate is cached by cache memory CACHE to cache on piece, makes winner
It wants processing module that can execute parallel, improves processing speed；Accomplish that the image three-dimensional for the above resolution ratio of full HD converts
Real-time processing, avoid the waste for writing region of memory bring memory bandwidth again.
In the present embodiment, its primitive meaning of cache memory CACHE refers to that access speed is accessed than general random
A kind of RAM that memory body (RAM) comes fastly, it is however generally that it uses DRAM technology unlike system memory main body, and uses
Expensive but SRAM technology faster, also there is the title of cache memory.Cache memory be present in main memory and CPU it
Between single-level memory, be made of static storage chip (SRAM), capacity is smaller but speed is more much higher than main memory, close to
The speed of CPU.It is that the high speed between central processing unit and main memory is small in the hierarchical structure of computer memory system
Capacity memory.It constitutes the memory of level-one with main memory together.Information between cache memory and main memory
Scheduling and transmission be to be carried out automatically by hardware.
As shown in Fig. 2, the function that the present embodiment is mainly realized is that source images A is obtained target area by 3D deformation process
Targeted graphical region B on D, when making observer to observe image perpendicular to the angle of target area D, 3D effect is presented in image.
In the present embodiment, the quadrangle that targeted graphical region is made of 4 vertex b0, b1, b2, b3, therefore
The coordinate of point in the case where known 4 apex coordinates on any side of quadrangle can be calculated by two point form linear equation
Come, so as to obtain the coordinate profiles of quadrangle, by the b0b2b3 in Fig. 2, this two sides are known as left profile for convenience of description,
This two sides b0b1b3 are known as right wheel exterior feature.
As shown in figure 3, in the present embodiment, the step S10 includes：
S11, targeted graphical region is obtained, calculates the profile in the targeted graphical region；
S12, point by point scanning is carried out according to preset spacing to the image in the profile；
S13, the coordinate for obtaining any point in the targeted graphical region.
In the present embodiment, as shown in figure 4, the direction y is motionless using b0 as starting point (x0, y0), by the incremental direction x0+ of x
Δ x, x0+2 Δ x, x0+3 Δ x ..., x0+k Δ x scanning, the right wheel exterior feature until encountering figure terminate current row scanning.Then it jumps
Start to scan to next line, the origin coordinates of next line is chosen from the left profile of image, and coordinate is (x0+k Δ y, y0+ Δ y)
Wherein (x0, y0) is the origin coordinates of lastrow, and the spacing of two rows is Δ y, and k is the slope of left profile, is swept by the direction that x is incremented by
It retouches, the right wheel exterior feature until encountering figure terminates.The every a line of multiple scanning is until figure last line just terminates like this.
In the present embodiment, the step S20 includes：
Obtain the apex coordinate in the targeted graphical region and the apex coordinate of the source images；
The coordinate of any point and the coordinate of corresponding source images in the targeted graphical region are calculated according to the apex coordinate
Between coefficient matrix；
The coordinate of source images is calculated according to the following formula：
U=(m11*x+m12*y+m13)/(m31*x+m32*y+m33)
V=(m21*x+m22*y+m23)/(m31*x+m32*y+m33)
Wherein, (u, v) be source images coordinate, (x, y) be targeted graphical region coordinate, m11, m12, m13, m21,
M22, m23, m31, m32, m33 are coefficient matrix.
In the present embodiment, the calculating of source images coordinate carries out simultaneously with scanning, puts beginning source figure from (x0, the y0) on D
As the changing rule of coordinate is as follows：
The every increase Δ x of the upper x coordinate of D, the value of molecule adds up m11* Δ x in above formula, and denominator value adds up m31* Δ x；
The every increase Δ y of the upper y-coordinate of D, the value of molecule adds up m12* Δ y in above formula, and denominator value adds up m32* Δ y；
The coordinate in corresponding source images A can be calculated by formula (molecule is divided by denominator), calculated coordinate has
Integer part and fractional part, wherein for taking a use, fractional part uses integer part for interpolation.
In the present embodiment, step S30 keeps in the calculated coordinate of step S20 to on-chip memory, so that step S10,
S20 can continue to calculate the source images coordinate of next object pixel, and the work without waiting step S40 is completed.Due to outer
Portion's memory is that system shares and speed is slower, if necessary to update CACHE, step S40 ability after usually wanting the long period
Receive next group of coordinate.The effect of step S30 is to make step S10, S20 and step S40, S50 can be with concurrent working, to be promoted
Efficiency.
In the present embodiment, the step S40 includes：
Judge that the cache memory whether there is the coordinate of required source images, if so, directly acquiring described
The coordinate of source images；
Otherwise, update it is one whole data cached in the cache memory, from update it is described it is data cached in
Obtain the coordinate of the source images；
The coordinate of the source images is converted to the physical address of source images.
In the present embodiment, by the physical address for being converted into source images of source images, image A is obtained from the buffer area CACHE
Data.The buffer area CACHE is for caching the source image data obtained from memory.If required pixel data is present in currently
In CACHE, then this data is directly returned to from CACHE, if pixel data is needed to be not present in current CACHE, it is whole to update one
CACHE line.According to practical application request, CACHE can be configured to one or more, per all the way correspond to a TAG table with
Record the start physical address of source images corresponding to each CACHE line in this road.The length of CACHE can be according to image
Tilt angle be configured to storage fixed number pixel.The number of pixels that can be stored in every CACHE line is the same.
The length m and CACHE of number n, the CACHE line of the size K and CACHE of on-chip memory need to meet per depth t all the way
Following formula：
K>=n*m*t
As shown in figure 5, being the structural schematic diagram of cache memory CACHE.
Because target image is lateral point by point scanning, the reading of corresponding source images has certain local correlations.Work as target
When image inclination angle is smaller, the corresponding lateral correlation of original image reading is stronger, and CACHE line length is longer to be promoted
Efficiency.When target image tilt angle is larger, it is stronger that corresponding original image reads longitudinal correlation, and usually SDRAM is only
Can by row laterally read, therefore CACHE line length is shorter can raising efficiency.The design can be according to the inclination angle of target image
The length m of number n, the CACHE line of reasonable disposition CACHE is spent with CACHE per depth t all the way, to guarantee on-chip memory
Make full use of.
In the present embodiment, the step S60 includes：
Judge that the displaing coordinate of current output image whether in the targeted graphical region, is inserted if so, output is described
It is worth pixel；
Otherwise, background color is exported.
In the present embodiment, the CACHE structure being connected using group optimize graph image projection transformation when due to reading source
The low problem of access SDR/DDR SDRAM access efficiency brought by image slices vegetarian refreshments is discontinuous.It is tied by flexible CACHE
Structure and replacement policy guarantee all to reach preferable improved efficiency in different deformation angles.Pass through caching source images coordinate simultaneously
It is cached on piece, main process modules is executed parallel, further improve processing speed.This method can accomplish needle
Real-time processing to the image three-dimensional transformation of the above resolution ratio of full HD, avoids and writes the storage of region of memory bring again
The waste of bandwidth.
Embodiment two
As shown in fig. 6, in the present embodiment, a kind of tripleplane's mapping device, including：
Module 10 is obtained, for obtaining targeted graphical region, obtains appointing in the targeted graphical region by point by point scanning
The coordinate of a bit；
Computing module 20 calculates the seat of corresponding source images for the coordinate according to any point in the targeted graphical region
Mark；
Preserving module 30, for saving the coordinate of the source images to cache memory；
Conversion module 40, for obtaining the coordinate of the source images from the cache memory, and by the source figure
The coordinate of picture is converted to the physical address of source images；
Interpolating module 50 is inserted for carrying out interpolation processing according to the coordinate of the physical address and the source images
It is worth pixel；
Output module 60, for being exported according to the displaing coordinate of current output image.
In the present embodiment, source images coordinate is cached by cache memory CACHE to cache on piece, makes winner
It wants processing module that can execute parallel, improves processing speed；Accomplish that the image three-dimensional for the above resolution ratio of full HD converts
Real-time processing, avoid the waste for writing region of memory bring memory bandwidth again.
In the present embodiment, its primitive meaning of cache memory CACHE refers to that access speed is accessed than general random
A kind of RAM that memory body (RAM) comes fastly, it is however generally that it uses DRAM technology unlike system memory main body, and uses
Expensive but SRAM technology faster, also there is the title of cache memory.Cache memory be present in main memory and CPU it
Between single-level memory, be made of static storage chip (SRAM), capacity is smaller but speed is more much higher than main memory, close to
The speed of CPU.It is that the high speed between central processing unit and main memory is small in the hierarchical structure of computer memory system
Capacity memory.It constitutes the memory of level-one with main memory together.Information between cache memory and main memory
Scheduling and transmission be to be carried out automatically by hardware.
As shown in Fig. 2, the function that the present embodiment is mainly realized is that source images A is obtained target area by 3D deformation process
Targeted graphical region B on D, when making observer to observe image perpendicular to the angle of target area D, 3D effect is presented in image.
In the present embodiment, the quadrangle that targeted graphical region is made of 4 vertex b0, b1, b2, b3, therefore
The coordinate of point in the case where known 4 apex coordinates on any side of quadrangle can be calculated by two point form linear equation
Come, so as to obtain the coordinate profiles of quadrangle, by the b0b2b3 in Fig. 2, this two sides are known as left profile for convenience of description,
This two sides b0b1b3 are known as right wheel exterior feature.
As shown in fig. 7, in the present embodiment, the acquisition module includes：
Profile acquiring unit 11 calculates the profile in the targeted graphical region for obtaining targeted graphical region；
Scanning element 12, for carrying out point by point scanning according to preset spacing to the image in the profile；
Coordinate acquiring unit 13, for obtaining the coordinate of any point in the targeted graphical region.
In the present embodiment, as shown in figure 4, the direction y is motionless using b0 as starting point (x0, y0), by the incremental direction x0+ of x
Δ x, x0+2 Δ x, x0+3 Δ x ..., x0+k Δ x scanning, the right wheel exterior feature until encountering figure terminate current row scanning.Then it jumps
Start to scan to next line, the origin coordinates of next line is chosen from the left profile of image, and coordinate is (x0+k Δ y, y0+ Δ y)
Wherein (x0, y0) is the origin coordinates of lastrow, and the spacing of two rows is Δ y, and k is the slope of left profile, is swept by the direction that x is incremented by
It retouches, the right wheel exterior feature until encountering figure terminates.The every a line of multiple scanning is until figure last line just terminates like this.
As shown in figure 8, in the present embodiment, the computing module includes：
Vertex acquiring unit 21, the vertex for the apex coordinate and the source images that obtain the targeted graphical region are sat
Mark；
Coefficient calculation unit 22, for calculating the coordinate of any point in the targeted graphical region according to the apex coordinate
With the coefficient matrix between the coordinate of corresponding source images；
The coordinate of source images is calculated according to the following formula：
U=(m11*x+m12*y+m13)/(m31*x+m32*y+m33)
V=(m21*x+m22*y+m23)/(m31*x+m32*y+m33)
Wherein, (u, v) be source images coordinate, (x, y) be targeted graphical region coordinate, m11, m12, m13, m21,
M22, m23, m31, m32, m33 are coefficient matrix.
In the present embodiment, the calculating of source images coordinate carries out simultaneously with scanning, puts beginning source figure from (x0, the y0) on D
As the changing rule of coordinate is as follows：
The every increase Δ x of the upper x coordinate of D, the value of molecule adds up m11* Δ x in above formula, and denominator value adds up m31* Δ x；
The every increase Δ y of the upper y-coordinate of D, the value of molecule adds up m12* Δ y in above formula, and denominator value adds up m32* Δ y；
The coordinate in corresponding source images A can be calculated by formula (molecule is divided by denominator), calculated coordinate has
Integer part and fractional part, wherein for taking a use, fractional part uses integer part for interpolation.
In the present embodiment, the conversion module includes：
Judge that the cache memory whether there is the coordinate of required source images, if so, directly acquiring described
The coordinate of source images；
Otherwise, update it is one whole data cached in the cache memory, from update it is described it is data cached in
Obtain the coordinate of the source images；
The coordinate of the source images is converted to the physical address of source images.
In the present embodiment, by the physical address for being converted into source images of source images, image A is obtained from the buffer area CACHE
Data.The buffer area CACHE is for caching the source image data obtained from memory.If required pixel data is present in currently
In CACHE, then this data is directly returned to from CACHE, if pixel data is needed to be not present in current CACHE, it is whole to update one
CACHE line.According to practical application request, CACHE can be configured to one or more, per all the way correspond to a TAG table with
Record the start physical address of source images corresponding to each CACHE line in this road.The length of CACHE can be according to image
Tilt angle be configured to storage fixed number pixel.The number of pixels that can be stored in every CACHE line is the same.
The length m and CACHE of number n, the CACHE line of the size K and CACHE of on-chip memory need to meet per depth t all the way
Following formula：
K>=n*m*t
As shown in figure 5, being the structural schematic diagram of cache memory CACHE.
Because target image is lateral point by point scanning, the reading of corresponding source images has certain local correlations.Work as target
When image inclination angle is smaller, the corresponding lateral correlation of original image reading is stronger, and CACHE line length is longer to be promoted
Efficiency.When target image tilt angle is larger, it is stronger that corresponding original image reads longitudinal correlation, and usually SDRAM is only
Can by row laterally read, therefore CACHE line length is shorter can raising efficiency.The design can be according to the inclination angle of target image
The length m of number n, the CACHE line of reasonable disposition CACHE is spent with CACHE per depth t all the way, to guarantee on-chip memory
Make full use of.
In the present embodiment, the output module includes：
Judge that the displaing coordinate of current output image whether in the targeted graphical region, is inserted if so, output is described
It is worth pixel；
Otherwise, background color is exported.
In the present embodiment, the CACHE structure being connected using group optimize graph image projection transformation when due to reading source
The low problem of access SDR/DDR SDRAM access efficiency brought by image slices vegetarian refreshments is discontinuous.It is tied by flexible CACHE
Structure and replacement policy guarantee all to reach preferable improved efficiency in different deformation angles.Pass through caching source images coordinate simultaneously
It is cached on piece, main process modules is executed parallel, further improve processing speed.This method can accomplish needle
Real-time processing to the image three-dimensional transformation of the above resolution ratio of full HD, avoids and writes the storage of region of memory bring again
The waste of bandwidth.
It should be noted that, in this document, the terms "include", "comprise" or its any other variant are intended to non-row
His property includes, so that the process, method, article or the device that include a series of elements not only include those elements, and
And further include other elements that are not explicitly listed, or further include for this process, method, article or device institute it is intrinsic
Element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including being somebody's turn to do
There is also other identical elements in the process, method of element, article or device.
The serial number of the above embodiments of the invention is only for description, does not represent the advantages or disadvantages of the embodiments.
The above is only a preferred embodiment of the present invention, is not intended to limit the scope of the invention, all to utilize this hair
Equivalent structure or equivalent flow shift made by bright specification and accompanying drawing content is applied directly or indirectly in other relevant skills
Art field, is included within the scope of the present invention.
Claims (10)
1. a kind of tripleplane's mapping method, which is characterized in that including：
Targeted graphical region is obtained, the coordinate of any point in the targeted graphical region is obtained by point by point scanning；
The coordinate of corresponding source images is calculated according to the coordinate of any point in the targeted graphical region；
The coordinate of the source images is saved to cache memory；
The coordinate of the source images is obtained from the cache memory, and the coordinate of the source images is converted into source images
Physical address；
Interpolation processing is carried out according to the coordinate of the physical address and the source images, obtains interpolating pixel；
It is exported according to the displaing coordinate of current output image.
2. a kind of tripleplane's mapping method according to claim 1, which is characterized in that the acquisition targeted graphical area
Domain includes by the coordinate that point by point scanning obtains any point in the targeted graphical region：
Targeted graphical region is obtained, the profile in the targeted graphical region is calculated；
Point by point scanning is carried out according to preset spacing to the image in the profile；
Obtain the coordinate of any point in the targeted graphical region.
3. a kind of tripleplane's mapping method according to claim 2, which is characterized in that described according to the targeted graphical
The coordinate of the corresponding source images of coordinate calculating of any point includes in region：
Obtain the apex coordinate in the targeted graphical region and the apex coordinate of the source images；
It is calculated in the targeted graphical region between the coordinate of any point and the coordinate of corresponding source images according to the apex coordinate
Coefficient matrix；
The coordinate of source images is calculated according to the following formula：
U=(m11*x+m12*y+m13)/(m31*x+m32*y+m33)
V=(m21*x+m22*y+m23)/(m31*x+m32*y+m33)
Wherein, (u, v) be source images coordinate, (x, y) be targeted graphical region coordinate, m11, m12, m13, m21, m22,
M23, m31, m32, m33 are coefficient matrix.
4. a kind of tripleplane's mapping method according to claim 3, which is characterized in that described to be deposited from the speed buffering
Reservoir obtains the coordinate of the source images, and includes by the physical address that the coordinate of the source images is converted to source images：
Judge that the cache memory whether there is the coordinate of required source images, if so, directly acquiring the source figure
The coordinate of picture；
Otherwise, update it is one whole data cached in the cache memory, from the data cached middle acquisition of update
The coordinate of the source images；
The coordinate of the source images is converted to the physical address of source images.
5. a kind of tripleplane's mapping method according to claim 4, which is characterized in that described according to current output image
Displaing coordinate carry out output include：
The displaing coordinate of current output image is judged whether in the targeted graphical region, if so, exporting the interpolation picture
Element；
Otherwise, background color is exported.
6. a kind of tripleplane's mapping device, which is characterized in that including：
It obtains module and any point in the targeted graphical region is obtained by point by point scanning for obtaining targeted graphical region
Coordinate；
Computing module calculates the coordinate of corresponding source images for the coordinate according to any point in the targeted graphical region；
Preserving module, for saving the coordinate of the source images to cache memory；
Conversion module, for obtaining the coordinate of the source images from the cache memory, and by the seat of the source images
Mark is converted to the physical address of source images；
Interpolating module obtains interpolating pixel for carrying out interpolation processing according to the coordinate of the physical address and the source images；
Output module, for being exported according to the displaing coordinate of current output image.
7. a kind of tripleplane's mapping device according to claim 6, which is characterized in that the acquisition module includes：
Profile acquiring unit calculates the profile in the targeted graphical region for obtaining targeted graphical region；
Scanning element, for carrying out point by point scanning according to preset spacing to the image in the profile；
Coordinate acquiring unit, for obtaining the coordinate of any point in the targeted graphical region.
8. a kind of tripleplane's mapping device according to claim 7, which is characterized in that the computing module includes：
Vertex acquiring unit, for obtaining the apex coordinate in the targeted graphical region and the apex coordinate of the source images；
Coefficient calculation unit, for according to the apex coordinate calculate in the targeted graphical region coordinate of any point with it is corresponding
Coefficient matrix between the coordinate of source images；
The coordinate of source images is calculated according to the following formula：
U=(m11*x+m12*y+m13)/(m31*x+m32*y+m33)
V=(m21*x+m22*y+m23)/(m31*x+m32*y+m33)
Wherein, (u, v) be source images coordinate, (x, y) be targeted graphical region coordinate, m11, m12, m13, m21, m22,
M23, m31, m32, m33 are coefficient matrix.
9. a kind of tripleplane's mapping device according to claim 8, which is characterized in that the conversion module includes：
Judge that the cache memory whether there is the coordinate of required source images, if so, directly acquiring the source figure
The coordinate of picture；
Otherwise, update it is one whole data cached in the cache memory, from the data cached middle acquisition of update
The coordinate of the source images；
The coordinate of the source images is converted to the physical address of source images.
10. a kind of tripleplane's mapping device according to claim 9, which is characterized in that the output module includes：
The displaing coordinate of current output image is judged whether in the targeted graphical region, if so, exporting the interpolation picture
Element；
Otherwise, background color is exported.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810771103.6A CN108921779A (en) | 2018-07-13 | 2018-07-13 | A kind of tripleplane's mapping method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810771103.6A CN108921779A (en) | 2018-07-13 | 2018-07-13 | A kind of tripleplane's mapping method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108921779A true CN108921779A (en) | 2018-11-30 |
Family
ID=64412740
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810771103.6A Pending CN108921779A (en) | 2018-07-13 | 2018-07-13 | A kind of tripleplane's mapping method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108921779A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11772276B2 (en) | 2020-01-02 | 2023-10-03 | Universal City Studios Llc | Systems and methods for optical performance captured animated figure with real-time reactive projected media |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1942901A (en) * | 2005-02-23 | 2007-04-04 | 索尼计算机娱乐公司 | Image processing device and image processing method |
| CN101257626A (en) * | 2008-01-31 | 2008-09-03 | 炬力集成电路设计有限公司 | Method, apparatus for access DRAM and medium player |
| US20140297192A1 (en) * | 2013-03-26 | 2014-10-02 | Cgg Services Sa | System and method for interpolating seismic data by matching pursuit in fourier transform |
| CN104796649A (en) * | 2014-01-21 | 2015-07-22 | 北京炬力北方微电子有限公司 | 3D projection method and device |
| CN104935831A (en) * | 2015-06-12 | 2015-09-23 | 中国科学院自动化研究所 | Parallel multiphase image interpolation device and method |
-
2018
- 2018-07-13 CN CN201810771103.6A patent/CN108921779A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1942901A (en) * | 2005-02-23 | 2007-04-04 | 索尼计算机娱乐公司 | Image processing device and image processing method |
| CN101257626A (en) * | 2008-01-31 | 2008-09-03 | 炬力集成电路设计有限公司 | Method, apparatus for access DRAM and medium player |
| US20140297192A1 (en) * | 2013-03-26 | 2014-10-02 | Cgg Services Sa | System and method for interpolating seismic data by matching pursuit in fourier transform |
| CN104796649A (en) * | 2014-01-21 | 2015-07-22 | 北京炬力北方微电子有限公司 | 3D projection method and device |
| CN104935831A (en) * | 2015-06-12 | 2015-09-23 | 中国科学院自动化研究所 | Parallel multiphase image interpolation device and method |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11772276B2 (en) | 2020-01-02 | 2023-10-03 | Universal City Studios Llc | Systems and methods for optical performance captured animated figure with real-time reactive projected media |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP0998709B1 (en) | Non-stalled requesting texture cache system and method | |
| CN102460503B (en) | Apparatus and method for displaying warped version of source image | |
| EP1066600B1 (en) | Block- and band-oriented traversal in three-dimensional triangle rendering | |
| EP0613098B1 (en) | Image processing apparatus and method of controlling the same | |
| DE69736611T2 (en) | Reorganization of memory references for pixels in a page-oriented memory architecture | |
| US6674443B1 (en) | Memory system for accelerating graphics operations within an electronic device | |
| CN110390382B (en) | Convolutional neural network hardware accelerator with novel feature map caching module | |
| US10726522B2 (en) | Method and system for correcting a distorted input image | |
| CN102890427B (en) | Method for preparing skewed data in field programmable gate array (FPGA) of direct-writing type photoetching system | |
| CN110958362A (en) | Image correction system and method based on block table look-up | |
| CN104796649B (en) | A kind of method and device of tripleplane | |
| CN108921779A (en) | A kind of tripleplane's mapping method and device | |
| GB2269081A (en) | Image processing apparatus | |
| KR20010012841A (en) | Image processor and image processing method | |
| CN109671127A (en) | A kind of method and device for realizing waveform drawing | |
| US6577776B1 (en) | Transforming video images | |
| CN106952215B (en) | Image pyramid feature extraction circuit, device and method | |
| CN116312412A (en) | Bicubic interpolation circuit based on FPGA | |
| CN109685715B (en) | Stepless rotation processing method for high-precision image in display system | |
| JP4408152B2 (en) | Texture mapping method and texture mapping apparatus | |
| US9245370B2 (en) | Image drawing apparatus, computer-readable medium storing program, and method of the same | |
| US20080036775A1 (en) | Method for generating characters | |
| CN106570827B (en) | The method and device for carrying out image rotation is converted using memory space address | |
| CN111696024B (en) | Infrared image mirroring method and device based on FPGA | |
| US7769247B2 (en) | Method and apparatus for data re-arrangement |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: No. 701, 7th floor, No. 10 Huayuan East Road, Haidian District, Beijing 100083 Applicant after: Beijing Juli North Microelectronics Co.,Ltd. Address before: 100083 701, 7 / F, Gaode building, 10 Huayuan East Road, Haidian District, Beijing Applicant before: BEIJING ACTIONS MICROELECTRONICS Co.,Ltd. |
|
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |