CN105894479A - Image filtering method and image filtering device - Google Patents
Image filtering method and image filtering device Download PDFInfo
- Publication number
- CN105894479A CN105894479A CN201610490476.7A CN201610490476A CN105894479A CN 105894479 A CN105894479 A CN 105894479A CN 201610490476 A CN201610490476 A CN 201610490476A CN 105894479 A CN105894479 A CN 105894479A
- Authority
- CN
- China
- Prior art keywords
- pixel
- value
- list item
- filtering
- item value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000001914 filtration Methods 0.000 title claims abstract description 287
- 238000000034 method Methods 0.000 title claims abstract description 173
- 238000004148 unit process Methods 0.000 claims 1
- 238000010586 diagram Methods 0.000 description 8
- 238000004590 computer program Methods 0.000 description 5
- 238000003860 storage Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The invention discloses an image filtering method and an image filtering device. The method comprises the following steps of: in the first period, performing the first filtering treatment on the first pixel; in the second period, performing the first filtering treatment on the second pixel, and performing the second filtering treatment on the first pixel subjected to the first filtering treatment; and in the third period, performing the first filtering treatment on the third pixel, performing the second filtering treatment on the second pixel subjected to the first filtering, and performing the third filtering treatment on the first pixel subjected to the second filtering. In the method, the filtering processes of adjacent pixels can be performed gradually, thus the efficiency of the flow process of the whole filtering process is greatly improved, and the hardware cost is saved.
Description
Technical field
The present invention relates to image processing field, particularly relate to a kind of image filtering method and device.
Background technology
Under Low light intensity, the video sequence of camera shooting often contains the strongest noise signal,
The quality of video sequence is affected greatly.At present, generally use the wave filter of smothing filtering to video
Noise in sequence processes.For video sequence, on time-space domain, the distribution of noise meets
Gaussian Profile.When the pixel needing two field picture each to video is filtered, existing filtering method
Not only the cycle is very long, and hardware spending is big.
Summary of the invention
For this reason, it may be necessary to provide the technical scheme of a kind of image filtering, in order to solve existing filtering method week
The problems such as the phase is very long, hardware spending is big.
For achieving the above object, inventor providing a kind of image filtering method, described method includes as follows
Step:
Within the period 1, the first pixel is carried out the first filtering process;
Within second round, the second pixel is carried out the first filtering process, and processes through the first filtering
The first pixel carry out the second filtering process, described second pixel is for adjacent with the first pixel and be positioned at first
On the right of pixel or following pixel;
Within the period 3, the 3rd pixel is carried out the first filtering process, to through the second of the first filtering
Pixel carries out the second filtering process, the first pixel through the second filtering is carried out the 3rd filtering and processes;Institute
State the 3rd pixel to be adjacent with the second pixel and be positioned on the right of the second pixel or following pixel;When described
When second pixel is edge pixel point, the next line that described 3rd pixel is expert at by the second pixel
First pixel, or be the second pixel column the right one row first pixel.
Further, described first is filtered into and filters from left to right, and the most described second pixel is and the first picture
Element is adjacent and is positioned at the pixel on the right of the first pixel, described " the first pixel carries out the first filtering process "
Including: calculate the first pixel and the pixel value difference of the adjacent leftmost pixel point of the first pixel, and according to
Pixel value difference and the corresponding relation of list item value, obtain the list item value that this pixel value difference is corresponding, by the first pixel
The pixel value of point is added with acquired list item value or subtracts each other.
Further, described " carry out the pixel value of the first pixel with acquired list item value being added or
Subtract each other " specifically include: when the pixel value of the first pixel is more than the leftmost pixel adjacent with the first pixel
During the pixel value put, the pixel value of the first pixel is subtracted each other with acquired list item value;When the first pixel
When the pixel value of point is less than the pixel value of the leftmost pixel point adjacent with the first pixel, by the first pixel
Pixel value be added with acquired list item value.
Further, described second is filtered into and filters from top to bottom, and the most described 3rd pixel is and second
Pixel is adjacent and is positioned at the pixel on the right of the second pixel, when described second pixel is edge pixel point,
First pixel of the next line that described 3rd pixel is expert at by the second pixel;Described " to
Two pixels carry out the first filtering process " including: calculate the first pixel and the pixel value difference of the second pixel,
And according to the corresponding relation of pixel value difference Yu list item value, obtain the list item value that this pixel value difference is corresponding, by
The pixel value of two pixels is added with acquired list item value or subtracts each other.
Further, described " the first pixel processed through the first filtering is carried out the second filtering process "
Including: calculate the first pixel and the adjacent top pixel of the first pixel processed through the first filtering
Pixel value difference, and according to the corresponding relation of pixel value difference Yu list item value, obtain this pixel value difference corresponding
List item value, is carried out the pixel value of the first pixel processed through the first filtering with acquired list item value
It is added or subtracts each other.
Further, described first is filtered into and filters from top to bottom, and the most described second pixel is and the first picture
Element is adjacent and is positioned at the pixel that the first pixel is following, described " the first pixel carries out the first filtering process "
Including: calculate the first pixel and the pixel value difference of the adjacent top pixel of the first pixel, and according to
Pixel value difference and the corresponding relation of list item value, obtain the list item value that this pixel value difference is corresponding, by the first pixel
The pixel value of point is added with acquired list item value or subtracts each other.
Further, described " carry out the pixel value of the first pixel with acquired list item value being added or
Subtract each other " specifically include: when the pixel value of the first pixel is more than the top pixel adjacent with the first pixel
During the pixel value put, the pixel value of the first pixel is subtracted each other with acquired list item value;When the first pixel
When the pixel value of point is less than the pixel value of the top pixel adjacent with the first pixel, by the first pixel
Pixel value be added with acquired list item value.
Further, described second is filtered into and filters from left to right, and the most described 3rd pixel is and second
Pixel is adjacent and is positioned at the pixel that the second pixel is following, when described second pixel is edge pixel point,
Described 3rd pixel is first pixel of the right one row of the second pixel column;Described " right
Second pixel carries out the first filtering process " including: the pixel calculating the first pixel and the second pixel is poor
Value, and according to the corresponding relation of pixel value difference Yu list item value, obtain the list item value that this pixel value difference is corresponding,
The pixel value of the second pixel is added with acquired list item value or subtracts each other.
Further, described " the first pixel processed through the first filtering is carried out the second filtering process "
Including: calculate the first pixel and the adjacent leftmost pixel point of the first pixel processed through the first filtering
Pixel value difference, and according to the corresponding relation of pixel value difference Yu list item value, obtain this pixel value difference corresponding
List item value, is carried out the pixel value of the first pixel processed through the first filtering with acquired list item value
It is added or subtracts each other.
Further, described " to through second filtering the first pixel carry out the 3rd filtering process " including:
The pixel calculating the first pixel pixel identical with the first pixel coordinate position in previous frame image is poor
Value, and according to the corresponding relation of pixel value difference Yu list item value, obtain the list item value that this pixel value difference is corresponding,
Carry out being added or phase with acquired list item value by the pixel value of the first pixel processed through the first filtering
Subtract.
Inventor additionally provides a kind of image filtering device, described device include the first filter unit, second
Filter unit and the 3rd filter unit;
Within the period 1, the first filter unit for carrying out the first filtering process to the first pixel;
Within second round, the first filter unit for the second pixel being carried out the first filtering process, second
Filter unit for the first pixel processed through the first filtering is carried out the second filtering process, described second
Pixel is adjacent with the first pixel and is positioned on the right of the first pixel or following pixel;
Within the period 3, the first filter unit for the 3rd pixel being carried out the first filtering process, second
Filter unit is for carrying out the second filtering process, the 3rd filter unit to the second pixel through the first filtering
Process for the first pixel through the second filtering is carried out the 3rd filtering;Described 3rd pixel is and second
Pixel is adjacent and is positioned on the right of the second pixel or following pixel;When described second pixel is edge picture
During vegetarian refreshments, first pixel of the next line that described 3rd pixel is expert at by the second pixel, or
Person is first pixel of the right one row of the second pixel column.
Further, described first is filtered into and filters from left to right, and the most described second pixel is and the first picture
Element is adjacent and is positioned at the pixel on the right of the first pixel, and described first filter unit includes the first pixel value difference
Computing module, the first list item value acquisition module, the first list item value computing module;It is described that " the first filtering is single
Unit by the first pixel being carried out the first filtering process " including: the first pixel value difference computing module by based on
Calculating the first pixel and the pixel value difference of the adjacent leftmost pixel point of the first pixel, the first list item value obtains
Module, for the corresponding relation according to pixel value difference Yu list item value, obtains the list item value that this pixel value difference is corresponding,
First list item value computing module is for being added the pixel value of the first pixel with acquired list item value
Or subtract each other.
It is further, described that " the first list item value computing module is for by the pixel value of the first pixel and institute
The list item value obtained is added or subtracts each other " specifically include: when the pixel value of the first pixel is more than with the
During the pixel value of the leftmost pixel point that one pixel is adjacent, the first list item value computing module is for by the first picture
The pixel value of vegetarian refreshments subtracts each other with acquired list item value;When the pixel value of the first pixel is less than and the first picture
During the pixel value of the leftmost pixel point that vegetarian refreshments is adjacent, the first list item value computing module is for by the first pixel
Pixel value be added with acquired list item value.
Further, described second is filtered into and filters from top to bottom, and the most described 3rd pixel is and second
Pixel is adjacent and is positioned at the pixel on the right of the second pixel, when described second pixel is edge pixel point,
First pixel of the next line that described 3rd pixel is expert at by the second pixel;Described " first
Filter unit for carrying out the first filtering process to the second pixel " including: the first pixel value difference computing module
For calculating the first pixel and the pixel value difference of the second pixel, the first list item value acquisition module is used for root
According to the corresponding relation of pixel value difference Yu list item value, obtain the list item value that this pixel value difference is corresponding, the first list item
Value computing module is for being added by the pixel value of the second pixel with acquired list item value or subtract each other.
Further, described second filter unit includes the second pixel value difference computing module, the second list item value
Acquisition module, the second list item value computing module;It is described that " the second filter unit is for through the first filtering
The first pixel processed carries out the second filtering process " including: the second pixel value difference computing module is used for calculating
The first pixel processed through the first filtering and the pixel of the adjacent top pixel of the first pixel are poor
Value, the second list item value acquisition module, for the corresponding relation according to pixel value difference Yu list item value, obtains this picture
The list item value that element difference is corresponding, the second list item value computing module is used for first will processed through the first filtering
The pixel value of pixel is added with acquired list item value or subtracts each other.
Further, described first is filtered into and filters from top to bottom, and the most described second pixel is and the first picture
Element is adjacent and is positioned at the pixel that the first pixel is following, and described first filter unit includes the first pixel value difference
Computing module, the first list item value acquisition module, the first list item value computing module;It is described that " the first filtering is single
Unit by the first pixel being carried out the first filtering process " including: the first pixel value difference computing module by based on
Calculating the first pixel and the pixel value difference of the adjacent top pixel of the first pixel, the first list item value obtains
Module, for the corresponding relation according to pixel value difference Yu list item value, obtains the list item value that this pixel value difference is corresponding,
First list item value computing module is for being added the pixel value of the first pixel with acquired list item value
Or subtract each other.
It is further, described that " the first list item value computing module is for by the pixel value of the first pixel and institute
The list item value obtained is added or subtracts each other " specifically include: when the pixel value of the first pixel is more than with the
During the pixel value of the top pixel that one pixel is adjacent, the first list item value computing module is for by the first picture
The pixel value of vegetarian refreshments subtracts each other with acquired list item value;When the pixel value of the first pixel is less than and the first picture
During the pixel value of the top pixel that vegetarian refreshments is adjacent, the first list item value computing module is for by the first pixel
Pixel value be added with acquired list item value.
Further, described second is filtered into and filters from left to right, and the most described 3rd pixel is and second
Pixel is adjacent and is positioned at the pixel that the second pixel is following, when described second pixel is edge pixel point,
Described 3rd pixel is first pixel of the right one row of the second pixel column;Described " the
One filter unit for carrying out the first filtering process to the second pixel " including: the first pixel value difference calculates mould
Block is for calculating the first pixel and the pixel value difference of the second pixel, and the first list item value acquisition module is used for
According to the corresponding relation of pixel value difference Yu list item value, obtain the list item value that this pixel value difference is corresponding, the first table
Entry value computing module is for carrying out being added or phase with acquired list item value by the pixel value of the second pixel
Subtract.
Further, described second filter unit includes the second pixel value difference computing module, the second list item value
Acquisition module, the second list item value computing module;It is described that " the second filter unit is for through the first filtering
The first pixel processed carries out the second filtering process " including: the second pixel value difference computing module is used for calculating
The first pixel processed through the first filtering and the pixel of the adjacent leftmost pixel point of the first pixel are poor
Value, the second list item value acquisition module, for the corresponding relation according to pixel value difference Yu list item value, obtains this picture
The list item value that element difference is corresponding, the second list item value computing module is used for first will processed through the first filtering
The pixel value of pixel is added with acquired list item value or subtracts each other.
Further, described 3rd filter unit includes the 3rd pixel value difference computing module, the 3rd list item value
Acquisition module, the 3rd list item value computing module;It is described that " the 3rd filter unit is for through the second filtering
The first pixel carry out the 3rd filtering process " including: the 3rd pixel value difference computing module be used for calculating first
The pixel value difference of the pixel that pixel is identical with the first pixel coordinate position in previous frame image, the 3rd
List item value acquisition module, for the corresponding relation according to pixel value difference Yu list item value, obtains this pixel value difference pair
The list item value answered, the 3rd list item value computing module is used for will be through the first pixel of the first filtering process
Pixel value is added with acquired list item value or subtracts each other.
Image filtering method described in technique scheme and device, described method comprises the steps:
In period 1, the first pixel is carried out the first filtering process;Within second round, the second pixel is entered
Row the first filtering processes, and the first pixel processed through the first filtering is carried out the second filtering process;?
In period 3, the 3rd pixel is carried out the first filtering process, the second pixel through the first filtering is entered
Row the second filtering processes, and the first pixel through the second filtering is carried out the 3rd filtering and processes.Said method
The filtering making each pixel adjacent can point echelon be carried out, and then makes whole filtering flowing water
Operation, substantially increases efficiency, saves hardware spending.
Accompanying drawing explanation
Fig. 1 is the flow chart of the image filtering method that one embodiment of the invention relates to;
Fig. 2 is the flow chart of the image filtering method that another embodiment of the present invention relates to;
Fig. 3 is the schematic diagram of the image filtering method that another embodiment of the present invention relates to;
Fig. 4 is the schematic diagram of the image filtering device that another embodiment of the present invention relates to.
Description of reference numerals:
101, the first filter unit;111, the first pixel value difference computing module;112, the second list item value obtains
Delivery block;113, the second list item value computing module;
102, the second filter unit;121, the second pixel value difference computing module;122, the second list item value obtains
Delivery block;123, the second list item value computing module;
103, the 3rd filter unit.
Detailed description of the invention
By describing the technology contents of technical scheme, structural feature in detail, being realized purpose and effect, below
In conjunction with specific embodiments and coordinate accompanying drawing to be explained in detail.
Refer to Fig. 1, for the flow chart of the image filtering method that one embodiment of the invention relates to.Described method
May be used for being filtered image processing, eliminate picture noise.Described method comprises the steps:
Initially enter step S101 within the period 1, the first pixel is carried out the first filtering process.For
For each two field picture of video, need to filter the most from left to right, filter two from the top down
Process, both can be adjusted according to actual needs order, i.e. can also filter the most from the top down,
Filter the most from left to right.In the present embodiment, described first it is filtered into and filters from left to right, then
Described second pixel is adjacent with the first pixel and is positioned at the pixel on the right of the first pixel.
As in figure 2 it is shown, described " the first pixel carries out the first filtering process " including: enter step S201
Calculate the first pixel and the pixel value difference of the adjacent leftmost pixel point of the first pixel, then enter step
S202, according to the corresponding relation of pixel value difference Yu list item value, obtains the list item value that this pixel value difference is corresponding, and
The pixel value of the first pixel is added with acquired list item value or subtracts each other by rear entrance step S203.
Specifically, filtering can be realized by below equation:
Wherein, x (k) is pixel the most to be filtered, and y (k-1) is the pixel after filtering processes, such as
First is filtered into and filters from left to right, and y (k-1) is then the pixel on the pixel left side the most to be filtered, Gamma
Value is index parameters, can configure according to actual needs.Owing to above-mentioned formula relates to exponent arithmetic,
If calculating not only algorithm complexity with hardware to too increase hardware spending, thus use in the present invention and look into
The form of table realizes, and i.e. chooses different pixel value differences according to gradient, and different pixels computed in advance is poor
The list item value that value is the most corresponding, and list item value is stored in form, described list item value is on the right of formula
The calculated value of remaining part in addition to x (k).In order to ensure the heart and immediate vicinity in the picture
Processing accuracy during pixel, the list item precision that will be close to image center increases, and i.e. pixel value difference is whole
When fractional part is between 0~7, use 3bit decimal essence;When pixel value difference integer part is between 8~15,
Use 2bit decimal precision;When pixel value difference integer part is between 16~31, use 1bit decimal precision;
When pixel value difference integer part is between 31~63, only with integer part.
In the present embodiment, step S203 specifically includes: when the pixel value of the first pixel is more than with the
During the pixel value of the leftmost pixel point that one pixel is adjacent, by the pixel value of the first pixel with acquired
List item value is subtracted each other;When the pixel value of the first pixel is less than the leftmost pixel point adjacent with the first pixel
During pixel value, the pixel value of the first pixel is added with acquired list item value.The purpose that filtering processes
To remove picture noise, i.e. allow on image the pixel value of neighbor pixel as close possible to.First filter
When the mode of ripple is for filtering from left to right, the leftmost pixel point that the first pixel is adjacent is filtered
Pixel, thus when the pixel value of the first pixel is more than the leftmost pixel point adjacent with the first pixel
During pixel value, in order to the pixel value that makes the first pixel and the pixel filtered pixel value more
Close, need to subtract each other the pixel value of the first pixel with acquired list item value.Based on similar reason,
When the pixel value of the first pixel is less than the pixel value of the leftmost pixel point adjacent with the first pixel, will
The pixel value of the first pixel is added with acquired list item value.
Then entrance step S102 is within second round, and the second pixel carries out the first filtering process, and right
The first pixel processed through the first filtering carries out the second filtering process.Described second pixel is and the first picture
Element is adjacent and is positioned on the right of the first pixel or following pixel, specifically, when first is filtered into from left-hand
During right filtering, the second pixel is the pixel on the right of the first pixel;When first is filtered into and filters down,
Second pixel is the pixel below the first pixel.It is filtered into filtering from left to right, the second filter with first
As a example by ripple is filtering from top to bottom, then described " the second pixel is carried out the first filtering process " including: meter
Calculate the first pixel and the pixel value difference of the second pixel, and close according to pixel value difference is corresponding with list item value
System, obtains the list item value that this pixel value difference is corresponding, by the pixel value of the second pixel and acquired list item
Value is added or subtracts each other.Second pixel when carrying out the first filtering (from left to right filtering), the first picture
Element has completed the first filtering, and i.e. for the second pixel, the first pixel is the most filtered pixel,
The process that second pixel carries out the first filtering is similar with the process that aforementioned first pixel carries out the first filtering.
Described " the first pixel processed through the first filtering is carried out the second filtering process " including: calculates
The first pixel processed through the first filtering and the pixel of the adjacent top pixel of the first pixel are poor
Value, and according to the corresponding relation of pixel value difference Yu list item value, obtain the list item value that this pixel value difference is corresponding,
Carry out being added or phase with acquired list item value by the pixel value of the first pixel processed through the first filtering
Subtract.Second filtering is processed as filtering from top to bottom, and i.e. adjacent with the first pixel top pixel is
Pixel after processing after filtering.
Then entrance step S103 is within the period 3, the 3rd pixel is carried out the first filtering process, to warp
The second pixel crossing the first filtering carries out the second filtering process, carries out the first pixel through the second filtering
3rd filtering processes.Described 3rd pixel is for adjacent with the second pixel and be positioned on the right of the second pixel or below
Pixel;When described second pixel is edge pixel point, described 3rd pixel is the second pixel
First pixel of the next line that point is expert at, or be the right one row of the second pixel column
First pixel.When first be filtered into from left to right filtering, second be filtered into filter from top to bottom time,
3rd pixel is adjacent with the second pixel and is positioned at the pixel on the right of the second pixel, i.e. relative to
For three pixels, when carrying out the first filtering, the second pixel is pixel the most after filtering, when
When described second pixel is edge pixel point, under described 3rd pixel is expert at by the second pixel
First pixel of a line;When first being filtered into filtering from top to bottom, second being filtered into and filter from left to right
During ripple, the 3rd pixel is adjacent with the second pixel and is positioned in the following pixel of the second pixel, i.e. phase
For the 3rd pixel, when carrying out the first filtering, the second pixel is pixel the most after filtering
Point, first pixel of the next line that described 3rd pixel is expert at by the second pixel, or be
First pixel of the right one row of the second pixel column.
In the present embodiment, described " the first pixel through the second filtering being carried out the 3rd filtering process "
Including: calculate the first pixel pixel identical with the first pixel coordinate position in previous frame image
Pixel value difference, and according to the corresponding relation of pixel value difference Yu list item value, obtain the table that this pixel value difference is corresponding
Entry value, carries out phase by the pixel value of the first pixel processed through the first filtering with acquired list item value
Add or subtract each other., the pixel on each two field picture is except carrying out left and right filtering with neighbor pixel and filtering up and down
Outside ripple, in addition it is also necessary to carry out the 3rd filtering, i.e. by calculating and the first pixel seat in former frame image
The pixel value difference of the pixel that cursor position is identical, and the mode tabled look-up obtains the list item value of correspondence, will be from
Body pixel value adds or deducts this list item value.
In other embodiments, described first is filtered into and filters from top to bottom, the most described second pixel be with
First pixel is adjacent and is positioned at the pixel that the first pixel is following, described " the first pixel is carried out the first filter
Ripple processes " including: calculate the first pixel and the pixel value difference of the adjacent top pixel of the first pixel,
And according to the corresponding relation of pixel value difference Yu list item value, obtain the list item value that this pixel value difference is corresponding, by
The pixel value of one pixel is added with acquired list item value or subtracts each other.Described " by the first pixel
Pixel value be added with acquired list item value or subtract each other " specifically include: when the picture of the first pixel
When element value is more than the pixel value of the top pixel adjacent with the first pixel, by the pixel of the first pixel
Value is subtracted each other with acquired list item value;When the pixel value of the first pixel is less than adjacent with the first pixel
During the pixel value of top pixel, the pixel value of the first pixel is added with acquired list item value.
In other embodiments, described second is filtered into and filters from left to right, and the most described 3rd pixel is
Adjacent with the second pixel and be positioned at the pixel that the second pixel is following, when described second pixel is edge picture
During vegetarian refreshments, described 3rd pixel is first pixel of the right one row of the second pixel column;
Described " the second pixel carries out the first filtering process " including: calculates the first pixel and the second pixel
Pixel value difference, and according to the corresponding relation of pixel value difference Yu list item value, obtain this pixel value difference corresponding
List item value, is added by the pixel value of the second pixel with acquired list item value or subtracts each other.Described " right
The first pixel processed through the first filtering carries out the second filtering process " including: calculate through the first filtering
The first pixel processed and the pixel value difference of the adjacent leftmost pixel point of the first pixel, and according to pixel
Difference and the corresponding relation of list item value, obtain the list item value that this pixel value difference is corresponding, will be through the first filtering
The pixel value of the first pixel processed is added with acquired list item value or subtracts each other.
As it is shown on figure 3, include three 9 pixels of row three row, it is filtered into first below and filters from top to bottom
Ripple, second be filtered into filtering from left to right as a example by, to the first of 3 pixels of first row the filtering and second
Filtering illustrates, specific as follows:
In the t0 moment (i.e. period 1), longitudinally filtering A0_0 pixel (the i.e. first pixel), obtains A0_0
Longitudinal filter result (i.e. through the first loudness of the first filtering) of pixel;
In the t1 moment (i.e. second round), A1_0 (the i.e. second pixel) pixel utilizes the vertical of A0_0 pixel
To filter result, carry out longitudinal filtering;A0_0 pixel (the i.e. first pixel) was carried out laterally in this t1 moment
Filtering (filters, the second filtering) the most from left to right;
In the t2 moment (i.e. in the period 3), A1_0 pixel (the second pixel) laterally filters (
Two filtering), A2_0 pixel (the 3rd pixel) utilizes longitudinal filter result of A1_0 pixel to carry out longitudinal filter
Ripple.
Utilize this stair-stepping filtering relation, can be filtered multiple pixels well processing, pole
Improve greatly filtration efficiency, save hardware area, reduce hardware cost.
As shown in Figure 4, inventor additionally provides a kind of image filtering device, and described device includes the first filter
Ripple unit the 101, second filter unit 102 and the 3rd filter unit 103;
Within the period 1, the first filter unit 101 for carrying out the first filtering process to the first pixel.
For each two field picture of video, need to filter the most from left to right, filter from the top down
Two processes, both can be adjusted according to actual needs order, i.e. can also first carry out from the top down
Filtering, then filter from left to right.In the present embodiment, described first it is filtered into and filters from left to right
Ripple, the most described second pixel is adjacent with the first pixel and is positioned at the pixel on the right of the first pixel.
Described first is filtered into and filters from left to right, and the most described second pixel is adjacent with the first pixel and position
Pixel on the right of the first pixel, described first filter unit includes the first pixel value difference computing module
111, first list item value acquisition module the 112, first list item value computing module 113;It is described that " the first filtering is single
Unit for carrying out the first filtering process to the first pixel " including: the first pixel value difference computing module 111 is used
In calculating the first pixel and the pixel value difference of the adjacent leftmost pixel point of the first pixel, the first list item value
Acquisition module 112, for the corresponding relation according to pixel value difference Yu list item value, obtains this pixel value difference corresponding
List item value, the first list item value computing module 113 for by the pixel value of the first pixel with acquired
List item value is added or subtracts each other.Specifically, filtering can be realized by below equation:
Wherein, x (k) is pixel the most to be filtered, and y (k-1) is the pixel after filtering processes, such as
First is filtered into and filters from left to right, and y (k-1) is then the pixel on the pixel left side the most to be filtered, Gamma
Value is index parameters, can configure according to actual needs.Owing to above-mentioned formula relates to exponent arithmetic,
If calculating not only algorithm complexity with hardware to too increase hardware spending, thus use in the present invention and look into
The form of table realizes, and i.e. chooses different pixel value differences according to gradient, and different pixels computed in advance is poor
The list item value that value is the most corresponding, and list item value is stored in form, described list item value is on the right of formula
The calculated value of remaining part in addition to x (k).In order to ensure the heart and immediate vicinity in the picture
Processing accuracy during pixel, the list item precision that will be close to image center increases, and i.e. pixel value difference is whole
When fractional part is between 0~7, use 3bit decimal essence;When pixel value difference integer part is between 8~15,
Use 2bit decimal precision;When pixel value difference integer part is between 16~31, use 1bit decimal precision;
When pixel value difference integer part is between 31~63, only with integer part.
It is in the present embodiment, described that " the first list item value computing module is for by the pixel of the first pixel
Value is added with acquired list item value or subtracts each other " specifically include: when the pixel value of the first pixel is big
When the pixel value of the leftmost pixel point adjacent with the first pixel, the first list item value computing module is used for will
The pixel value of the first pixel subtracts each other with acquired list item value;When the first pixel pixel value less than with
During the pixel value of the leftmost pixel point that the first pixel is adjacent, the first list item value computing module is for by first
The pixel value of pixel is added with acquired list item value.The purpose that filtering processes is to make an uproar to remove image
Sound, i.e. allow on image the pixel value of neighbor pixel as close possible to.The mode of the first filtering is from left-hand
During right filtering, the leftmost pixel point that the first pixel is adjacent is the pixel filtered, thus when the
When the pixel value of one pixel is more than the pixel value of the leftmost pixel point adjacent with the first pixel, in order to make
The pixel value of the first pixel is more nearly with the pixel value of the pixel filtered, and needs first
The pixel value of pixel subtracts each other with acquired list item value.Based on similar reason, when the picture of the first pixel
When element value is less than the pixel value of the leftmost pixel point adjacent with the first pixel, by the pixel of the first pixel
Value is added with acquired list item value.
Within second round, the first filter unit 102 is used for carrying out the second pixel the first filtering process,
Second filter unit 102 the first pixel being used for processing through the first filtering carries out the second filtering process,
Described second pixel is adjacent with the first pixel and is positioned on the right of the first pixel or following pixel.Specifically
Ground, when first is filtered into and filters from left to right, the second pixel is the pixel on the right of the first pixel;When
First is filtered into when filtering down, and the second pixel is the pixel below the first pixel.With the first filter
Ripple is filtering from left to right, second be filtered into filtering from top to bottom as a example by, then described " the first filter unit
For the second pixel being carried out the first filtering process " including: the first pixel value difference computing module 111 calculates
First pixel and the pixel value difference of the second pixel, the first list item value acquisition module 112 is poor according to pixel
Value and the corresponding relation of list item value, obtain the list item value that this pixel value difference is corresponding, and the first list item value calculates mould
The pixel value of the second pixel is added with acquired list item value or subtracts each other by block 113.Second pixel
When carrying out the first filtering (from left to right filtering), the first pixel has completed the first filtering, i.e. relative to
For second pixel, the first pixel is the most filtered pixel, and the second pixel carries out the mistake of the first filtering
The process that journey carries out the first filtering with aforementioned first pixel is similar.
Described second filter unit includes second pixel value difference computing module the 121, second list item value acquisition module
122, the second list item value computing module 123;It is described that " the second filter unit is at the first filtering
First pixel of reason carries out the second filtering process " including: the second pixel value difference computing module 121 calculates warp
Cross the first pixel and the pixel value difference of the adjacent top pixel of the first pixel that the first filtering processes,
Second list item value acquisition module 122, according to the corresponding relation of pixel value difference Yu list item value, obtains this pixel poor
The list item value that value is corresponding, the first pixel that the second list item value computing module 123 will process through the first filtering
The pixel value of point is added with acquired list item value or subtracts each other.Second filtering is processed as filtering from top to bottom
Ripple, i.e. adjacent with the first pixel top pixel is the pixel after processing the most after filtering.
Within the period 3, the first filter unit 101 is used for the 3rd pixel is carried out the first filtering process,
Second filter unit 102 for the second pixel through the first filtering being carried out the second filtering process, the 3rd
Filter unit 103 processes for the first pixel through the second filtering is carried out the 3rd filtering;Described 3rd
Pixel is adjacent with the second pixel and is positioned on the right of the second pixel or following pixel;When described second picture
When vegetarian refreshments is edge pixel point, the first of the next line that described 3rd pixel is expert at by the second pixel
Individual pixel, or be first pixel of the right one row of the second pixel column.When the first filter
Ripple for from left to right filtering, second be filtered into filter from top to bottom time, the 3rd pixel is and the second pixel
Adjacent and be positioned at the pixel on the right of the second pixel, i.e. for the 3rd pixel, carrying out
During one filtering, the second pixel is pixel the most after filtering, when described second pixel is edge picture
During vegetarian refreshments, first pixel of the next line that described 3rd pixel is expert at by the second pixel;When
First being filtered into filtering from top to bottom, second be filtered into when filtering from left to right, the 3rd pixel is and the
Two pixels are adjacent and are positioned in the following pixel of the second pixel, i.e. for the 3rd pixel,
When carrying out the first filtering, the second pixel is pixel the most after filtering, and described 3rd pixel is
First pixel of the next line that two pixels are expert at, or be the right of the second pixel column
First pixel of one row.
It is in the present embodiment, described that " the 3rd filter unit is for the first pixel through the second filtering
Carry out the 3rd filtering to process " including: calculate the first pixel and the first pixel coordinate in previous frame image
The pixel value difference of the pixel that position is identical, and according to the corresponding relation of pixel value difference Yu list item value, obtain
The list item value that this pixel value difference is corresponding, by pixel value and the institute of the first pixel through the first filtering process
The list item value obtained is added or subtracts each other.Pixel on each two field picture is except clicking on neighbor
Filtering and up and down outside filtering about row, in addition it is also necessary to carry out the 3rd filtering, i.e. by calculating and former frame the
The pixel value difference of the pixel that the first pixel coordinate position is identical in image, and the mode tabled look-up obtains
Corresponding list item value, adds self pixel value or deducts this list item value.
In other embodiments, described first is filtered into and filters from top to bottom, the most described second pixel be with
First pixel is adjacent and is positioned at the pixel that the first pixel is following, described, and " the first filter unit is for the
One pixel carries out the first filtering process " including: the first pixel value difference computing module 111 calculates the first pixel
The pixel value difference of the top pixel that point is adjacent with the first pixel, the second list item value acquisition module 112
According to the corresponding relation of pixel value difference Yu list item value, obtain the list item value that this pixel value difference is corresponding, the first list item
The pixel value of the first pixel is added with acquired list item value or subtracts each other by value computing module 123.
It is described that " pixel value of the first pixel is entered by the first list item value computing module 123 with acquired list item value
Row is added or subtracts each other " specifically include: when the pixel value of the first pixel is more than adjacent with the first pixel
During the pixel value of top pixel, the pixel value of the first pixel is subtracted each other with acquired list item value;When
When the pixel value of the first pixel is less than the pixel value of the top pixel adjacent with the first pixel, by the
The pixel value of one pixel is added with acquired list item value.
In other embodiments, described second is filtered into and filters from left to right, and the most described 3rd pixel is
Adjacent with the second pixel and be positioned at the pixel that the second pixel is following, when described second pixel is edge picture
During vegetarian refreshments, described 3rd pixel is first pixel of the right one row of the second pixel column;
Described " the first filter unit for carrying out the first filtering process to the second pixel " including: the first pixel is poor
Value computing module 111 calculates the first pixel and the pixel value difference of the second pixel, and the second list item value obtains
Module 112, according to the corresponding relation of pixel value difference Yu list item value, obtains the list item value that this pixel value difference is corresponding,
The pixel value of the second pixel is added by the first list item value computing module 123 with acquired list item value
Or subtract each other.Described second filter unit includes that second pixel value difference computing module the 121, second list item value obtains
Module the 122, second list item value computing module 123.It is described that " the second filter unit is for through the first filter
The first pixel that ripple processes carries out the second filtering process " including: the second pixel value difference computing module 121 is counted
The pixel calculating the first pixel and the adjacent leftmost pixel point of the first pixel processed through the first filtering is poor
Value, the second list item value acquisition module 122, according to the corresponding relation of pixel value difference Yu list item value, obtains this picture
The list item value that element difference is corresponding, the second list item value computing module 123 will be through the first of the first filtering process
The pixel value of pixel is added with acquired list item value or subtracts each other.
Image filtering method described in technique scheme and device, described method comprises the steps:
In period 1, the first pixel is carried out the first filtering process;Within second round, the second pixel is entered
Row the first filtering processes, and the first pixel processed through the first filtering is carried out the second filtering process;?
In period 3, the 3rd pixel is carried out the first filtering process, the second pixel through the first filtering is entered
Row the second filtering processes, and the first pixel through the second filtering is carried out the 3rd filtering and processes.Said method
The filtering making each pixel adjacent can point echelon be carried out, and then makes whole filtering flowing water
Operation, substantially increases efficiency, saves hardware spending.
It should be noted that in this article, the relational terms of such as first and second or the like is used merely to
One entity or operation are separated with another entity or operating space, and not necessarily requires or imply
Relation or the order of any this reality is there is between these entities or operation.And, term " includes ",
" comprise " or its any other variant is intended to comprising of nonexcludability, so that include that one is
The process of row key element, method, article or terminal device not only include those key elements, but also include not
There are other key elements being expressly recited, or also include setting for this process, method, article or terminal
Standby intrinsic key element.In the case of there is no more restriction, by statement " including ... " or " comprising ... "
The key element limited, it is not excluded that in including the process of described key element, method, article or terminal device
There is also other key element.Additionally, in this article, " being more than ", " being less than ", " exceeding " etc. are interpreted as not
Including this number;More than " ", " below ", " within " etc. be interpreted as including this number.
Those skilled in the art are it should be appreciated that the various embodiments described above can be provided as method, device or meter
Calculation machine program product.These embodiments can use complete hardware embodiment, complete software implementation or knot
The form of the embodiment in terms of conjunction software and hardware.Whole or portion in the method that the various embodiments described above relate to
Can instruct relevant hardware by program step by step to complete, described program can be stored in calculating
In the storage medium that machine equipment can read, all or part of for perform described in the various embodiments described above method
Step.Described computer equipment, includes but not limited to: personal computer, server, all-purpose computer,
Special-purpose computer, the network equipment, embedded device, programmable device, intelligent mobile terminal, intelligence man
Occupy equipment, wearable intelligent equipment, vehicle intelligent equipment etc.;Described storage medium, including but do not limit
In: RAM, ROM, magnetic disc, tape, CD, flash memory, USB flash disk, portable hard drive, storage card, note
Recall rod, webserver storage, network cloud storage etc..
The various embodiments described above are with reference to according to the method described in embodiment, equipment (system) and computer program
The flow chart of product and/or block diagram describe.It should be understood that flow process can be realized by computer program instructions
Stream in each flow process in figure and/or block diagram and/or square frame and flow chart and/or block diagram
Journey and/or the combination of square frame.Can provide these computer program instructions to computer equipment processor with
Produce a machine so that the instruction performed by the processor of computer equipment is produced for realizing at stream
The function specified in one flow process of journey figure or multiple flow process and/or one square frame of block diagram or multiple square frame
Device.
These computer program instructions may be alternatively stored in the meter that computer equipment can be guided to work in a specific way
Calculate in machine equipment readable memory so that the instruction being stored in this computer equipment readable memory produces
Including the manufacture of command device, this command device realize one flow process of flow chart or multiple flow process and/
Or the function specified in one square frame of block diagram or multiple square frame.
These computer program instructions also can be loaded on computer equipment so that holds on a computing device
Row sequence of operations step is to produce computer implemented process, thus the finger performed on a computing device
Order provides for realizing at one flow process of flow chart or multiple flow process and/or one square frame or multiple of block diagram
The step of the function specified in square frame.
Although being described the various embodiments described above, but those skilled in the art once learn
Basic creative concept, then can make other change and amendment to these embodiments, so above institute
Stating only embodiments of the invention, not thereby limit the scope of patent protection of the present invention, every utilization is originally
Equivalent structure or equivalence flow process that description of the invention and accompanying drawing content are made convert, or directly or indirectly use
In the technical field that other are relevant, within being the most in like manner included in the scope of patent protection of the present invention.
Claims (20)
1. an image filtering method, it is characterised in that described method comprises the steps:
Within the period 1, the first pixel is carried out the first filtering process;
Within second round, the second pixel is carried out the first filtering process, and processes through the first filtering
The first pixel carry out the second filtering process, described second pixel is for adjacent with the first pixel and be positioned at first
On the right of pixel or following pixel;
Within the period 3, the 3rd pixel is carried out the first filtering process, to through the second of the first filtering
Pixel carries out the second filtering process, the first pixel through the second filtering is carried out the 3rd filtering and processes;Institute
State the 3rd pixel to be adjacent with the second pixel and be positioned on the right of the second pixel or following pixel;When described
When second pixel is edge pixel point, the next line that described 3rd pixel is expert at by the second pixel
First pixel, or be the second pixel column the right one row first pixel.
Image filtering method the most according to claim 1, it is characterised in that described first is filtered into
Filtering from left to right, the most described second pixel is adjacent with the first pixel and is positioned at the picture on the right of the first pixel
Vegetarian refreshments, described " the first pixel carries out the first filtering process " including: calculates the first pixel and first
The pixel value difference of the leftmost pixel point that pixel is adjacent, and according to the corresponding relation of pixel value difference Yu list item value,
Obtain the list item value that this pixel value difference is corresponding, the pixel value of the first pixel is entered with acquired list item value
Row is added or subtracts each other.
Image filtering method the most according to claim 2, it is characterised in that described " by the first picture
The pixel value of vegetarian refreshments is added with acquired list item value or subtracts each other " specifically include: when the first pixel
Pixel value more than the pixel value of the leftmost pixel point adjacent with the first pixel time, by the first pixel
Pixel value subtracts each other with acquired list item value;When the pixel value of the first pixel is less than and the first pixel phase
During the pixel value of adjacent leftmost pixel point, the pixel value of the first pixel is added with acquired list item value.
4. according to the image filtering method described in Claims 2 or 3, it is characterised in that described second filter
Ripple is for filter from top to bottom, and the most described 3rd pixel is adjacent with the second pixel and is positioned at the second pixel right side
The pixel on limit, when described second pixel is edge pixel point, described 3rd pixel is the second picture
First pixel of the next line that vegetarian refreshments is expert at;Described " the second pixel is carried out the first filtering process "
Including: calculate the first pixel and the pixel value difference of the second pixel, and according to pixel value difference and list item value
Corresponding relation, obtain the list item value that this pixel value difference is corresponding, by the pixel value of the second pixel with obtained
The list item value taken is added or subtracts each other.
Image filtering method the most according to claim 4, it is characterised in that described " to through
The first pixel that one filtering processes carries out the second filtering process " including: calculate through the first filtering process
First pixel and the pixel value difference of the adjacent top pixel of the first pixel, and according to pixel value difference with
The corresponding relation of list item value, obtains the list item value that this pixel value difference is corresponding, by process through the first filtering
The pixel value of the first pixel is added with acquired list item value or subtracts each other.
Image filtering method the most according to claim 1, it is characterised in that described first is filtered into
Filtering from top to bottom, the most described second pixel is adjacent with the first pixel and is positioned at the picture that the first pixel is following
Vegetarian refreshments, described " the first pixel carries out the first filtering process " including: calculates the first pixel and first
The pixel value difference of the top pixel that pixel is adjacent, and according to the corresponding relation of pixel value difference Yu list item value,
Obtain the list item value that this pixel value difference is corresponding, the pixel value of the first pixel is entered with acquired list item value
Row is added or subtracts each other.
Image filtering method the most according to claim 6, it is characterised in that described " by the first picture
The pixel value of vegetarian refreshments is added with acquired list item value or subtracts each other " specifically include: when the first pixel
Pixel value more than the pixel value of the top pixel adjacent with the first pixel time, by the first pixel
Pixel value subtracts each other with acquired list item value;When the pixel value of the first pixel is less than and the first pixel phase
During the pixel value of adjacent top pixel, the pixel value of the first pixel is added with acquired list item value.
8. according to the image filtering method described in claim 6 or 7, it is characterised in that described second filter
Ripple is for filter from left to right, and the most described 3rd pixel is adjacent with the second pixel and is positioned under the second pixel
The pixel on limit, when described second pixel is edge pixel point, described 3rd pixel is the second picture
First pixel of the right one row of vegetarian refreshments column;Described " the second pixel is carried out at the first filtering
Reason " including: calculate the first pixel and the pixel value difference of the second pixel, and according to pixel value difference and table
The corresponding relation of entry value, obtains the list item value that this pixel value difference is corresponding, by the pixel value of the second pixel with
Acquired list item value is added or subtracts each other.
Image filtering method the most according to claim 4, it is characterised in that described " to through
The first pixel that one filtering processes carries out the second filtering process " including: calculate through the first filtering process
First pixel and the pixel value difference of the adjacent leftmost pixel point of the first pixel, and according to pixel value difference with
The corresponding relation of list item value, obtains the list item value that this pixel value difference is corresponding, by process through the first filtering
The pixel value of the first pixel is added with acquired list item value or subtracts each other.
Image filtering method the most according to claim 1, it is characterised in that described " to process
First pixel of the second filtering carries out the 3rd filtering and processes " including: calculate the first pixel and previous frame figure
The pixel value difference of the pixel that the first pixel coordinate position is identical in Xiang, and according to pixel value difference and list item
The corresponding relation of value, obtains the list item value that this pixel value difference is corresponding, first will processed through the first filtering
The pixel value of pixel is added with acquired list item value or subtracts each other.
11. 1 kinds of image filtering devices, it is characterised in that described device include the first filter unit,
Two filter units and the 3rd filter unit;
Within the period 1, the first filter unit for carrying out the first filtering process to the first pixel;
Within second round, the first filter unit for the second pixel being carried out the first filtering process, second
Filter unit for the first pixel processed through the first filtering is carried out the second filtering process, described second
Pixel is adjacent with the first pixel and is positioned on the right of the first pixel or following pixel;
Within the period 3, the first filter unit for the 3rd pixel being carried out the first filtering process, second
Filter unit is for carrying out the second filtering process, the 3rd filter unit to the second pixel through the first filtering
Process for the first pixel through the second filtering is carried out the 3rd filtering;Described 3rd pixel is and second
Pixel is adjacent and is positioned on the right of the second pixel or following pixel;When described second pixel is edge picture
During vegetarian refreshments, first pixel of the next line that described 3rd pixel is expert at by the second pixel, or
Person is first pixel of the right one row of the second pixel column.
12. image filtering devices according to claim 11, it is characterised in that described first filtering
For filtering from left to right, the most described second pixel is adjacent with the first pixel and is positioned on the right of the first pixel
Pixel, described first filter unit includes that the first pixel value difference computing module, the first list item value obtain mould
Block, the first list item value computing module;It is described that " the first filter unit is for carrying out the first filter to the first pixel
Ripple processes " including: the first pixel value difference computing module is for calculating the first pixel and the first pixel phase
The pixel value difference of adjacent leftmost pixel point, the first list item value acquisition module is for according to pixel value difference and list item
The corresponding relation of value, obtains the list item value that this pixel value difference is corresponding, and the first list item value computing module is used for will
The pixel value of the first pixel is added with acquired list item value or subtracts each other.
13. image filtering devices according to claim 12, it is characterised in that described " the first table
Entry value computing module is for carrying out being added or phase with acquired list item value by the pixel value of the first pixel
Subtract " specifically include: when the pixel value of the first pixel is more than the leftmost pixel point adjacent with the first pixel
Pixel value time, the first list item value computing module is for by the pixel value of the first pixel and acquired table
Entry value is subtracted each other;When the pixel value of the first pixel is less than the picture of the leftmost pixel point adjacent with the first pixel
During element value, the first list item value computing module is for by the pixel value of the first pixel and acquired list item value
It is added.
14. according to the image filtering device described in claim 12 or 13, it is characterised in that described
Two are filtered into and filter from top to bottom, and the most described 3rd pixel is for adjacent with the second pixel and be positioned at the second picture
Pixel on the right of element, when described second pixel is edge pixel point, described 3rd pixel is
First pixel of the next line that two pixels are expert at;It is described that " the first filter unit is for second
Pixel carries out the first filtering process " including: the first pixel value difference computing module is for calculating the first pixel
With the pixel value difference of the second pixel, the first list item value acquisition module is for according to pixel value difference and list item value
Corresponding relation, obtain the list item value that this pixel value difference is corresponding, the first list item value computing module is for by the
The pixel value of two pixels is added with acquired list item value or subtracts each other.
15. image filtering devices according to claim 14, it is characterised in that described second filtering
Unit includes that the second pixel value difference computing module, the second list item value acquisition module, the second list item value calculate mould
Block;It is described that " the second filter unit is for carrying out the second filtering to the first pixel processed through the first filtering
Process " including: the second pixel value difference computing module is for calculating the first pixel processed through the first filtering
The pixel value difference of the top pixel that point is adjacent with the first pixel, the second list item value acquisition module is used for root
According to the corresponding relation of pixel value difference Yu list item value, obtain the list item value that this pixel value difference is corresponding, the second list item
Value computing module is used for will be through the pixel value of the first pixel of the first filtering process and acquired list item
Value is added or subtracts each other.
16. image filtering devices according to claim 11, it is characterised in that described first filtering
For filtering from top to bottom, the most described second pixel is adjacent with the first pixel and to be positioned at the first pixel following
Pixel, described first filter unit includes that the first pixel value difference computing module, the first list item value obtain mould
Block, the first list item value computing module;It is described that " the first filter unit is for carrying out the first filter to the first pixel
Ripple processes " including: the first pixel value difference computing module is for calculating the first pixel and the first pixel phase
The pixel value difference of adjacent top pixel, the first list item value acquisition module is for according to pixel value difference and list item
The corresponding relation of value, obtains the list item value that this pixel value difference is corresponding, and the first list item value computing module is used for will
The pixel value of the first pixel is added with acquired list item value or subtracts each other.
17. image filtering devices according to claim 16, it is characterised in that described " the first table
Entry value computing module is for carrying out being added or phase with acquired list item value by the pixel value of the first pixel
Subtract " specifically include: when the pixel value of the first pixel is more than the top pixel adjacent with the first pixel
Pixel value time, the first list item value computing module is for by the pixel value of the first pixel and acquired table
Entry value is subtracted each other;When the pixel value of the first pixel is less than the picture of the top pixel adjacent with the first pixel
During element value, the first list item value computing module is for by the pixel value of the first pixel and acquired list item value
It is added.
18. according to the image filtering device described in claim 16 or 17, it is characterised in that described
Two are filtered into and filter from left to right, and the most described 3rd pixel is for adjacent with the second pixel and be positioned at the second picture
The following pixel of element, when described second pixel is edge pixel point, described 3rd pixel is the
First pixel of the right one row of two pixel columns;It is described that " the first filter unit is for the
Two pixels carry out the first filtering process " including: the first pixel value difference computing module is for calculating the first pixel
Putting the pixel value difference with the second pixel, the first list item value acquisition module is for according to pixel value difference and list item
The corresponding relation of value, obtains the list item value that this pixel value difference is corresponding, and the first list item value computing module is used for will
The pixel value of the second pixel is added with acquired list item value or subtracts each other.
19. image filtering devices according to claim 14, it is characterised in that described second filtering
Unit includes that the second pixel value difference computing module, the second list item value acquisition module, the second list item value calculate mould
Block;It is described that " the second filter unit is for carrying out the second filtering to the first pixel processed through the first filtering
Process " including: the second pixel value difference computing module is for calculating the first pixel processed through the first filtering
The pixel value difference of the leftmost pixel point that point is adjacent with the first pixel, the second list item value acquisition module is used for root
According to the corresponding relation of pixel value difference Yu list item value, obtain the list item value that this pixel value difference is corresponding, the second list item
Value computing module is used for will be through the pixel value of the first pixel of the first filtering process and acquired list item
Value is added or subtracts each other.
20. image filtering devices according to claim 11, it is characterised in that described 3rd filtering
Unit includes that the 3rd pixel value difference computing module, the 3rd list item value acquisition module, the 3rd list item value calculate mould
Block;Described " the 3rd filter unit processes for the first pixel through the second filtering is carried out the 3rd filtering "
Including: the 3rd pixel value difference computing module is for calculating the first pixel and the first pixel in previous frame image
The pixel value difference of the pixel that point coordinates position is identical, the 3rd list item value acquisition module is for poor according to pixel
Value and the corresponding relation of list item value, obtain the list item value that this pixel value difference is corresponding, and the 3rd list item value calculates mould
Block is for carrying out phase by the pixel value of the first pixel processed through the first filtering with acquired list item value
Add or subtract each other.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610490476.7A CN105894479B (en) | 2016-06-28 | 2016-06-28 | A kind of image filtering method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610490476.7A CN105894479B (en) | 2016-06-28 | 2016-06-28 | A kind of image filtering method and device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105894479A true CN105894479A (en) | 2016-08-24 |
CN105894479B CN105894479B (en) | 2018-08-31 |
Family
ID=56719342
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610490476.7A Active CN105894479B (en) | 2016-06-28 | 2016-06-28 | A kind of image filtering method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105894479B (en) |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6526181B1 (en) * | 1998-01-27 | 2003-02-25 | Eastman Kodak Company | Apparatus and method for eliminating imaging sensor line noise |
JP2005229491A (en) * | 2004-02-16 | 2005-08-25 | Matsushita Electric Ind Co Ltd | Method for removing noises from digital images |
CN101739670A (en) * | 2009-11-27 | 2010-06-16 | 西安电子科技大学 | Non-local mean space domain time varying image filtering method |
CN102289798A (en) * | 2011-08-18 | 2011-12-21 | 青岛海信信芯科技有限公司 | method and device for reducing image ringing noise |
CN103606132A (en) * | 2013-10-31 | 2014-02-26 | 西安电子科技大学 | Multiframe digital image denoising method based on space domain and time domain combination filtering |
CN104050646A (en) * | 2014-07-04 | 2014-09-17 | 西安电子科技大学 | Method for restraining image impulse noise based on iteration non-local mean value |
CN104809705A (en) * | 2015-04-29 | 2015-07-29 | 厦门美图之家科技有限公司 | Image denoising method and system based on threshold value block matching |
-
2016
- 2016-06-28 CN CN201610490476.7A patent/CN105894479B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6526181B1 (en) * | 1998-01-27 | 2003-02-25 | Eastman Kodak Company | Apparatus and method for eliminating imaging sensor line noise |
JP2005229491A (en) * | 2004-02-16 | 2005-08-25 | Matsushita Electric Ind Co Ltd | Method for removing noises from digital images |
CN101739670A (en) * | 2009-11-27 | 2010-06-16 | 西安电子科技大学 | Non-local mean space domain time varying image filtering method |
CN102289798A (en) * | 2011-08-18 | 2011-12-21 | 青岛海信信芯科技有限公司 | method and device for reducing image ringing noise |
CN103606132A (en) * | 2013-10-31 | 2014-02-26 | 西安电子科技大学 | Multiframe digital image denoising method based on space domain and time domain combination filtering |
CN104050646A (en) * | 2014-07-04 | 2014-09-17 | 西安电子科技大学 | Method for restraining image impulse noise based on iteration non-local mean value |
CN104809705A (en) * | 2015-04-29 | 2015-07-29 | 厦门美图之家科技有限公司 | Image denoising method and system based on threshold value block matching |
Non-Patent Citations (5)
Title |
---|
BUADES A 等: "Nonlocal image and movie denoising", 《COMPUTER VISION》 * |
RABBANI H 等: "Speckle noise reduction of medical ultrasound images in complex wavelet domain using mixture priors", 《BIOMEDICAL ENGINEERING》 * |
权利 等: "十字基自适应指导滤波在图像降噪中的应用", 《计算机应用》 * |
王常虹 等: "使用中值-各向异性扩散的超声图像去噪算法", 《哈尔滨工程大学学报》 * |
陆广地: "基于改进数学形态学滤波的农产品图像处理", 《江苏农业科学》 * |
Also Published As
Publication number | Publication date |
---|---|
CN105894479B (en) | 2018-08-31 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
EP3542525B1 (en) | Methods and systems for filling light for video images | |
US20220415044A1 (en) | Method, apparatus, medium and device for extracting river drying-up region and frequency | |
US20210097651A1 (en) | Image processing method and apparatus, electronic device, and storage medium | |
CN109416727A (en) | Glasses minimizing technology and device in a kind of facial image | |
CN104517265A (en) | Intelligent buffing method and intelligent buffing device | |
US10832382B2 (en) | Method for filtering spurious pixels in a depth-map | |
CN109785246A (en) | A kind of noise-reduction method of non-local mean filtering, device and equipment | |
CN112001923B (en) | Retina image segmentation method and device | |
CN103208094B (en) | For the method and system to image applications wave filter | |
CN104658030A (en) | Secondary image mixing method and apparatus | |
CN113240578B (en) | Image special effect generation method and device, electronic equipment and storage medium | |
CN109461125A (en) | Point 2-d gaussian filters device and image processing method based on FPGA | |
CN109284761A (en) | A kind of image characteristic extracting method, device, equipment and readable storage medium storing program for executing | |
CN105894450A (en) | Image processing method and device | |
KR20160066425A (en) | Method and apparatus for blurring an image | |
CN103871035A (en) | Image denoising method and device | |
CN103020908A (en) | Image denoising method and device | |
CN105894479A (en) | Image filtering method and image filtering device | |
CN105989579A (en) | Image enhancement method and device | |
CN109388311A (en) | A kind of image display method, device and equipment | |
CN106485666B (en) | A kind of information indicating method and apparatus | |
CN109376627A (en) | A kind of method for detecting human face based on deep learning, device and equipment | |
US8737759B2 (en) | Image blurring by partitioning a non-separable fir filter | |
CN108062773B (en) | Image processing method and device and robot | |
CN110930290A (en) | Data processing method and device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CP01 | Change in the name or title of a patent holder |
Address after: 350003 building, No. 89, software Avenue, Gulou District, Fujian, Fuzhou 18, China Patentee after: Ruixin Microelectronics Co., Ltd Address before: 350003 building, No. 89, software Avenue, Gulou District, Fujian, Fuzhou 18, China Patentee before: Fuzhou Rockchips Electronics Co.,Ltd. |
|
CP01 | Change in the name or title of a patent holder |