CN101352048A - Device of processing dead pixel - Google Patents
Device of processing dead pixel Download PDFInfo
- Publication number
- CN101352048A CN101352048A CNA2006800499742A CN200680049974A CN101352048A CN 101352048 A CN101352048 A CN 101352048A CN A2006800499742 A CNA2006800499742 A CN A2006800499742A CN 200680049974 A CN200680049974 A CN 200680049974A CN 101352048 A CN101352048 A CN 101352048A
- Authority
- CN
- China
- Prior art keywords
- value
- threshold value
- detecting unit
- pixel
- comparison
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N9/00—Details of colour television systems
- H04N9/64—Circuits for processing colour signals
- H04N9/646—Circuits for processing colour signals for image enhancement, e.g. vertical detail restoration, cross-colour elimination, contour correction, chrominance trapping filters
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformation in the plane of the image
- G06T3/40—Scaling the whole image or part thereof
- G06T3/4015—Demosaicing, e.g. colour filter array [CFA], Bayer pattern
Abstract
A dead pixel processing device is disclosed. The dead pixel processing device separates an inputted Bayer pattern image into corresponding component data; calculates the distribution (pattern) of pixels based on the median of each data; calculates a comparing value based on a component having a center pixel; outputs an external flag which informs whether there are a dead pixel and/or a hot pixel by using the comparing value; compares the values of the center pixel and adjacent pixels in the component having the center pixel; calculates a measuring value based on the center pixel; outputs an internal flag by using the measuring value; and corrects the dead pixel or the hot pixel. With the present invention, an image can be corrected by detecting a corresponding dead pixel and hot pixel.
Description
The cross reference of related application
According to 35U.S.C..sctn 119 (a)-(d), the application requires the foreign priority of the PCT/KR2006/005844 that submits to December in 2006 28, in the mode of reference its full content is incorporated into this.
Technical field
The present invention relates to device of processing dead pixel, more particularly, relate to the device of processing dead pixel of imageing sensor.
Background technology
In imageing sensor, arrange a plurality of pixels with two-dimensional structure.Each pixel will be imported light according to the brightness of input light and be converted to the signal of telecommunication.Can discern the light quantity that is input to each pixel by the signal of telecommunication of measuring after changing.Whereby, can construct with the pixel is the image of unit.
Imageing sensor is made of to several thousand storage devices the device and the hundreds of that have the pel array of hundreds of thousands to millions of pixels, analogue data is converted to numerical data.During each relevant treatment, may there be the possibility that causes error because of a large amount of devices.These errors cause generating in the pixel bad point of pixel (dead pixel) or focus (hot pixel).Bad point becomes the grade of imageing sensor and the key factor in the price.Pixel data by bad dot generation is compared excessive or too small with the pixel data that neighbor generates.
Conventional bad some processing method mainly compares pixel and benchmark pixel.If the difference between pixel and the benchmark pixel not in preset range, then is bad point with this pixel definition and proofreaies and correct.Yet conventional bad some processing method is used many hardware resources.At the dense graph picture or have under the situation of image of high-level luminance difference, widely different between the neighbor.Therefore, conventional bad some processing method is defined as correct pixel bad point mistakenly, and proofreading and correct correct pixel.As a result, the image after the recovery is compared with corresponding real image and is had many distortions.
Summary of the invention
For overcoming the above problems the device of processing dead pixel that the invention provides that designs, compare to determine with pattern and other components of the component by in 6 * 6 masks, will having center pixel whether pixel is bad point or focus.
In addition, the invention provides a kind of device of processing dead pixel, carried out a pixel value of proofreading and correct for relevant bad point with output.
In addition, the invention provides a kind of device of processing dead pixel, carried out 6 * 6 pixel values of proofreading and correct for relevant bad point with output.
By following execution mode, other purposes of the present invention will become and be more readily understood.
In order to realize above purpose, according to an embodiment of the invention, device of processing dead pixel comprises: separative element, and its Baeyer pattern (Bayer pattern) image with input is divided into corresponding component data; And comparing unit, its intermediate value based on each data is come the distribution (pattern) of calculating pixel, calculate comparison value based on component with center pixel, utilize this comparison value to come output notice whether to have the trappings of bad point and/or focus, the value of the center pixel in the component with center pixel and the value of neighbor are compared; Calculate measured value based on this center pixel; And utilize this measured value to export inner mark.Here, this device of processing dead pixel may further include correcting unit, and its output utilizes the pixel value after the trappings that receive from this comparing unit and inner mark are proofreaied and correct.
If determine to exist bad point/focus, then this correcting unit has the intermediate value of the component data of center pixel with reference to these trappings and the output of this inner mark.In addition, if determine not bad point/focus, then this correcting unit is exported this center pixel with reference to these trappings and this inner mark.
This comparing unit has: the central value detecting unit, and this central value detecting unit is used for each intermediate value of detected components data; The first absolute value detecting unit is used to detect the absolute value that deducts the obtained value of described intermediate value from each element; The difference detected value, be used to detect difference, that is, deduct the absolute value of this first absolute value detecting unit at the dateout of component data at each obtained value of the dateout of the component data except that described component data with center pixel with described center pixel from this first absolute value detecting unit; The comparison value detecting unit is used for the output of this difference detecting unit is sued for peace, and at the comparison value of each element detected components data; The trappings generation unit is used for this comparison value and each bad some threshold value and focus threshold value are compared, and generates and whether have a bad threshold value and/or the relevant trappings of focus threshold value; The second absolute value detecting unit is used to detect the absolute value that deducts the obtained value of this center pixel value from each element of component data with center pixel; And the inner mark generation unit, be used for the threshold value by utilizing brightness classification usefulness and the intermediate value of this component data and set the inner mark threshold value, and utilize this threshold value to generate inner mark.
When being compared, the comparison value of every element and this evil idea point threshold value having only under the situation of comparison value greater than bad some threshold value of this center pixel, and/or when with the comparison value of every element and focus threshold ratio than the time have only under the situation of value greater than the focus threshold value of this center pixel, this trappings generation unit is set at trappings and has bad point/focus.In addition, this inner mark generation unit is sued for peace to the intermediate value of each component data; Threshold value according to this threshold setting inner mark; And at the absolute value of all every elements that this second absolute value detecting unit calculates all under the situation greater than the threshold value of this inner mark, this inner mark is set at there is bad point/focus.
Another execution mode according to the present invention, a kind of image processor that is used for imaging device of handling bad point, this image processor comprises: separative element is divided into corresponding component data with the Baeyer mode image of importing; The intermediate value detecting unit detects each intermediate values of this detected components data; The first absolute value detecting unit is used to detect the absolute value that deducts the obtained value of this intermediate value from each element of this component data; The difference detected value, be used to detect difference, that is, deduct the absolute value of this first absolute value detecting unit at the dateout of component data at each obtained value of the dateout of the component data except that described component data with center pixel with center pixel from this first absolute value detecting unit; The comparison value detecting unit is used for the output of this difference detecting unit is sued for peace, and at the comparison value of each element detected components data; The trappings generation unit is used for this comparison value and each bad some threshold value and focus threshold value are compared, and generates and whether have a bad threshold value and/or the relevant trappings of focus threshold value; The second absolute value detecting unit is used to detect the absolute value that deducts the obtained value of this center pixel value from each element of component data with center pixel; And the inner mark generation unit, be used for the threshold value by utilizing brightness classification usefulness and the intermediate value of this component data and set the inner mark threshold value, and utilize this threshold value to generate inner mark; And correcting unit, output utilizes respectively the pixel value after the trappings that receive from this trappings generation unit and this inner mark generation unit and inner mark are proofreaied and correct.
Another execution mode according to the present invention provides a kind of device of processing dead pixel, and this device of processing dead pixel comprises: separative element is used for input Baeyer mode image is divided into corresponding component data; And a bad some detecting unit, based on the distribution (pattern) of the median calculation pixel of each data, the relatively pattern of other components; Calculate comparison value based on each component; This comparison value and bad some threshold value are compared; Determine whether corresponding pixel is bad point; And export calibrated component data.
This evil idea point detecting unit has: the first central value detecting unit, and this first central value detecting unit is used for each intermediate value of detected components data; The first absolute value detecting unit is used to detect the absolute value that deducts the obtained value of intermediate value from each element; First comparing unit is used to calculate each absolute value that deducts the obtained value of first output from second output of the first absolute value detecting unit; Described absolute value is sued for peace; And at each component and each Data Detection comparison value; The threshold setting unit, the threshold value by utilizing brightness classification usefulness and the intermediate value of the detected component data of the first intermediate value detecting unit are set the threshold value of inner mark; And the first output determining unit, be used for a detected comparison value of this first comparison value detecting unit and a bad some threshold value are compared, and according to whether existing bad point to determine dateout.
Here, this first output determining unit compares a detected comparison value of this first comparison value detecting unit and a bad some threshold value, and if not greater than the comparison value of this evil idea point threshold value then export this component data.This first output determining unit compares this detected comparison value of first comparison value detecting unit and this evil idea point threshold value, if and a comparison value threshold value is arranged greater than bad some threshold value, then determine each element of component data, this first output determining unit will compare from element-specific absolute value that deducts the obtained value of adjacent element and the threshold value of being determined by this threshold value determining unit simultaneously, if and all absolute values then are defined as this element the intermediate value of adjacent element all greater than this threshold value.This first output determining unit compares this detected comparison value of first comparison value detecting unit and this evil idea point threshold value, if and have a comparison value threshold value greater than bad some threshold value then determine component data each element, this first output determining unit will deduct the absolute value of the obtained value of adjacent element and threshold value that this threshold value determining unit is determined compares from element-specific simultaneously, if, then coherent element is defined as output not greater than the absolute value of this threshold value.
Preferably, this device of processing dead pixel also has the focus detecting unit, and this focus detecting unit is used for calculating distribution (pattern) according to the component data that receives from this evil idea point detecting unit, calculates comparison value based on each component; With this comparison value and focus threshold ratio; Determine whether corresponding pixel is focus; And export calibrated component data.
Here, this focus detecting unit has the second central value detecting unit, and this second central value detecting unit is used to detect each intermediate value of the component data that receives from this evil idea point detecting unit; The second absolute value detecting unit is used to detect the absolute value that deducts the obtained value of the detected intermediate value of this second intermediate value detecting unit from each element; Second comparing unit is used to calculate each absolute value that deducts the obtained value of first output from second output of this second absolute value detecting unit; Described absolute value is sued for peace, and at each component and each Data Detection comparison value; And the second output determining unit, be used for and will compare, and according to whether existing focus to determine dateout by this detected comparison value of second comparison value determining unit and this focus threshold value.
At this moment, this second output determining unit will be compared by the detected comparison value of the second comparison value detecting unit and this focus threshold value, and if less than greater than the comparison value of this focus threshold value then export this component data.This second output determining unit will be compared by the detected comparison value of the second comparison value detecting unit and this focus threshold value, if and a comparison value is arranged greater than this focus threshold value, then determine each element of component data, this second output determining unit will deduct the absolute value of the obtained value of adjacent element and threshold value that this threshold value determining unit is determined compares from element-specific simultaneously, if all described absolute values all greater than this threshold value, then are defined as this element the intermediate value of described adjacent element.This second output determining unit compares the detected comparison value of the second comparison value detecting unit and this focus threshold value, if and a comparison value is arranged greater than the focus threshold value, then determine each element of component data, this second output determining unit will deduct the absolute value of the obtained value of adjacent element and threshold value that this threshold value determining unit is determined compares from element-specific simultaneously, if and, then coherent element is defined as output not greater than the absolute value of this threshold value.
Description of drawings
Fig. 1 is an illustration according to the block diagram of the execution mode of the device of processing dead pixel of first embodiment of the invention;
The example of input 6 * 6 Baeyer mode images of Fig. 2 is illustration Fig. 1;
Fig. 3 is an illustration is divided into the example of 3 * 3 components with the Baeyer mode image of Fig. 2;
The detailed diagram of the execution mode of the pattern comparing unit of Fig. 4 is illustration Fig. 1;
The example of the work of Fig. 5 is the illustration intermediate value detecting unit of Fig. 4, the first absolute value detecting unit, difference detecting unit and comparison value;
The example of the work of the second absolute value detecting unit of Fig. 6 is illustration Fig. 4;
Fig. 7 is an illustration imports the example of the original image of device of processing dead pixel of the present invention;
Fig. 8 has been an illustration by detecting the example of the image after the bad point of being correlated with among Fig. 7 is proofreaied and correct;
Fig. 9 is an illustration according to the block diagram of the execution mode of the device of processing dead pixel of second embodiment of the invention;
The badly detailed diagram of a detecting unit of Fig. 9 that Figure 10 has been an illustration;
The example of the work of Figure 11 is the illustration intermediate value detecting unit of Figure 10, absolute value detecting unit and comparison value detected value;
Figure 12 is an illustration from the example of the image of the bad some detecting unit output of Fig. 9; And
The detailed diagram of the execution mode of the focus detecting unit of Figure 13 is illustration Fig. 9.
Embodiment
With reference to accompanying drawing, will make above purpose, feature and advantage more apparent by following description.
Because the present invention may exist various changes and execution mode, therefore some execution mode is wherein described and illustration with reference to accompanying drawing.Yet this never is to limit the invention to these specific execution modes, but should be interpreted as comprising that the institute that spirit and scope of the invention covers changes, equivalent and substitute.In whole accompanying drawing, similar key element is provided identical label.In whole specification of the present invention, when determine describing certain technology and can obscure the present invention and will put, omit relevant detailed description to this technology.
When describing each key element, for example can use the term of " first " and " second " and so on, but above these key elements should not be subjected to the restriction of above term.Above term only is used for key element is distinguished from each other.For example, under the situation of the scope that does not break away from claim of the present invention, can be with first element called after second key element, can also be with the second key element called after first element.Term " and/or " should comprise the combination of a plurality of projects of enumerating or any one of a plurality of projects of enumerating.
When a key element being described as " being connected to (connected) " or " being linked into (accessed) " another key element, not only can being interpreted as being directly connected to or being linked into another key element but also can being interpreted as between the two, having other key elements.On the other hand, if a key element is described as " directly connect " or " directly inserting " to another key element, then should be interpreted as not having between the two other key elements.
The term that uses in specification only is intended to describe some execution mode, and absolutely not in order to limit the present invention.Unless spell out in addition, otherwise the odd number expression comprises plural implication.In this description, for example " comprise " or " by ... form " expression be intended to specific characteristic, quantity, step, operation, element, parts or its combination, should not be interpreted as getting rid of any existence or the possibility of one or more other features, quantity, step, operation, element, parts or its combination.
Unless otherwise defined, otherwise all terms of technical term and scientific and technical terminology that comprise used herein all have and the identical implication of those skilled in the art's common sense implication.Any term that defines in general dictionary all should be interpreted as having identical meanings in the background at correlative technology field, and, unless definition clearly in addition, otherwise should not be interpreted as having the ideal or the undue implication of literalization.
Hereinafter, describe preferred implementation with reference to the accompanying drawings in detail.With figure number irrespectively to the given identical label of identical or corresponding key element, and identical or corresponding key element is not repeated any redundant description.
Fig. 1 is an illustration according to the block diagram of the execution mode of the device of processing dead pixel of first embodiment of the invention.
According to Fig. 1, device of processing dead pixel can comprise component separative element 110, pattern comparing unit 120 and correcting unit 130.
Component separative element 110 will be imported 6 * 6 Baeyer mode images and be divided into as shown in Figure 343 * 3 components.The example of input 6 * 6 Baeyer mode images of Fig. 2 is illustration Fig. 1, Fig. 3 is an illustration is divided into the example of 3 * 3 components with the Baeyer mode image of Fig. 2.
As shown in Figures 2 and 3, component separative element 110 of the present invention is divided into 6 * 6 Baeyer mode images only have same components 3 * 3 data of (for example, being respectively green (G) of red (R), GR capable (Gr), G and the indigo plant (B) of GB capable (Gb)).
" B5 " that supposes Fig. 3 is the center pixel of Fig. 2, describes the correction of " B5 " by example.Then, this example is carried out vague generalization.
The detailed diagram of the execution mode of the pattern comparing unit of Fig. 4 is illustration Fig. 1.
As shown in Figure 4, pattern comparing unit 120 can comprise intermediate value detecting unit 121, the first absolute value detecting unit 122, difference detecting unit 123, comparison value detecting unit 124, trappings generation unit 125, the second absolute value detecting unit 126 and inner mark generation unit 127.Though relevant drawings is expressed as the line that does not comprise input signal Gr, Gb, R and B with signal flow, obviously, should be appreciated that to those skilled in the art to comprise various other data.
Intermediate value detecting unit 121 is determined each intermediate value (being called as M1, M2, M3 and M4) of 3 * 3 signal Gr, R, B and the Gb of input.
For example, under the situation of Gr, the first absolute value detecting unit 122 detects the absolute value Abs_Gri that deducts the obtained value of M1 from each element Gri (i=1～9).Similarly for R, B and Gb, the first absolute value detecting unit 122 detects absolute value Abs_Ri, Abs_Bi and the Abs_Gbi that deducts the obtained value of M2, M3 and M4 from each element Ri, Bi and Gbi (i=1～9) respectively.Calculate Abs_Gri, Abs_Ri, Abs_Bi and Abs_Gbi by following formula 1.
[formula 1]
Abs_Gri＝abs(Gri-M1)
Abs_Ri＝abs(Ri-M2)
Abs_Bi＝abs(Bi-M3)
Abs_Gbi＝abs(Gbi-M4) i＝1～9
Difference detecting unit 123 detects absolute value Diff1_i, Diff2_i, the Diff3_i that deducts the value of Abs_Gri, Abs_Ri and Abs_Gbi from Abs_Bi.Abs_Bi is the dateout of the first absolute value detecting unit 122 at the data B with center pixel B5.Abs_Gri, Abs_Ri and Abs_Gbi are the dateout of the first absolute value detecting unit 122 at remainder data.By following formula 2 this is represented.
[formula 2]
Diff1_i＝abs(Abs_Bi-Abs_Gri)
Diff2_i＝abs(Abs_Bi-Abs_Ri)
Diff3_i＝abs(Abs_Bi-Abs_Gbi) i＝1～9
Comparison value detecting unit 124 generates the value COMP_i to Diff1_i, Diff2_i, Diff3_i summation, generates trappings, by following formula 3 this is represented.
[formula 3]
COMP_i＝Diff1_i+Diff2_i+Diff3_i i＝1～9
Fig. 5 illustration aforementioned processing.The example of the work of Fig. 5 is the illustration intermediate value detecting unit of Fig. 4, the first absolute value detecting unit, difference detecting unit and comparison value.
Trappings generation unit 125 compares the comparison value of above detection with bad some threshold value dead_threshold and focus threshold value hot_threshold, and generates and whether have the bad some sign relevant with focus.Here, dead_threshold and hot_threshold are adjustable.
In other words, trappings generation unit 125 compares the comparison value and the dead_threshold of each element.Under situation about existing greater than the comparison value of dead_threshold, if i is 5 (that is, the comparison value of center pixel is greater than dead_threshold), then trappings generation unit 125 will with whether exist the sign dead_flag of bad spot correlation to be set at 1.Otherwise sign dead_flag is 0.
Similarly, trappings generation unit 125 compares the comparison value and the hot_threshold of each element.Under situation about existing greater than the comparison value of hot_threshold, if i is 5 (that is, the comparison value of center pixel is greater than dead_threshold), then trappings generation unit 125 will with whether exist the relevant sign hot_flag of focus to be set at 1.Otherwise sign hot_flag is 0.
Then, if dead_flag or hot_flag are 1, then trappings generation unit 125 is set at 1 with trappings external_flag.Otherwise (that is, being under 0 the situation at dead_flag and hot_flag), trappings generation unit 125 is set at 0 with trappings external_flag, these trappings is outputed to the correcting unit 130 of Fig. 1.
The second absolute value detecting unit 126 detects the absolute value Abs_Ci that deducts (not comprising B5) the obtained value of this center pixel value from each element of the data B with center pixel B5.By following formula 4 this is represented.And, Fig. 6 illustration this process.The example of the work of the second absolute value detecting unit of Fig. 6 is illustration Fig. 4.
[formula 4]
Abs_Ci=abs (Bi-BS) (i=1～9 do not comprise 5)
Inner mark generation unit 127 is used for the threshold value Y_Threshold that the brightness that is applied is classified and intermediate value M1, M2, M3 and the M4 of 3 * 3 data, sets the threshold value thr of inner mark.Inner mark generation unit 127 utilizes threshold value thr to generate inner mark.Threshold value Y_Threshold can adjust according to system.
When the arithmetic mean with the intermediate value id of each component is defined as Y_m (that is,
The time, if Y_m less than Y_Threshold, then inner mark generation unit 127 is set at 0 with thr.Otherwise inner mark generation unit 127 is set at 50 with thr.Yet, being not limited to this situation, this numerical value can be adjusted according to system.Inner mark generation unit 127 compares the thr that sets with the absolute value of each element of the second absolute value detecting unit, 126 generations.As long as any one of 8 values is less than thr, inner mark generation unit 127 will be set at 0 with inner mark.Otherwise (that is, each absolute value of all elements is all greater than thr), inner mark generation unit 127 are set at 1 with inner mark, and this inner mark is sent to correcting unit 130.
If trappings and inner mark all are 1, then correcting unit 130 is exported the intermediate value of 3 * 3 data B.Otherwise, correcting unit 130 former states output B5.
In other words, though above description supposition center pixel is that B5 proofreaies and correct bad point, the present invention can be applicable to the situation that center pixel is any component.
Fig. 7 is an illustration imports the example of the original image of device of processing dead pixel of the present invention, and Fig. 8 has been an illustration by detecting the example of the image after relevant bad point is proofreaied and correct among Fig. 7.
As shown in Figure 7 and Figure 8, detect and proofread and correct bad point and focus.
Fig. 9 is an illustration according to the block diagram of the execution mode of the device of processing dead pixel of second embodiment of the invention.
As shown in Figure 9, device of processing dead pixel of the present invention can comprise component separative element 910, bad some detecting unit 920 and focus detecting unit 930.
As shown in Figure 9, component separative element 910 will be imported 3 * 3 Baeyer mode images and be divided into as shown in Figure 3 four 3 * 3 components.As mentioned above, Fig. 2 can be the execution mode of 6 * 6 Baeyer mode images of input Fig. 9, and Fig. 3 is an illustration is divided into the example of 3 * 3 components with the Baeyer mode image of Fig. 2.
As shown in Figures 2 and 3, component separative element 910 of the present invention is divided into 6 * 6 Baeyer mode images only have same components 3 * 3 data of (for example, being respectively green (R) of red (R), GR capable (Gr), G and the indigo plant (B) of GB capable (Gb)).For convenience, as shown in Figure 3,3 * 3 components of Gr, R, B and Gb are called as C1, C2, C3 and C4 respectively.
Bad some detecting unit 920 calculates the pattern (hereinafter, the pixel value that is called based on intermediate value distributes) of each component data; The pattern that compares other components; Calculate measured value according to each component; And utilize measured value to determine whether related pixel is bad point.With reference to relevant drawings this is described in detail.
The badly detailed diagram of a detecting unit of Fig. 9 that Figure 10 has been an illustration.
As shown in figure 10, bad some detecting unit 920 can comprise intermediate value detecting unit 921, absolute value detecting unit 922, comparison value detecting unit 923, threshold setting unit 924 and output determining unit 925.Though relevant drawings is expressed as the line that does not comprise input signal C1, C2, C3 and C4 with signal flow, obviously, should be appreciated that to those skilled in the art to comprise various other data.
Intermediate value detecting unit 921 is determined each intermediate value (being called as M1, M2, M3 and M4 respectively) of input 3 * 3 data C1, C2, C3 and C4.
For example, under the situation of C1, absolute value detecting unit 922 detects the absolute value Abs_1i that deducts the value of M1 from each Elements C 1i (i=1～9).Similarly for C2, C3 and C4, the first absolute value detecting unit 922 detects absolute value Abs_2i, Abs_3i and the Abs_4i that deducts the value of M2, M3 and M4 from each Elements C 2i, C3i and C4i (i=1～9) respectively.Calculate Abs_1i, Abs_2i, Abs_3i and Abs_4i by following formula 5.
[formula 5]
Abs_1i＝abs(Gri-M1)
Abs_2i＝abs(Ri-M2)
Abs_3i＝abs(Bi-M3)
Abs_4i＝abs(Gbi-M4)
Wherein i is the natural number between 1 to 9, comprises 1 and 9.
The dateout of comparison value detecting unit 923 calculating from each dateout that absolute value detecting unit 922 calculates deducts the absolute value of the obtained value of other each dateouts respectively, these absolute values sued for peace calculate this comparison value.In other words, comparison value detecting unit 923 calculates each absolute value that deducts the value of Abs_2i, Abs_3i and Abs_4i from Abs_1i, and to these absolute value summations.Similarly, comparison value detecting unit 923 calculates COMP_2i, COMP_3i and COMP_4i.By following formula 6 this is represented.
[formula 6]
COMP?1i＝abs(Abs_1i-Abs_2i)-abs(Abs_1i-Abs_3i)+abs(Abs_1i-Abs_4i)
COMP_2i＝abs(Abs_2i-Abs_1i)+abs(Abs_2i-Abs_3i)+abs(Abs_2i-Abs_4i)
COMP_3i＝abs(Abs_3i-Abs_1i)+abs(Abs_3i-Abs_2i)-abs(Abs_3i-Abs_4i)
COMP_4i＝abs(Abs_4i-Abs_1i)+abs(Abs_4i-Abs_2i)+abs(Abs_4i-Abs_3i)
Wherein i is the natural number between 1 to 9, comprises 1 and 9.The example of the work of Figure 11 is the illustration intermediate value detecting unit of Figure 10, absolute value detecting unit and comparison value detected value.
Each of 4 comparison values that output determining unit 925 detects above comparison value detecting unit 923 compares with a bad some threshold value dead_threshold who uses.If, then do not export original image greater than the comparison value of dead_threshold.In other words, former state output 3 * 3 image C j (j=1～4).Figure 12 is an illustration from the example of the image of the bad some detecting unit output of Fig. 9.As shown in figure 12, bad some detecting unit 920 output 3 * 3 image C j (j=1～4).Correspondingly, the data from 925 outputs of output determining unit are original 6 * 6 images.In Figure 12, " Size3 * 3_Cj1 " refers to first element of 3 * 3 data Cj.Below describe and to use 3 * 3 data Cj.
If only there is a element, then exports the pixel of determining unit 925 by each element of polishing and generate image once more greater than dead_threshold.
For polishing data Size3 * 3_Cj1 (first element of 3 * 3 data Cj, j=1～4), output determining unit 925 will be from data Size3 * 3_Cj1 (promptly, first element) (that is, other data Size3 * 3_Cj2, Size3 * 3_Cj5 and Size3 * 3_Cj4) each absolute value and the threshold value thr of obtained value compare among Figure 12 to deduct adjacent element.As long as an absolute value is arranged less than this thr, with regard to dateout Size3 * 3_Cj1.Otherwise, export the intermediate value of other data Size3 * 3_Cj2, Size3 * 3_Cj5 and Size3 * 3_Cj4.
In addition, in order to compensate Size3 * 3_Cj2 (second element of 3 * 3 data Cj, j=1～4), output determining unit 925 will be from data Size3 * 3_Cj2 (promptly, second element) (that is, other data Size3 * 3_Cj1, Size3 * 3_Cj3, Size3 * 3_Cj4, Size3 * 3_Cj5 and Size3 * 3_Cj6) each absolute value and the threshold value thr of obtained value compare among Figure 12 to deduct adjacent element.As long as an absolute value is arranged less than this thr, with regard to dateout Size3 * 3_Cj2.Otherwise, export the intermediate value of other data Size3 * 3_Cj1, Size3 * 3_Cj3, Size3 * 3_Cj4, Size3 * 3_Cj5 and Size3 * 3_Cj6.Like this, each output valve of all elements of specified data Size3 * 3_Cj (j=1～4) and be sent to the focus determining unit 930 of Fig. 9.
Focus determining unit 930 is calculated the pattern of each component and is then relatively calculated measured value based on the pattern of other components.Focus determining unit 930 utilizes this measured value to determine whether respective pixel is focus.Hereinafter, describe focus determining unit 930 in detail with reference to relevant drawings.
The detailed diagram of the execution mode of the focus detecting unit of Figure 13 is illustration Fig. 9.
As shown in figure 13, focus detecting unit 930 can comprise intermediate value detecting unit 931, absolute value detecting unit 932, comparison value detecting unit 933 and output determining unit 934.Though relevant drawings is expressed as the line that does not comprise input signal Si ze3 * 3C1, Size3 * 3C2, Size3 * 3C3 and Size3 * 3C4 with signal flow, obviously, should be appreciated that to those skilled in the art to comprise various other data.
Intermediate value detecting unit 931 is determined each intermediate value (being called as M5, M6, M7 and M8 respectively) of input 3 * 3 data C1,3 * 3 data C2,3 * 3 data C3 and 3 * 3 data C4.Yet intermediate value M5, M6, M7 and the M8 that intermediate value detecting unit 931 detects also can be identical with the intermediate value that the intermediate value detecting unit of going bad some detecting unit 920 921 of Figure 10 detects.
For example, under the situation of Size3 * 3C1, absolute value detecting unit 932 detects the absolute value Abs_5i that deducts the obtained value of M5 from each elements Si ze3 * 3C1 (i=1～9).Similarly for Size3 * 3C2i, Size3 * 3C3i and Size3 * 3C4i, the first absolute value detecting unit 932 detects absolute value Abs_6i, Abs_7i and the Abs_8i that deducts the obtained value of M6, M7 and M8 from each elements Si ze3 * 3C2i, Size3 * 3C3i and Size3 * 3C4i (i=1～9) respectively respectively.Calculate Abs_5i, Abs_6i, Abs_7i and Abs_8i by following formula 7.
[formula 7]
Abs_5i＝abs(Size3×3C1i-M5)
Abs_6i＝abs(Size3×3C2i-M6)
Abs_7i＝abs(Size3×3C3i-M7)
Abs_8i＝abs(Size3×3C4i-M8)
Wherein i is the natural number between 1 to 9, comprises 1 and 9.
Comparison value detecting unit 933 calculates the absolute value that a dateout from each dateout that absolute value detecting unit 932 calculates deducts the obtained value of other dateouts respectively, and absolute value is sued for peace calculate comparison value.In other words, comparison value detecting unit 933 calculates each absolute value that deducts the obtained value of Abs_6i, Abs_7i and Abs_8i from Abs_5i respectively, and to these absolute values summations.Similarly, comparison value detecting unit 933 calculates COMP_6i, COMP_7i and COMP_8i.Represent this computing by following formula 8.
[formula 8]
COMP_5i＝abs(Abs_5i-Abs_6i)+abs(Abs_5i-Abs_7i)-abs(Abs_5i-Abs_8i)
COMP_6i＝abs(Abs_6i-Abs_5i)+abs(Abs_6i-Abs_7i)+abs(Abs_6i-Abs_8i)
COMP_7i＝abs(Abs_7i-Abs_5i)+abs(Abs_7i-Abs_6i)+abs(Abs_7i-Abs_8i)
COMP_8i＝abs(Abs_8i-Abs_5i)+abs(Abs_8i-Abs_6i)+abs(Abs_8i-Abs_7i)
Wherein i is the natural number between 1 to 9, comprises 1 and 9.
If only there is a element, then exports the pixel of determining unit 934 by each element of polishing and generate image once more greater than dead_threshold.
For polishing data Size3 * 3_Cj1 (first element of 3 * 3 data Cj, j=1～4), output determining unit 934 will be from data Size3 * 3_Cj1 (promptly, first element) (that is, the threshold value determining unit 924 of each absolute value of the obtained value of other data Size3 * 3_Cj2, Size3 * 3_Cj5 and Size3 among Figure 12 * 3_Cj4) and threshold cell 920 is determined and the threshold value thr that send compare to deduct adjacent element.As long as an absolute value is arranged less than this thr, with regard to dateout Size3 * 3_Cj1.Otherwise export the intermediate value of other data Size3 * 3_Cj2, Size3 * 3_Cj5 and Size3 * 3_Cj4.
In addition, for offset data Size3 * 3_Cj2 (second element of 3 * 3 data Cj, j=1～4), output determining unit 925 will be from data Size3 * 3_Cj2 (promptly, second element) (that is, other data Size3 * 3_Cj1, Size3 * 3_Cj3, Size3 * 3_Cj4, Size3 * 3_Cj5 and Size3 * 3_Cj6) each absolute value and the threshold value thr of obtained value compare among Figure 12 to deduct adjacent element.As long as an absolute value is arranged less than this thr, with regard to dateout Size3 * 3_Cj2.Otherwise export the intermediate value of other data Size3 * 3_Cj1, Size3 * 3_Cj3, Size3 * 3_Cj4, Size3 * 3_Cj5 and Size3 * 3_Cj6.Like this, can specified data Size3 * 3_Cj (j=1～4) thus each output valve dateout Size3 * 3_Cj (j=1～4) of all elements.Therefore, output 6 * 6 data.
As mentioned above, Fig. 7 can be an illustration imports the example of the original image of device of processing dead pixel of the present invention, and Fig. 8 can be an illustration by detecting the example of the image after relevant bad point is proofreaied and correct among Fig. 7.Because second execution mode can provide identical with first embodiment of the invention or similar image, therefore omit associated description or figure.
In other words, shown in relevant drawings, can come correcting image by detecting corresponding bad point and focus.
As mentioned above, the present invention can compare by pattern and other components that will have the component of center pixel in 6 * 6 masks (mask), determines accurately whether pixel is bad point or focus, and exports the correcting image of corresponding 1 pixel.
In addition, the present invention can compare by pattern and other components that will have the component of center pixel in 6 * 6 masks, determines that accurately whether pixel is bad point or focus, and 6 * 6 pixel values of output after proofreading and correct for relevant bad point.
Accompanying drawing and detailed description are example of the present invention, only are used to describe the present invention and never are to limit or the restriction the spirit and scope of the present invention.Therefore, those of ordinary skill in the art is to be understood that and also exists a large amount of the change with other to be equal to execution mode.True scope of the present invention must a spirit by appended claim limit.
Claims (17)
1, a kind of device of processing dead pixel, this device of processing dead pixel comprises:
Separative element, its Baeyer mode image with input is divided into corresponding component data; And
Comparing unit, its intermediate value based on each data is come the distribution (pattern) of calculating pixel, calculate comparison value based on component with center pixel, utilize this comparison value to come output notice whether to have the trappings of bad point and/or focus, the value of the center pixel in the component with center pixel and the value of neighbor are compared; Calculate measured value based on this center pixel; And utilize this measured value to export inner mark, wherein
This comparing unit comprises:
The central value detecting unit, it detects each intermediate value in the described component data;
The first absolute value detecting unit, it detects the absolute value that deducts the obtained value of described intermediate value from each element;
The difference detected value, it detects difference, that is, deduct the absolute value of this first absolute value detecting unit at the dateout of component data at each obtained value of the dateout of the component data except that described component data with center pixel with center pixel from this first absolute value detecting unit;
The comparison value detecting unit, its output to this difference detecting unit is sued for peace, and at the comparison value of each element detected components data;
The trappings generation unit, it compares this comparison value and each bad some threshold value and focus threshold value, and generates and whether have a bad threshold value and/or the relevant trappings of focus threshold value;
The second absolute value detecting unit, it detects the absolute value that deducts the obtained value of this center pixel value from each element of the component data with center pixel; And
The inner mark generation unit, it utilizes the threshold value of brightness classification usefulness and the intermediate value of this component data to set the inner mark threshold value, and utilizes this threshold value to generate inner mark.
2, device of processing dead pixel according to claim 1, this device of processing dead pixel further comprises correcting unit, the output of this correcting unit utilizes the pixel data of 1 pixel after the trappings that receive from described comparing unit and inner mark are proofreaied and correct.
3, device of processing dead pixel according to claim 2, if wherein determine to exist bad point/focus, then described correcting unit has the intermediate value of the component data of described center pixel with reference to described trappings and described inner mark output device.
4, device of processing dead pixel according to claim 2, if wherein determine not bad point/focus, then described correcting unit is exported described center pixel with reference to described trappings and described inner mark.
5, device of processing dead pixel according to claim 1, wherein when being compared, the comparison value of every element and described bad some threshold value having only under the situation of comparison value greater than bad some threshold value of described center pixel, and/or when the comparison value of every element and focus threshold value are compared, having only under the situation of comparison value greater than the focus threshold value of described center pixel, described trappings generation unit is set at trappings and has bad point/focus.
6, device of processing dead pixel according to claim 1, wherein said inner mark generation unit is sued for peace to the intermediate value of each component data; Threshold value according to described threshold setting inner mark; And at the absolute value of all every elements that the described second absolute value detecting unit calculates all under the situation greater than the threshold value of described inner mark, described inner mark is set at there is bad point/focus.
7, a kind of image processor that is used for imaging device of handling bad point, described image processor comprises:
Separative element, its Baeyer mode image with input is divided into corresponding component data;
The intermediate value detecting unit, it detects each intermediate value of described detected components data;
The first absolute value detecting unit, it detects the absolute value that deducts the obtained value of described intermediate value from each element of described component data;
The difference detected value, it detects difference, that is, deduct the absolute value of the described first absolute value detecting unit at the dateout of component data at each obtained value of the dateout of the component data except that described component data with center pixel with center pixel from the described first absolute value detecting unit;
The comparison value detecting unit, its output to described difference detecting unit is sued for peace, and at the comparison value of each element detected components data;
The trappings generation unit, it compares described comparison value and each bad some threshold value and focus threshold value, and generates and whether have a bad threshold value and/or the relevant trappings of focus threshold value;
The second absolute value detecting unit, it detects the absolute value that deducts the obtained value of described center pixel value from each element of the component data with center pixel; And
The inner mark generation unit, it utilizes the threshold value of brightness classification usefulness and the intermediate value of described component data to set the inner mark threshold value, and utilizes this threshold value to generate inner mark; And
Correcting unit, its output utilize respectively the pixel value after the trappings that receive from described trappings generation unit and described inner mark generation unit and inner mark are proofreaied and correct.
8, a kind of device of processing dead pixel, described device of processing dead pixel comprises:
Separative element, it will be imported the Baeyer mode image and be divided into corresponding component data; And
Bad some detecting unit, it is based on the distribution (pattern) of the median calculation pixel of each data; The pattern that compares other components; Calculate comparison value based on each component; Described comparison value and bad some threshold value are compared; Determine whether corresponding pixel is bad point; And export calibrated component data.
9, device of processing dead pixel according to claim 8, wherein said bad some detecting unit comprises:
The first central value detecting unit, it detects each intermediate value of described component data;
The first absolute value detecting unit, it detects the absolute value that deducts the obtained value of described intermediate value from each element;
First comparing unit, it calculates each absolute value that deducts the obtained value of first output from second output of the first absolute value detecting unit; Described absolute value is sued for peace; And at each component and each Data Detection comparison value;
The threshold setting unit, it utilizes the intermediate value of the threshold value of brightness classification usefulness and the detected component data of the first intermediate value detecting unit to set the threshold value of inner mark; And
The first output determining unit, it compares detected comparison value of the described first comparison value detecting unit and bad some threshold value, and according to whether existing bad point to determine dateout.
10, device of processing dead pixel according to claim 9, the wherein said first output determining unit will the detected comparison value of the described first comparison value detecting unit compares with a bad some threshold value, and if not greater than the comparison value of a described bad threshold value then export described component data.
11, device of processing dead pixel according to claim 9, the wherein said first output determining unit compares described detected comparison value of the first comparison value detecting unit and described bad some threshold value, if and had a comparison value threshold value to go bad the some threshold value then each element of definite described component data greater than described, simultaneously
The described first output determining unit will compare from element-specific absolute value that deducts the obtained value of adjacent element and the threshold value of being determined by described threshold value determining unit, if and all absolute values then are defined as this element the intermediate value of adjacent element all greater than described threshold value.
12, device of processing dead pixel according to claim 9, the wherein said first output determining unit compares described detected comparison value of the first comparison value detecting unit and described bad some threshold value, if and had a comparison value to go bad the some threshold value then each element of definite described component data greater than described, simultaneously
The described first output determining unit will deduct the absolute value of the obtained value of adjacent element from element-specific and compare with threshold value that described threshold value determining unit is determined, if not greater than the absolute value of this threshold value, then this coherent element will be defined as exporting.
13, according to Claim 8 with 9 in wherein described device of processing dead pixel, described device of processing dead pixel further comprises the focus detecting unit, described focus detecting unit calculates distribution (pattern) according to the component data that receives from described bad some detecting unit, calculates comparison value based on each component; Described comparison value and focus threshold value are compared; Determine whether corresponding pixel is focus; And export calibrated component data.
14, device of processing dead pixel according to claim 13, wherein said focus detecting unit comprises:
The second central value detecting unit, it detects each intermediate value of the component data that receives from described bad some detecting unit;
The second absolute value detecting unit, it detects the absolute value that deducts the obtained value of the detected intermediate value of the described second intermediate value detecting unit from each element;
Second comparing unit, it calculates each absolute value that deducts the obtained value of first output from second output of the described second absolute value detecting unit; Described absolute value is sued for peace; And at each component and each Data Detection comparison value; And
The second output determining unit, it will be compared by described detected comparison value of the second comparison value determining unit and described focus threshold value, and according to whether existing focus to determine dateout.
15, device of processing dead pixel according to claim 14, the wherein said second output determining unit will be compared by the detected comparison value of the second comparison value detecting unit and described focus threshold value, and if less than greater than the comparison value of described focus threshold value then export described component data.
16, device of processing dead pixel according to claim 14, the wherein said second output determining unit will be compared by detected comparison value of the second comparison value detecting unit and described focus threshold value, and if a comparison value threshold value were arranged greater than described focus threshold value then determine each element of described component data, wherein
The described second output determining unit will deduct the absolute value of the obtained value of adjacent element and threshold value that described threshold value determining unit is determined compares from element-specific, if all described absolute values all greater than described threshold value, then are defined as this element the intermediate value of described adjacent element.
17, device of processing dead pixel according to claim 14, the wherein said second output determining unit compares detected comparison value of the second comparison value detecting unit and described focus threshold value, and if a comparison value threshold value were arranged greater than described focus threshold value then determine each element of described component data, simultaneously
The described second output determining unit will deduct the absolute value of the obtained value of adjacent element and threshold value that described threshold value determining unit is determined compares from element-specific, if and, then this coherent element is defined as output not greater than the absolute value of described threshold value.
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| KR1020050133602 | 2005-12-29 | ||
| KR1020050133602A KR100683415B1 (en) | 2005-12-29 | 2005-12-29 | Method for processing dead pixel |
| KR1020050133563 | 2005-12-29 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101352048A true CN101352048A (en) | 2009-01-21 |
Family
ID=38103804
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2006800499742A Pending CN101352048A (en) | 2005-12-29 | 2006-12-28 | Device of processing dead pixel |
Country Status (2)
| Country | Link |
|---|---|
| KR (1) | KR100683415B1 (en) |
| CN (1) | CN101352048A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102576455A (en) * | 2009-07-02 | 2012-07-11 | 斯基恩特-X公司 | Reduction of noise caused by absorption of X-ray photons in pixelated image sensors |
| CN104935838A (en) * | 2015-06-04 | 2015-09-23 | 上海集成电路研发中心有限公司 | Image restoration method |
| CN105516694A (en) * | 2015-12-08 | 2016-04-20 | 上海集成电路研发中心有限公司 | Method and system for avoiding dark corner of lens |
| CN105718849A (en) * | 2015-10-21 | 2016-06-29 | 深圳芯启航科技有限公司 | Pixel point scanning method and apparatus applied to fingerprint sensor |
| CN107016670A (en) * | 2017-03-27 | 2017-08-04 | 福州瑞芯微电子股份有限公司 | A kind of dead pixel points of images detection method and device |
| CN107613229A (en) * | 2017-09-30 | 2018-01-19 | 杭州雄迈集成电路技术有限公司 | A kind of dead pixels of image sensor surveys means for correcting and method |
| CN110035281A (en) * | 2019-04-29 | 2019-07-19 | 昆山丘钛微电子科技有限公司 | A kind of dead pixel detection method, device and electronic equipment |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100906606B1 (en) * | 2007-06-28 | 2009-07-09 | 엠텍비젼 주식회사 | Method and apparatus for processing dead pixel |
| KR101114586B1 (en) | 2010-02-08 | 2012-03-02 | 삼성전자주식회사 | Apparatus and method for removing of defect pixel |
| KR101175737B1 (en) * | 2010-12-14 | 2012-08-21 | 엠텍비젼 주식회사 | Defect pixel processing method and device |
| KR102193913B1 (en) * | 2019-06-27 | 2020-12-22 | 한화시스템 주식회사 | Apparatus for detecting soft defective pixels in infrared camera for detection of acquired bad pixels |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2000217039A (en) | 1999-01-21 | 2000-08-04 | Sanyo Electric Co Ltd | Point defect detection method and point defect pixel value correction method |
| JP4453332B2 (en) * | 2003-10-17 | 2010-04-21 | ソニー株式会社 | Defective pixel detection method, detection apparatus, and imaging apparatus |
| KR100607803B1 (en) * | 2004-04-27 | 2006-08-02 | 매그나칩 반도체 유한회사 | Method for color interpolation of image sensor |
| JP2005341020A (en) | 2004-05-25 | 2005-12-08 | Matsushita Electric Ind Co Ltd | Pixel defect detector and pixel defect correction device |
-
2005
- 2005-12-29 KR KR1020050133602A patent/KR100683415B1/en active IP Right Grant
-
2006
- 2006-12-28 CN CNA2006800499742A patent/CN101352048A/en active Pending
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102576455A (en) * | 2009-07-02 | 2012-07-11 | 斯基恩特-X公司 | Reduction of noise caused by absorption of X-ray photons in pixelated image sensors |
| CN102576455B (en) * | 2009-07-02 | 2015-05-20 | 斯基恩特-X公司 | Reduction of noise caused by absorption of X-ray photons in pixelated image sensors |
| CN104935838A (en) * | 2015-06-04 | 2015-09-23 | 上海集成电路研发中心有限公司 | Image restoration method |
| CN104935838B (en) * | 2015-06-04 | 2018-04-06 | 上海集成电路研发中心有限公司 | A kind of method of image restoring |
| CN105718849A (en) * | 2015-10-21 | 2016-06-29 | 深圳芯启航科技有限公司 | Pixel point scanning method and apparatus applied to fingerprint sensor |
| CN105516694A (en) * | 2015-12-08 | 2016-04-20 | 上海集成电路研发中心有限公司 | Method and system for avoiding dark corner of lens |
| CN107016670A (en) * | 2017-03-27 | 2017-08-04 | 福州瑞芯微电子股份有限公司 | A kind of dead pixel points of images detection method and device |
| CN107016670B (en) * | 2017-03-27 | 2019-06-28 | 福州瑞芯微电子股份有限公司 | A kind of dead pixel points of images detection method and device |
| CN107613229A (en) * | 2017-09-30 | 2018-01-19 | 杭州雄迈集成电路技术有限公司 | A kind of dead pixels of image sensor surveys means for correcting and method |
| CN110035281A (en) * | 2019-04-29 | 2019-07-19 | 昆山丘钛微电子科技有限公司 | A kind of dead pixel detection method, device and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| KR100683415B1 (en) | 2007-02-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101352048A (en) | Device of processing dead pixel | |
| US8131111B2 (en) | Device of processing dead pixel | |
| US9626760B2 (en) | System and method to align and merge differently exposed digital images to create a HDR (High Dynamic Range) image | |
| CN109068025B (en) | Lens shadow correction method and system and electronic equipment | |
| CN104296876B (en) | Scene asymmetric correction method and its device based on FPGA | |
| CN103389310B (en) | Online sub-pixel optical component damage detection method based on radiation calibration | |
| CN102907103A (en) | Image processing device, image processing method and program | |
| CN108109147A (en) | A kind of reference-free quality evaluation method of blurred picture | |
| AU2008261138A1 (en) | Measure display SFR using a camera and phase shifting | |
| CN104732531B (en) | A kind of high-resolution remote sensing image signal to noise ratio curve self-adapting acquisition methods | |
| CN106686281B (en) | Fuse circuit board noise suppression ability test system | |
| CN105869129B (en) | For the thermal infrared images residue non-uniform noise minimizing technology after nonuniformity correction | |
| CN102236790A (en) | Image processing method and device | |
| CN107392948B (en) | Image registration method of amplitude-division real-time polarization imaging system | |
| CN113781375B (en) | Vehicle-mounted vision enhancement method based on multi-exposure fusion | |
| US20140355900A1 (en) | Degradation restoration system, degradation restoration method and program | |
| CN101567965A (en) | Image processing apparatus | |
| CN113822809A (en) | Dim light enhancement method and system | |
| Lluis-Gomez et al. | Chromatic aberration correction in RAW domain for image quality enhancement in image sensor processors | |
| KR100696167B1 (en) | Method for processing dead pixel | |
| CN104751159B (en) | Video coloured silk field detection method and device | |
| CN102376083A (en) | Cross-color image processing systems and methods for sharpness enhancement | |
| KR100740625B1 (en) | Method for sharing line memory in image processing system | |
| CN113724202B (en) | Image sensor correction effect quantitative evaluation method based on one-dimensional Fourier transform | |
| CN109919894A (en) | A kind of non-reference picture quality appraisement method and system based on human visual system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Open date: 20090121 |