CN101145336A - Image processing method and scaling system - Google Patents

Image processing method and scaling system Download PDF

Info

Publication number
CN101145336A
CN101145336A CNA2006101272210A CN200610127221A CN101145336A CN 101145336 A CN101145336 A CN 101145336A CN A2006101272210 A CNA2006101272210 A CN A2006101272210A CN 200610127221 A CN200610127221 A CN 200610127221A CN 101145336 A CN101145336 A CN 101145336A
Authority
CN
China
Prior art keywords
value
pixel
pixel value
resolution
composing
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
Application number
CNA2006101272210A
Other languages
Chinese (zh)
Other versions
CN100541604C (en
Inventor
贾惠芬
唐英原
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Princeton Technology Corp
Original Assignee
Princeton Technology Corp
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Princeton Technology Corp filed Critical Princeton Technology Corp
Priority to CNB2006101272210A priority Critical patent/CN100541604C/en
Publication of CN101145336A publication Critical patent/CN101145336A/en
Application granted granted Critical
Publication of CN100541604C publication Critical patent/CN100541604C/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Image Processing (AREA)
  • Editing Of Facsimile Originals (AREA)

Abstract

The present invention provides an image processing method and a zooming system which are used for zooming an image from a first resolution to a second resolution, wherein, P<SUB>T</SUB>, the pixel value of a target pixel T under the second resolution, consists of a certain number (k) of P<SUB>i</SUB>, the pixel value of an original pixel i under the first resolution. The number k is equal to or larger than one, and the number i refers to an integer larger than or equal to one and smaller than or equal to k. The method comprises the following procedures of first, finding out a composition proportion R<SUB>i</SUB> corresponding to the pixel value of each original pixel i, P<SUB>i</SUB>; then performing certain calculation on the composition proportion R<SUB>i</SUB> corresponding to P<SUB>i</SUB>, the pixel value of each original pixel i, to obtain a corresponding weighted value W<SUB>i</SUB>, wherein, the W<SUB>i</SUB> is an integer between 1 and 2<SUP>n</SUP>, and the sum of W<SUB>i</SUB> is 2<SUP>n</SUP>; and finally performing calculation on P<SUB>i</SUB>, the pixel value of each original pixel i, and the corresponding weighted value W<SUB>i</SUB> to produce the pixel value of the target pixel T, P<SUB>T</SUB> and complete the image processing. The present invention can reduce the complexity and the cost of hardware.

Description

Image treatment method and panntographic system
Technical field
The invention relates to a kind of image treatment method and system, particularly can be used for dwindling or the image treatment method and the system of magnified image resolution (resolution) relevant for a kind of.
Background technology
In the image processing, in order to meet some special-purpose, the dwindling or be enlarged into widely used means of image (image).The dwindling of image (image) resolution is meant image is converted to the demonstration of a low resolution by the demonstration of a high-resolution that the amplification of image then is meant image is shown the demonstration that is converted to a high-resolution by a lower resolution.When original image is represented with the image of new resolution different pixels point, because image resolution changes, the pixel value of each pixel is made up of according to different proportion of composing one in the original image pixel value to a plurality of pixels in the then new image.For instance, suppose that pixel value that the pixel value of the point of A ' in the new image order by original A, B is formed, 1/3 or the A point of the pixel value that 2/3 and B of the pixel value that the pixel value of then new A ' point may be ordered by A ordered and the pixel value that B is ordered respectively account for 1/2 and form.
A kind of being used for dwindled or the known skill of amplifying is to utilize linear interpolation an image resolution, is that the pixel value with a plurality of pixels makes up according to different proportion of composing, to produce the pixel value of the pixel in the new image.For instance, as shown in Figure 1, when supposing that desire uses 1 dimensional linear interpolation method to become at 300 by 500 resolution image, original pixel value PA, PB, PC, PD and the PE of ordering of A, B, C, D and the E in 500 images can produce A ', B ' in 300 new images and pixel value PA ', PB ' and the PC ' of C ' point according to following ratio:
PA &prime; = 300 &times; PA + 200 &times; PB 500 , . . . . . . . . . . . . . . . . . . . . . . . ( 1 )
PB &prime; = 100 &times; PB + 300 &times; PC + 100 &times; PD 500 , . . . . . . . . . . . ( 2 )
PC &prime; = 200 &times; PD + 300 &times; PE 500 , . . . . . . . . . . . . . . . . . . . . . ( 3 )
Wherein, 200 and 300 in (1) formula represented the middle A point of pixel value PA ' of A ' point and the proportion of composing RA and the RB of B point pixel value respectively.Simultaneously, by (1) to (3) formula as can be known, the proportion of composing of A to E each point pixel value depends on pixel resolution fully, and all is no more than 300, that is is no more than the pixel resolution of new image, and denominator is 500, is exactly original resolution.Utilize linear interpolation, though directly the pixel resolution of convergent-divergent image carries out dwindling of image or amplifies simple, yet this kind method need take advantage of in a large number/computing of division, so needs on the hardware to have than the taking advantage of of long number/divider and finish above-mentioned computing.Generally speaking, take advantage of/figure place of divider the more, its hardware cost and computational complexity are just higher.Above example when image resolution is become at 300 by 500, just needs at least nine take advantage of/divider on hardware.Thus, when image became big or becomes two dimension or multiplanar image, what hardware was required took advantage of/the also increase relatively of divider figure place, not only makes hardware cost improve, and makes that also operation time is longer, very inconvenient.
Therefore, need a kind ofly hardware cost be can effectively reduce, and image resolution Zoom method and the system of operation time shortened.
Summary of the invention
In view of this, the invention provides a kind of image treatment method and panntographic system, the above-mentioned calculating process that carries out the new pixel of generation in the image resolution conversion can be transformed into the multiplying of less figure place, and then reduce the figure place and the contraction in division device of required multiplier, therefore can effectively reduce hardware cost, and shorten operation time.
Based on above-mentioned purpose, the invention provides a kind of image treatment method, in order to an image is scaled to one second resolution by one first resolution, the pixel value P of the object pixel T under this second resolution wherein TBe pixel value P by k original pixels i under this first resolution iForm, k 〉=1 and i=1,2 ..., k.This method comprises the following steps.At first, find out the pixel value P of each these original pixels i iA corresponding proportion of composing R iThen, with the pixel value P of each these original pixels i iCorresponding this proportion of composing R iObtain the weighted value W of a correspondence via a computing i, W wherein iFor between 1 to 2 nInteger, and W iSummation be 2 n, and n is a positive integer.At last, with the pixel value P of each these original pixels i iAnd weighted value W that should correspondence iCarry out computing, produce the pixel value P of this object pixel T TTo finish image processing.
Image treatment method of the present invention should be via this computing to obtain this corresponding weighted value W iStep comprise: the inquiry comparison list (look-up table) to obtain this corresponding weighted value W i, wherein this table of comparisons has 2 nIndividual form field, and each these field has a reference value.
Image treatment method of the present invention should be via this computing to obtain this corresponding weighted value W iStep comprise: obtain a k weighted value W with following formula k:
W k = 2 n - &Sigma; l = 1 k - 1 Wi .
Image treatment method of the present invention, this inquires about this table of comparisons to obtain this corresponding weighted value W iStep more comprise: the pixel value P that inquires about an original pixels i iThis proportion of composing R iJudge this proportion of composing R of this inquiry iWhether less than a lower limit; And this proportion of composing R that works as this inquiry iBe during, with this proportion of composing R of this inquiry less than this lower limit iThe pixel value P of pairing this original pixels i iWeighted value be made as 0.
Image treatment method of the present invention is when this reference value of a field is to be the pixel value P more than or equal to this original pixels i iThis proportion of composing R iThe time, the weighted value W that this is corresponding iIt is the field preface that equals this field.
Image treatment method of the present invention is as the pixel value P of this original pixels i iThis proportion of composing R iBe between two reference values the time, the weighted value W that this is corresponding iBe to equal the wherein field preface of a field with decision with rounding-off method or unconditional casting out.
Image treatment method of the present invention, this reference value are relevant with this first resolution.
Image treatment method of the present invention, this produces the pixel value P of this object pixel T TStep more comprise: obtain the pixel value P of this object pixel T according to following formula T:
P T = &Sigma; i = 1 k P i &times; W i 2 n .
The present invention also provides a kind of panntographic system, in order to an image is scaled to one second resolution by one first resolution, and the pixel value P of an object pixel T of this second resolution wherein TBe pixel value P by k original pixels i under this first resolution iForm, k 〉=1 and i=1,2 ..., k.This system comprises a microprocessor, a storer, comparison list index generator (Table index generator) and a convergent-divergent control module (scalar).This microprocessor produces comparison list in order to according to this first resolution, and wherein this table of comparisons has 2 nIndividual form field, and each form field has a reference value.This storer is in order to provide these original pixels P iThis table of comparisons index generator is according to the pixel value P of an original pixels i iA corresponding proportion of composing R i, by obtaining a weighted value W in this table of comparisons i, and this weighted value W iBe between 1~2 nBetween integer, wherein n is a positive integer.This convergent-divergent control module is according to the pixel value P of these original pixels i iAnd weighted value W that should correspondence iComputing, produce the pixel value P of this object pixel T TWherein, this microprocessor is found out the pixel value P of these original pixels i iCorresponding this proportion of composing R i, and with the pixel value P of each these original pixels i iCorresponding this proportion of composing R i, via this table of comparisons index generator or via calculating to obtain this weighted value W i, and with this weighted value W iBe sent to this convergent-divergent control module to produce the pixel value P of this object pixel T T
Panntographic system of the present invention, this convergent-divergent control module comprises the multiplier of at least one n figure place, in order to carry out this computing to produce the pixel value P of this object pixel T T
Panntographic system of the present invention, this microprocessor can obtain a k weighted value W according to following formula k:
W k = 2 n - &Sigma; l = 1 k = 1 Wi .
Panntographic system of the present invention is when this reference value of a field is to be the pixel value P more than or equal to this original pixels i iThis proportion of composing R iThe time, the weighted value W that this is corresponding iIt is the field preface that equals this field.
Panntographic system of the present invention is as the pixel value P of this original pixels i iThis proportion of composing R iBe in this table of comparisons between two reference values time, the weighted value W that this is corresponding iBe to equal the wherein field preface of a field with decision with rounding-off method or unconditional casting out.
Panntographic system of the present invention, this reference value are relevant with this first resolution.
Panntographic system of the present invention, this convergent-divergent control module obtain the pixel value P of this object pixel according to following formula T:
P T = &Sigma; l = 1 K P i &times; W i 2 n .
Panntographic system of the present invention, when this original pixels was two dimension, this microprocessor produced one second table of comparisons.
Image treatment method of the present invention and panntographic system can effectively reduce hardware cost, and shorten operation time.
Description of drawings
Fig. 1 shows that a known image resolution dwindles the method synoptic diagram.
Fig. 2 shows an image resolution panntographic system according to the embodiment of the invention.
Fig. 3 shows a table of comparisons T according to the embodiment of the invention.
Fig. 4 shows an image processing schematic flow sheet according to the embodiment of the invention.
Fig. 5 shows that an one dimension image resolution according to the embodiment of the invention dwindles the method synoptic diagram.
Fig. 6 A is that demonstration one known bidimensional image resolution is dwindled the method synoptic diagram.
Fig. 6 B shows that one dwindles the method synoptic diagram according to the bidimensional image resolution of the embodiment of the invention.
Fig. 7 A shows the table of comparisons according to another embodiment of the present invention.
Fig. 7 B shows a known one dimension image resolution amplification method synoptic diagram.
Fig. 7 C shows an one dimension image resolution amplification method synoptic diagram according to the embodiment of the invention.
Embodiment
For above-mentioned and other purposes, feature and advantage of the present invention can be become apparent, cited below particularlyly go out preferred embodiment, and cooperate appended graphicly, be described in detail below.
The present invention will originally be used for the proportion of composing of linear interpolation, utilize the mode of computing, convert less value to, therefore can utilize the multiplier of less figure place can reach above-mentioned need be than the operation effect of the taking advantage of of long number/divider.According to the embodiment of the invention, when the figure place of supposing multiplier is n, then these proportion of composing will see through the mode of tabling look-up by factorial/division operations such as hundred figure places originally, be reduced to 2 nBe taking advantage of/division operation of grade.Because the figure place of multiplier is reduced to n, and has omitted divider, its hardware cost reduces relatively, also makes computing become simple and is easy to practical operation.
Please refer to Fig. 2, Fig. 2 shows a panntographic system synoptic diagram according to the embodiment of the invention.Panntographic system 200 comprises a microprocessor 210, a storer 220, comparison list index generator 230 and a scale controller 240.Deposited the pixel value of the original pixels of the required original resolution of computing (first resolution) in the storer 220, and the pixel value of these original pixels is sent to scale controller 240 has carried out computing.Table of comparisons T places a storage device (not illustrating), is produced by microprocessor 210, and therefore following inquiry table of comparisons T is actually the table of comparisons in this storage device of inquiry.Table of comparisons T is divided into a plurality of fields, and wherein field CX is in order to X the field of expression in the form, and X is defined as the field preface of field CX.For instance, field C1 represents the 1st field in the form, so its field preface is 1, and field C2 represents the 2nd field in the form, so its field preface is 2, by that analogy.In other words, the value of field preface X can push away by field CX.In addition, each field CX has the reference value UX of a correspondence, and along with the field preface X of field CX is bigger, its corresponding reference value UX is also bigger.For instance, suppose the corresponding reference value U1 of field C1, the corresponding reference value U2 of field C2, because the field preface of field C2 is greater than the field preface of field C1, so its corresponding reference value U2 will be greater than reference value U1.Wherein, the scope of the field preface X of field CX is decided by the figure place n of the multiplier that uses, and the scope of reference value UX is then decided by the original resolution before the convergent-divergent.Table of comparisons index generator 230 will produce the table of comparisons index of an inquiry usefulness finding the pairing field of a specific composition ratio, and the field preface of this field is defined the weighted value of specific composition ratio for this reason.At least the multiplier (not illustrating) that has comprised a n figure place in the scale controller 240 is in order to carry out multiplying to produce the new pixel behind the convergent-divergent.Scale controller 240 in order to according to the pixel value of the original pixels that obtains new resolution (second resolution) with and corresponding weighted value image is carried out convergent-divergent, produce the pixel value of the object pixel behind the convergent-divergent.Therefore, the image processing action of image resolution conversion can be finished, and then a display (not illustrating) can be driven to show the image after changing by panntographic system 200.
At first, provide image-zooming preceding original resolution Q and desire that image is carried out the new resolution P of convergent-divergent to panntographic system.Then, microprocessor 210 receives this original resolution Q and produces comparison list and each field corresponding reference value.
Please refer to Fig. 3, Fig. 3 shows a synoptic diagram according to the table of comparisons T of the embodiment of the invention.In this example, original resolution Q is 500, and new resolution P is 300, therefore utilizes Q to produce the reference value of field in the table of comparisons.Wherein, the field preface of field C1 is 1, and its corresponding reference value U1 is 31, and the field preface of field C2 is 2, and its corresponding reference value U2 is 63, by that analogy.As shown in Figure 3, in present embodiment, because the figure place n of multiplier is 4, table of comparisons T is divided into 2 according to the size of n nIndividual field, promptly (C1~C16), each field has a particular reference value to 16 fields.It should be noted that the reference value in the table of comparisons T is adjustable 16 positive integers, according to original resolution Q, ascending arrangement.In the present embodiment, be divided into 16 equal portions with original resolution Q is rough, and the reference value of field is ascending arrangement, and the reference value of each field can be added up a special value and gets by the reference value of previous field.For instance, as shown in Figure 3, the reference value U9 of field C9 is 281, suppose that resolution Q is 500, and then the reference value U10 of field C10 can push away according to following manner:
U10=U9+Q/16=281+500/16≈313
In addition, the reference value U16 in the field C16 is made as 500 (being original resolution Q), and the reference value U1 in the field C1 is made as minimum possible reference value.Numerical range when these reference values will be used for defining the inquiry table of comparisons.
Fig. 4 is the image treatment method process flow diagram 400 that shows according to the embodiment of the invention.This method can produce the pixel value of the new pixel behind the convergent-divergent in order to the pixel value according to one or more original image vegetarian refreshments, is applied on all original image vegetarian refreshments, just can reach and dwindle or the effect of magnified image resolution.As above-mentioned, the figure place n that supposes multiplier in the present embodiment is 4.At first, as step S410, find the combination of forming the required original image vegetarian refreshments of new pixel and the shared proportion of composing of pixel value of each original image vegetarian refreshments wherein.In this embodiment, the convergent-divergent computing is to adopt linear interpolation, try to achieve required original pixels combination with and the proportion of composing of pixel value.Such proportion of composing is determined by original resolution and new resolution fully, shown in (1) to (3) formula, all can learn in advance via simple calculating and derivation, and this is well known to those skilled in the art, and its method is described in detail in this omission.Then, as step S420, inquiry table of comparisons T or via calculating converts the proportion of composing of the pixel value of each pixel in the original pixels combination to 1~2 nBetween scope in an integer.As described above, table of comparisons T is divided into 16 fields in advance according to original resolution Q, and each field has a reference value, as shown in Figure 3.One proportion of composing will be used to therewith reference value compares, and finding one near the reference value of this proportion of composing, thereby obtains 1~2 of this portfolio ratio correspondence nBetween integer.The field preface of the pairing form field of this reference value of finding out is promptly represented the weighted value after its conversion.For instance, please refer to Fig. 3, suppose that the reference value of desiring to compare is 300, because the reference value U9 of field C9 is 281, the reference value U10 of field C10 is 313,281<300<313, the reference value U10 (313) that therefore can to retrieve its corresponding reference value be field C10, its weighted value just is made as the field preface 10 of field C10.So according to the present invention, initial value is that 300 proportion of composing will replace with a less numerical value 10.At last, as step S430, the pixel value of the pixel that each pixel weighted value that above-mentioned steps S420 is obtained is corresponding with it carries out computing, to produce the pixel value of new pixel.
It should be noted that, in this embodiment, the reference value of form field is ascending in regular turn ordering, finds out wherein the minimum reference value more than or equal to this proportion of composing when comparing the reference value of a proportion of composing and form field, and its field preface of corresponding field be its weighted value.Yet in other embodiment, its comparison also can be according to other rules, and for example rounding-off method or unconditional casting out or the like decide the weighted value (integer) after the conversion.In addition, suppose that the pixel value of a new pixel is made up of the pixel value of k original pixels, and k 〉=1, then equal 2 for the summation that meets weighted value n, the weighted value W of k original pixels kCan not see through and table look-up, but utilize following formula to try to achieve:
W k = 2 n - &Sigma; l = 1 k - 1 Wi
W wherein iThe weighted value of k-1 original pixels correspondence before the expression can be tried to achieve according to the above-mentioned process of tabling look-up.
For instance, be example with Fig. 5, show that with an original resolution Q be the image synoptic diagram that 500 image dwindles into 300 of new resolution.Wherein, the figure place n of multiplier is 4 and adopts linear interpolation to carry out computing.With reference to formula (1) as can be known, new pixel A ' pixel value PA ' formed with proportion of composing 300 and 200 respectively by pixel value PA and the PB of original pixels A and original pixels B, be new pixel A ' collection of pixels for (A, B), proportion of composing RA be 300 and RB be 200.Utilize method shown in Figure 4, produce the table of comparisons T of 16 fields as shown in Figure 3 according to original resolution Q.Then, inquire about table of comparisons T, find out the weighted value WA of corresponding RA.Because RA is between reference value U9 and U10, therefore, can learn that its corresponding weighted value WA is 10 (because of the field preface of the corresponding field C10 of reference value U10 is 10).So the weighted value WB1 of B correspondence can push away according to following formula:
WB1=16-WA=16-10=6
Then, formula (1) can be made into the form of following formula (4):
PA &prime; = 10 &times; PA + 6 &times; PB 16 . . . . . . . . . . . . ( 4 )
By above-mentioned formula (4) as can be known, pixel A ' pixel value PA ' can utilize 4 multiplier to carry out computing, and be equivalent to (result of 10 * PA+6 * PB), four positions that move to right reach the purpose that does not need divider divided by 16 part.
Similarly, pixel B ' pixel value PB ' can be then decide by the pixel value PB of remaining original pixels B and the pixel value PC of original pixels C.According to formula (2) as can be known
PB &prime; = 100 &times; PB + 300 &times; PC + 100 &times; PD 500 . . . . . . . . . . . . ( 2 ) ,
Inquiry table of comparisons T can learn 100 between U3 (94) and U4 (125), so the weighted value WB2 of B is 4.After obtaining the weighted value WB2 of B, because RC is 300, from the above, so the weighted value WC=WA=10 of C just can utilize following formula to obtain the weighted value WD1 of D:
WD1=16-WB2-WC=16-4-10=2
So, formula (2) can be made into the form of following formula (5):
PB &prime; = 4 &times; PB + 10 &times; PC + 2 &times; PD 16 . . . . . . . . . . . . ( 5 ) ,
Wherein PD is the pixel value of pixel D.By above-mentioned formula (5) as can be known, pixel B ' pixel value PB ' also can utilize 4 multiplier to carry out computing.Utilize similar computing method, formula (3) can be made into the form of following formula (6):
PC &prime; = 7 &times; PD + 9 &times; PE 16 . . . . . . . . . . . . . . . . ( 6 )
Similarly, the pixel value PC ' of pixel C ' also can utilize 4 multiplier to carry out computing.It should be noted that when the pixel value of a new pixel is made up of the pixel value of a plurality of points, can utilize the proportion of composing of the pixel value of differentiating each pixel, and will ignore less than the proportion of composing of a set value and simplify computing.Because proportion of composing represents that it accounts for a ratio in the new pixel, therefore, when proportion of composing during less than a lower limit, represent that its ratio shared in the pixel value of new pixel is very little, this proportion of composing can be ignored or the weighted value that it is corresponding is made as 0, get next point and inquire about the table of comparisons and find out weighted value, to simplify computing.Therefore, can before tabling look-up, judge a lower limit earlier according to method of the present invention, when the proportion of composing of the pixel value of an original image vegetarian refreshments during less than this lower limit, this pixel will be left in the basket or the weighted value that it is corresponding is made as 0, and the proportion of composing of the pixel value of this pixel can add the proportion of composing of the pixel value of next original image vegetarian refreshments to be inquired about to carry out the table of comparisons.For instance, suppose that lower limit is decided to be 16, the pixel value PD ' of new pixel D ' can be tried to achieve by following formula:
PD &prime; = 10 &times; PB + 300 &times; PC + 190 &times; PD 500 . . . . . . . . . . . . . . . . . . ( 7 )
When carrying out table of comparisons when inquiry, because wherein the proportion of composing of the pixel value of pixel B is 10,10 less than lower limit 16, this pixel will be left in the basket, get the proportion of composing 300 of the pixel value of next pixel C, add the proportion of composing 10 of the pixel value of preceding pixel B, table look-up with 310.Therefore, formula (7) can change following form into:
PD &prime; = 10 &times; PC + 6 &times; PD 16 . . . . . . . . . . . . . . . . ( 8 )
In sum, if the pixel value of the pixel P ' among the new resolution P by k original image vegetarian refreshments among the original resolution Q (1,2 ..., pixel value k) (P1, P2 ..., Pk) respectively with proportion of composing (R1, R2 ..., Rk) form, then
P &prime; = &Sigma; i = 1 K Pi &times; Ri Q . . . . . . . . . . . . . . . . . . . ( 9 )
Wherein, Ri is the integer between 1~P.According to method of the present invention, formula (9) can be rewritten into
P &prime; = &Sigma; i = 1 k Pi &times; Wi 2 n , . . . . . . . . . . . . . . . . . . . . . . . . ( 10 )
Wherein, Wi is 1~2 nBetween integer and
&Sigma; i = 1 k Wi = 2 n .
Therefore,, only need the multiplier of n position just can finish image-zooming, can reduce the complexity and the cost of hardware according to method of the present invention required multiplier figure place and reduce divider can reduce its computing the time.In addition, also can suitably adjust the size of n value to obtain more suitable degree of accuracy and usefulness.Especially, when image was the image of multidimensional, the effect of its saving more obviously.
Fig. 6 A shows the original image vegetarian refreshments and the new graph of a relation between the pixel when image become 300 * 300 resolution by 500 * 500 resolution.As shown in the figure, solid line part and dotted portion are represented the distribution scenario of each pixel under the resolution of 500 * 500 resolution and 300 * 300 respectively.Wherein, the pixel value PA ' of new pixel A ' is made of pixel value PA, PB, PC and the PD of original image vegetarian refreshments A, B, C and D.If adopt known two-dimentional bilinear interpolation method, the pixel value PA ' of then new pixel A ' can obtain according to following formula:
PA &prime; = ( PA &times; 300 &times; 300 + PB &times; 200 &times; 300 + PC &times; 300 &times; 200 + PD &times; 200 &times; 200 ) 500 &times; 500 . . . . ( 11 )
In order to finish the computing of (11) formula, need two taking advantage of more than nine/dividers at least, and its operand is quite huge and consuming time.Fig. 6 B shows the image treatment method according to the embodiment of the invention, original image vegetarian refreshments when image is become 300 * 300 resolution by 500 * 500 resolution and the new graph of a relation between the pixel.Similarly, solid line part and dotted portion are represented the distribution scenario of each pixel under the resolution of 500 * 500 resolution and 300 * 300 respectively.According to the embodiment of the invention, be example as if multiplier with 4, its operator is 2 4=16, because pixel has level and vertical coordinate, therefore can set up the table of comparisons of a horizontal coordinate and the table of comparisons of a vertical coordinate respectively, the pixel value proportion of composing 300 * 300 of tabling look-up respectively again pixel A changes into weighted value I x* I Y(WA), wherein, I xAnd I YThe weighted value of representing its level and vertical coordinate to table look-up respectively and getting, and I xAnd I YBe 1~2 nBetween integer.After obtaining the weighted value WA of pixel A, the weighted value WB of pixel B, C, D, WC and WD just can be extrapolated by WA, and be as follows:
WB=(16-I x)×I Y,WC=(16-I Y)×I x,WD=(16-I x)×(16-I Y),
Therefore, formula (11) will be rewritten into
PA &prime; = ( PA &times; I x &times; I y + PB &times; ( 16 - I x ) &times; I y + PC &times; I x &times; ( 16 - I y ) + PD &times; ( 16 - I x ) &times; ( 16 - I y ) 16 &times; 16
By following formula as can be known, nine multiplier computings of 300 * 300 of script have been reduced to I x* I v4 multiplier computings, originally also economize the operation result that slightly only needs molecule eight positions that move to right divided by 500 * 500, its operand significantly reduces and also will shorten many operation time.
Though above-mentioned example is all with the example that is reduced into of image resolution, yet the present invention also can be applicable to the amplification of image.Fig. 7 A shows the table of comparisons according to use in the image amplification of the embodiment of the invention.Be known amplification method among Fig. 7 B, it is to be that pixel A-C in 300 becomes the example that resolution is pixel the A '-E ' in 500 with image by resolution.If adopt known one dimension bilinear interpolation method, pixel value PA '-PE ' of then new pixel A '-E ' can obtain according to following formula:
PA &prime; = 300 &times; PA 300 , . . . . . . . . . . . . (1)’
PB &prime; = 200 &times; PA + 100 &times; PB 300 , . . . . . . . . . . . . (2)’
PC &prime; = 300 &times; PB 300 , . . . . . . . . . . . . . . . . . . (3)’
PD &prime; = 100 &times; PB + 200 &times; PC 300 , . . . . . . . . . . . . (4)’
PE &prime; = 300 &times; PC 300 , . . . . . . . . . . . . . . . . . . (5)’
For instance, for example ask the pixel value of B ', by (2) ' formula can learn that the proportion of composing of the pixel value PA ' of A among the B ' is 200, table look-up (Fig. 7 A) obtains its corresponding weighted value is 11.Therefore, (2) ' formula can be rewritten as
PB &prime; = 11 &times; PA + ( 16 - 11 ) &times; PB 16 , . . . . . . . . . . . . . . . . . . (2)”
The pixel value algorithm of rest of pixels point A ', C ', D ' and E ' point can convert its proportion of composing between 1~16 integer according to aforesaid method and rule, shown in Fig. 7 C.Therefore, method of the present invention is applied in the amplification of image, its operand also will significantly reduce and also will shorten many operation time.
The above only is preferred embodiment of the present invention; so it is not in order to limit scope of the present invention; any personnel that are familiar with this technology; without departing from the spirit and scope of the present invention; can do further improvement and variation on this basis, so the scope that claims were defined that protection scope of the present invention is worked as with the application is as the criterion.
Being simply described as follows of symbol in the accompanying drawing:
R x、R i: proportion of composing
P T: object pixel
P i: original pixels
W i: weighted value
N: figure place
200: panntographic system
210: microprocessor
220: memory
230: table of comparisons index generator
240: scale controller
250: table of comparisons T
CX, C1-C16: field
X: field preface
UX, U1-U16: reference value
Q: original resolution
P: new resolution
S410-S430: step

Claims (16)

1. an image treatment method is characterized in that, this image treatment method is in order to be scaled to one second resolution with an image by one first resolution, wherein the pixel value P of the object pixel T under this second resolution TBe pixel value P by k original pixels i under this first resolution iForm, k 〉=1 and i=1,2 ..., k, it comprises the following steps:
Find out the pixel value P of each this original pixels i iA corresponding proportion of composing R i
Pixel value P with each this original pixels i iCorresponding this proportion of composing R iObtain the weighted value W of a correspondence via a computing i, W wherein iFor between 1 to 2 nInteger, and W iSummation be 2 n, and n is a positive integer; And
Pixel value P with each this original pixels i iAnd weighted value W that should correspondence iCarry out computing, produce the pixel value P of this object pixel T TTo finish image processing.
2. image treatment method according to claim 1 is characterized in that, should be via this computing to obtain this corresponding weighted value W iStep comprise:
The inquiry comparison list is to obtain this corresponding weighted value W i, wherein this table of comparisons has 2 nIndividual form field, and each this field has a reference value.
3. image treatment method according to claim 1 is characterized in that, should be via this computing to obtain this corresponding weighted value W iStep comprise:
Obtain a k weighted value W with following formula k:
W k = 2 n - &Sigma; i = 1 k - 1 Wi .
4. image treatment method according to claim 2 is characterized in that, this inquires about this table of comparisons to obtain this corresponding weighted value W iStep more comprise:
Inquire about the pixel value P of an original pixels i iThis proportion of composing R i
Judge this proportion of composing R of this inquiry iWhether less than a lower limit; And
This proportion of composing R when this inquiry iBe during, with this proportion of composing R of this inquiry less than this lower limit iThe pixel value P of pairing this original pixels i iWeighted value be made as 0.
5. image treatment method according to claim 2 is characterized in that, when this reference value of a field is to be the pixel value P more than or equal to this original pixels i iThis proportion of composing R iThe time, the weighted value W that this is corresponding iIt is the field preface that equals this field.
6. image treatment method according to claim 2 is characterized in that, as the pixel value P of this original pixels i iThis proportion of composing R iBe between two reference values the time, the weighted value W that this is corresponding iBe to equal the wherein field preface of a field with decision with rounding-off method or unconditional casting out.
7. image treatment method according to claim 2 is characterized in that, this reference value is relevant with this first resolution.
8. image treatment method according to claim 1 is characterized in that, this produces the pixel value P of this object pixel T TStep more comprise:
Obtain the pixel value P of this object pixel T according to following formula T:
P T = &Sigma; i = 1 k P i &times; W i 2 n .
9. a panntographic system is characterized in that, this panntographic system is in order to be scaled to one second resolution with an image by one first resolution, wherein the pixel value P of an object pixel T of this second resolution TBe pixel value P by k original pixels i under this first resolution iForm, k 〉=1 and i=1,2 ..., k, this system comprises:
One microprocessor in order to according to this first resolution, produces comparison list, and wherein this table of comparisons has 2 nIndividual form field, and each form field has a reference value;
One storer is in order to provide this original pixels P i
The comparison list index generator is according to the pixel value P of an original pixels i iA corresponding proportion of composing R i, by obtaining a weighted value W in this table of comparisons i, and this weighted value W iBe between 1~2 nBetween integer, wherein n is a positive integer; And
One convergent-divergent control module is according to the pixel value P of this original pixels i iAnd weighted value W that should correspondence iComputing, produce the pixel value P of this object pixel T T,
Wherein this microprocessor is found out the pixel value P of this original pixels i iCorresponding this proportion of composing R i, and with the pixel value P of each this original pixels i iCorresponding this proportion of composing R i, via this table of comparisons index generator or via calculating to obtain this weighted value W i, and with this weighted value W iBe sent to this convergent-divergent control module to produce the pixel value P of this object pixel T T
10. panntographic system according to claim 9 is characterized in that this convergent-divergent control module comprises the multiplier of at least one n figure place, in order to carry out this computing to produce the pixel value P of this object pixel T T
11. panntographic system according to claim 9 is characterized in that, this microprocessor can obtain a k weighted value W according to following formula k:
W k = 2 n - &Sigma; i = 1 k - 1 Wi .
12. panntographic system according to claim 9 is characterized in that, when this reference value of a field is to be the pixel value P more than or equal to this original pixels i iThis proportion of composing R iThe time, the weighted value W that this is corresponding iIt is the field preface that equals this field.
13. panntographic system according to claim 9 is characterized in that, as the pixel value P of this original pixels i iThis proportion of composing R iBe in this table of comparisons between two reference values time, the weighted value W that this is corresponding iBe to equal the wherein field preface of a field with decision with rounding-off method or unconditional casting out.
14. panntographic system according to claim 9 is characterized in that, this reference value is relevant with this first resolution.
15. panntographic system according to claim 9 is characterized in that, this convergent-divergent control module obtains the pixel value P of this object pixel according to following formula T:
P T = &Sigma; i = 1 k P i &times; W i 2 n .
16. panntographic system according to claim 9 is characterized in that, when this original pixels was two dimension, this microprocessor produced one second table of comparisons.
CNB2006101272210A 2006-09-12 2006-09-12 Image treatment method and panntographic system Expired - Fee Related CN100541604C (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CNB2006101272210A CN100541604C (en) 2006-09-12 2006-09-12 Image treatment method and panntographic system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CNB2006101272210A CN100541604C (en) 2006-09-12 2006-09-12 Image treatment method and panntographic system

Publications (2)

Publication Number Publication Date
CN101145336A true CN101145336A (en) 2008-03-19
CN100541604C CN100541604C (en) 2009-09-16

Family

ID=39207826

Family Applications (1)

Application Number Title Priority Date Filing Date
CNB2006101272210A Expired - Fee Related CN100541604C (en) 2006-09-12 2006-09-12 Image treatment method and panntographic system

Country Status (1)

Country Link
CN (1) CN100541604C (en)

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102282586A (en) * 2009-01-19 2011-12-14 诺基亚公司 Method and apparatus for reducing size of image data
CN101751660B (en) * 2009-12-25 2012-06-27 四川长虹电器股份有限公司 Picture zooming method
CN102831576A (en) * 2012-06-14 2012-12-19 北京暴风科技股份有限公司 Video image zooming method and system
CN103237214A (en) * 2013-04-12 2013-08-07 华为技术有限公司 Method and device for coding and decoding depth images
CN104505015A (en) * 2015-01-13 2015-04-08 京东方科技集团股份有限公司 Display method for a display panel, display panel and display device
CN108630139A (en) * 2018-05-08 2018-10-09 京东方科技集团股份有限公司 Image display processing method and device, display device and storage medium
CN113284472A (en) * 2021-05-13 2021-08-20 深圳乐播科技有限公司 Screen projection resolution conversion method, device, equipment and storage medium

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102282586A (en) * 2009-01-19 2011-12-14 诺基亚公司 Method and apparatus for reducing size of image data
US9042678B2 (en) 2009-01-19 2015-05-26 Nokia Corporation Method and apparatus for reducing size of image data
CN101751660B (en) * 2009-12-25 2012-06-27 四川长虹电器股份有限公司 Picture zooming method
CN102831576A (en) * 2012-06-14 2012-12-19 北京暴风科技股份有限公司 Video image zooming method and system
CN102831576B (en) * 2012-06-14 2014-09-24 北京暴风科技股份有限公司 Video image zooming method and system
CN103237214A (en) * 2013-04-12 2013-08-07 华为技术有限公司 Method and device for coding and decoding depth images
CN103237214B (en) * 2013-04-12 2016-06-08 华为技术有限公司 The decoding method of depth image and coding and decoding device
CN104505015A (en) * 2015-01-13 2015-04-08 京东方科技集团股份有限公司 Display method for a display panel, display panel and display device
CN108630139A (en) * 2018-05-08 2018-10-09 京东方科技集团股份有限公司 Image display processing method and device, display device and storage medium
CN108630139B (en) * 2018-05-08 2021-03-05 京东方科技集团股份有限公司 Image display processing method and device, display device and storage medium
CN113284472A (en) * 2021-05-13 2021-08-20 深圳乐播科技有限公司 Screen projection resolution conversion method, device, equipment and storage medium

Also Published As

Publication number Publication date
CN100541604C (en) 2009-09-16

Similar Documents

Publication Publication Date Title
CN100541604C (en) Image treatment method and panntographic system
CN109584156B (en) Microscopic sequence image splicing method and device
CN104067296B (en) The method and apparatus that one-dimensional signal extracts
CN103366342B (en) Be applied to the subsection linearity inser value method that video image amplifies
CN110728210A (en) Semi-supervised target labeling method and system for three-dimensional point cloud data
CN110147750B (en) Image searching method and system based on motion acceleration and electronic equipment
CN103870516A (en) Image retrieving method, real-time drawing prompting method and device thereof
CN102881000A (en) Super-resolution method, device and equipment for video image
CN109493400A (en) Handwriting samples generation method, device, computer equipment and storage medium
CN104660951A (en) Super-resolution amplification method of ultra-high definition video image converted from high definition video image
CN102982147B (en) A kind of method and apparatus improving data message integration efficiency
CN102542298A (en) Electronic device and image similarity degree comparison method thereof
Cao et al. Fast incremental structure from motion based on parallel bundle adjustment
CN104318591A (en) Dynamic drawing method with boundary plane flow field
CN105719330A (en) Animation curve generation method and device
CN111260723B (en) Barycenter positioning method of bar and terminal equipment
CN106408499B (en) Method and device for acquiring reverse mapping table for image processing
CN113407537A (en) Data processing method and device and electronic equipment
CN106775596A (en) A kind of infrared image linear interpolation expands hardware processing method
CN107240064A (en) Method for correcting image, device and terminal
CN110717860A (en) Image scaling processing method based on Newton&#39;s three-correlation-connected-fraction mixed rational interpolation
CN113763559B (en) Geometric motion detail reconstruction method for fitting depth image
CN112685865B (en) Multi-topology model merging method, device, equipment and storage medium
CN114120011B (en) Method for improving image matching precision and speed by using group intelligent algorithm
CN111062924B (en) Image processing method, device, terminal and storage medium

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20090916

Termination date: 20180912