CN101527780B - 2D image segmentation apparatus and method, and apparatus and method for removing red eye from image - Google Patents
2D image segmentation apparatus and method, and apparatus and method for removing red eye from image Download PDFInfo
- Publication number
- CN101527780B CN101527780B CN2008101815752A CN200810181575A CN101527780B CN 101527780 B CN101527780 B CN 101527780B CN 2008101815752 A CN2008101815752 A CN 2008101815752A CN 200810181575 A CN200810181575 A CN 200810181575A CN 101527780 B CN101527780 B CN 101527780B
- Authority
- CN
- China
- Prior art keywords
- pixel
- pixel groups
- groups
- sets
- adjacent pixels
- 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.)
- Expired - Fee Related
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/36—Applying a local operator, i.e. means to operate on image points situated in the vicinity of a given point; Non-linear local filtering operations, e.g. median filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/10—Image acquisition
- G06V10/12—Details of acquisition arrangements; Constructional details thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
- G06V10/751—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Multimedia (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Databases & Information Systems (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Nonlinear Science (AREA)
- Image Processing (AREA)
Abstract
The present invention relates to a 2D image segmentation apparatus and method, and apparatus and method for removing red eye from image. A Two-Dimensional (2D) image segmentation apparatus for segmenting pixels of a progressive input 2D image includes a group information storing unit storing information of pixel groups including a plurality of adjacent pixels; a pixel determining unit determining coordinates of an input pixel, and determining whether the input pixel is an effective pixel for segmentation; a group scanning unit scanning a adjacent pixel group disposed in a scan range preset from the effective pixel in the group storing unit when the input pixel is determined as the effective pixel by the pixel determining unit; and a group information updating unit updating information of a pixel group stored in the group information storing unit according to whether there is the scanned adjacent pixel group and a pixel group including an input pixel preceding the effective pixel. Moreover, a method is provided that removes red-eye using area information of a pixel group determined by the segmentation apparatus.
Description
The application requires to be submitted on March 7th, 2008 the 2008-21236 korean patent application of Korea S Department of Intellectual Property and the priority of the 2008-44444 korean patent application submitted on May 14th, 2008; Disclosing of this application is completely integrated in this, for reference.
Technical field
The present invention relates to digital image processing field; More specifically; Relate to a kind of two dimension (2D) image segmentation apparatus and method of cutting apart the pixel of input picture line by line with the AD HOC of detected image, and a kind of equipment and method of blood-shot eye illness that is used for removing the 2D image of removing blood-shot eye illness in the cut zone.
Background technology
Recently, the mode identification technology through image processing identification given shape is applied to various industrial fields.The example of this mode identification technology comprises known mode identification technology, such as detecting the eyelid shake in case prevent fatigue driving PRS, from through scanner scanning to the character recognition system and the fingerprint recognition system of image recognition character.
Such mode identification technology needs a kind of pixel that cut apart entire image so that the technology of the shape that will discern from the entire image detection of generation such as camera.
Usually, the partitioning algorithm that in mode identification technology, uses comprises: first analyzing and processing, the characteristic of the part of the input picture that analysis will detect from input picture; And second analyzing and processing, input picture is converted into the binary picture of only representing customizing messages and uses the shape information of the part of the input picture that will detect through first analyzing and processing that said binary picture is identified as one group subsequently.For example; The pattern recognition process of identification eye areas that is used for removing the blood-shot eye illness of the image that camera takes is such processing: produce the binary picture of the redness in the image that only expression takes through first analyzing and processing, and each pixel of cutting apart said binary picture through second analyzing and processing is organized accordingly confirming with shape of eyes.In such pattern recognition process, can carry out partitioning algorithm.
The example of the image segmentation algorithm of prior art comprises: meadow fire (grass-fire) algorithm that detecting pattern is cut apart with execution on eight directions of a point, and in advance input pattern information and use the algorithm with the correlation detection pattern of input picture.Yet, under the situation of existing partitioning algorithm, because the operating time is very slow and need complicated operations, so must use microprocessor to carry out operational processes.Therefore, under the situation of the partitioning algorithm of prior art, cannot be to the pixel executive real-time dividing processing of input picture line by line, and because expensive microprocessor causes cost to increase.In addition, be difficult to the partitioning algorithm of prior art need to be applied to the device (such as portable phone) of miniaturization.
On the other hand, the image segmentation algorithm of prior art can be applicable to remove the technology of the blood-shot eye illness in the digital picture.Just, use the partitioning algorithm of relevant art that the blood-shot eye illness appearance zone execution in two dimension (2D) digital picture is cut apart, and can use the coordinate of cut zone to use blood-shot eye illness removal algorithm.
Usually, the example of removing the prior art of blood-shot eye illness comprises: substitute that the scheme in zone appears in blood-shot eye illness and through covering the scheme of application drawing picture with particular color.
The zone that the definite input of the scheme pixel of replacing with particular color belongs to, and check when the input pixel is arranged in operated zone that subsequently the color of respective pixel is in skin-coloured regions or and the corresponding color of blood-shot eye illness.At this moment, be during when check result shows the color of respective pixel with the corresponding color of blood-shot eye illness, replace corresponding pixel with black.Because this scheme needs a spot of operation and small memory space, therefore need very fast service speed and small memory space.Yet when using the particular color replacement to confirm as the pixel of blood-shot eye illness, because eye color becomes dullness, so the fact of image manipulation can be by clearly performance.
Based on the scheme of covering is such scheme: through being applied to make color near black in operated zone with the covering of big or small identical size in the zone with low coefficient.Owing to can make eye color natural based on the scheme of covering, but therefore reduce the true possibility of user's identification colors operation.Yet, need store space based on the scheme of covering with operated zone, owing to stored information along with the size with operated zone increases, thereby the size of storage space increases.In addition, under the situation based on the scheme of covering, owing to the size variation with operated zone, it also must be variable therefore covering.In addition, owing to become slowly, therefore needing to be not suitable for the system of high speed operation based on the scheme of covering along with operating area increases service speed.
Summary of the invention
One side of the present invention provides a kind of two dimension (2D) image segmentation apparatus and method, and said equipment and method can be cut apart each pixel executive real-time of 2D image of input line by line, and because the simplification of operation and do not need the additional hardware resource.
Another aspect of the present invention provides a kind of equipment and method that is used for removing the blood-shot eye illness of 2D image, and said equipment and method can be through carrying out the correction degree and come to remove more naturally blood-shot eye illness and the input pixel is used for the gain of correction in real time according to being included in locations of pixels in the blood-shot eye illness distinctively.
According to an aspect of the present invention, a kind of 2D image segmentation apparatus of the pixel of the 2D image of input line by line that is used to cut apart is provided, comprises: the group information memory cell, storage comprises the information of the pixel groups of a plurality of neighbors; Pixel is confirmed the unit, confirms the coordinate of input pixel, and confirms whether the input pixel is the valid pixel that is used to cut apart; The group scanning element when the input pixel is confirmed that by pixel valid pixel is confirmed as in the unit, scans being arranged in from the sets of adjacent pixels of the preset sweep limits of valid pixel in the group information memory cell; And the group information updating unit, according to whether existing sets of adjacent pixels that scans and the pixel groups that comprises valid pixel input pixel before to update stored in the information of the pixel groups in the group information memory cell.
According to a further aspect in the invention, a kind of 2D image partition method of the 2D image of input line by line that is used to cut apart is provided, comprises: confirm the coordinate of input pixel, and confirm whether the input pixel is the valid pixel that is used to cut apart; When input pixel when being confirmed as valid pixel, be arranged in from the sets of adjacent pixels of the preset sweep limits of valid pixel in the group information memory cell to the information of storage pixel group and scan; And according to whether existing sets of adjacent pixels that scans and the pixel groups that comprises valid pixel input pixel before to update stored in the information of the pixel groups in the group information memory cell.
According to a further aspect in the invention; A kind of equipment that is used for removing the blood-shot eye illness of 2D image is provided; Comprise: the reference length computing unit receives the coordinate information in the zone comprise the blood-shot eye illness in the entire image, and confirms the centre coordinate in said zone and to the reference length of centre coordinate; The length in pixels computing unit, order receives each pixel of image, and calculates the length in pixels as the distance between the center in each pixel and said zone; Interval confirm reference length and length in pixels to be compared the unit, select the pixel of length in pixels less than reference length, and according to the interval of the pixel of from the interval of being scheduled to, confirming to comprise selection to the distance of centre coordinate; And the image manipulation unit, be provided for the gain of each interval luminance gain and color, and use luminance gain and color gain pixel value with the pixel of revising selection.
According to a further aspect in the invention, a kind of method that is used for removing the blood-shot eye illness of 2D image is provided, comprises: the coordinate information that receives the zone that comprises the blood-shot eye illness in the entire image; The centre coordinate in the zone that confirm to receive and to the reference length of centre coordinate; Order receives each pixel of image, and calculates the length in pixels as the distance between the center in each pixel and said zone; Reference length and length in pixels are compared, select the pixel of length in pixels less than reference length, and according to the interval of the pixel of from predetermined interval, confirming to comprise selection to the distance of centre coordinate; And be provided for the gain of each interval luminance gain and color, and use luminance gain and color gain pixel value with the pixel of revising selection.
Description of drawings
Through the detailed description below in conjunction with accompanying drawing, above-mentioned and other aspects of the present invention, characteristic and other advantages will more be expressly understood, wherein:
Fig. 1 is the block diagram of two dimension (2D) image segmentation apparatus according to the embodiment of the invention;
Fig. 2 is the flow chart that illustrates according to the 2D image partition method of the embodiment of the invention;
Fig. 3 A and Fig. 3 B are the exemplary diagram that is used to describe the binary picture of using the embodiment of the invention;
Fig. 4 is the exemplary diagram of a divided pixel groups;
Fig. 5 illustrates a plurality of registers of the group information of 32 groups of storage;
Fig. 6 A, Fig. 6 B, Fig. 7 A, Fig. 7 B, Fig. 8 A and Fig. 8 B are the exemplary diagram of describing according to the embodiment of the invention of method that is used to be provided with the sets of adjacent pixels sweep limits;
Fig. 9 is the flow chart that is used for upgrading when scanning when sets of adjacent pixels the method for the group information that is stored in group information memory cell 11 in advance that illustrates according to the embodiment of the invention;
Figure 10 is the exemplary diagram that illustrates according to the type that is stored in a plurality of pixel groups in the group information memory cell of the embodiment of the invention;
Figure 11 is the block diagram of equipment of blood-shot eye illness that is used for removing the 2D image according to the embodiment of the invention;
Figure 12 is the block diagram of method of blood-shot eye illness that is used for removing the 2D image according to the embodiment of the invention;
Figure 13 is the flow chart of operation that illustrates in greater detail the correction pixel value of Figure 12;
Figure 14 is the concept map of the coordinate, reference length and the coordinate length calculation that are used for describing the embodiments of the invention application region;
Figure 15 illustrates the curve chart of the change rate when according to the relation between reference length and the coordinate length pixel of selecting being carried out correction in an embodiment of the present invention;
Figure 16 and Figure 17 are the curve charts that illustrates in an embodiment of the present invention according to the change of the locations of pixels luminance gain of selecting;
Figure 18 and Figure 19 are the curve charts that illustrates in an embodiment of the present invention according to the change of the locations of pixels color gain of selecting;
Figure 20 and Figure 21 are the curve charts of change that the luminance gain that adds bright processing in the interval that is applied to comprise centre coordinate in an embodiment of the present invention is shown;
Figure 22 is comprising the exemplary diagram of having used the eye image that adds bright processing on the interval of centre coordinate in an embodiment of the present invention.
Embodiment
Below will exemplary embodiment of the present invention be described in more detail with reference to accompanying drawing.Yet the present invention can implement with different form, and should not be understood that the embodiment that is limited in this elaboration.On the contrary, provide these embodiment to make that the disclosure will be thorough and complete, and scope of the present invention is fully conveyed to those skilled in the art.Therefore, notice that size of component illustrated in the accompanying drawings can be by exaggerative for clear.
Fig. 1 is the block diagram of two dimension (2D) image segmentation apparatus according to the embodiment of the invention.
With reference to Fig. 1, comprise according to the 2D image segmentation apparatus 10 of the embodiment of the invention: group information memory cell 11, storage comprises the information of the pixel groups of a plurality of neighbors; Pixel is confirmed unit 12, confirms the coordinate of the pixel of the 2D image of input line by line, and confirms whether the input pixel is the valid pixel that is used to cut apart; Group scanning element 13 when the input pixel is confirmed that by pixel valid pixel is confirmed as in unit 12, is arranged in the scanning group information memory cell 11 from the sets of adjacent pixels of the preset sweep limits of valid pixel; And group information updating unit 14, according to whether existing sets of adjacent pixels that scans and the pixel groups that comprises valid pixel input pixel before to update stored in the information of the pixel groups in the group information memory cell 11.
2D image segmentation apparatus 10 according to the embodiment of the invention also can comprise: stack features is confirmed unit 15; The pixel quantity that comprises in the pixel groups according to the renewal completion that is confirmed as group information in the pixel groups that is stored in the group information memory cell 11 confirms with the ratio of width to height of the pixel groups of the renewal completion that is confirmed as group information whether corresponding pixel groups is effective, and is confirmed as the group information of the pixel groups of inactive pixels group from 11 deletions of group information memory cell.
Fig. 2 is the flow chart that illustrates according to the 2D image partition method of the embodiment of the invention.
With reference to Fig. 2, comprise that according to the 2D image partition method of the embodiment of the invention operation S11 is to S17.Operation S11 confirms the coordinate of the pixel of the 2D image of input line by line, and confirms whether the input pixel is the valid pixel that is used to cut apart.Confirmed that by pixel unit 12 confirms as when cutting apart invalid pixel at operation S12 when the input pixel, operation S13 is in the holding state that receives follow-up input pixel.When importing pixel when operation S12 is confirmed that by pixel the valid pixel that is used to cut apart is confirmed as in unit 12; Operation S14 scans the sets of adjacent pixels that is arranged in sweep limits, and said sweep limits is preset in the valid pixel from the group information memory cell 11 of the information of storage pixel group.Whether operation S15 exists sets of adjacent pixels that scans and the pixel groups that comprises valid pixel input pixel before to update stored in the information of the pixel groups in the group information memory cell 11 to the S17 basis.
2D image partition method according to the embodiment of the invention also can comprise: operation S18, confirm with the ratio of width to height of the pixel groups of the renewal completion that is confirmed as group information whether corresponding pixel groups is effective according to the pixel quantity that comprises in the pixel groups (said pixel groups is the pixel groups that is confirmed as the renewal completion of group information in the pixel groups that is stored in the group information memory cell 11); And the operation (not shown) of group information that is confirmed as the pixel groups of inactive pixels group from the deletion of group information memory cell 11.
Below, will describe the 2D image segmentation apparatus and the method according to the embodiment of the invention of parts with reference to accompanying drawing in detail with Fig. 1 and Fig. 2.
Usually; Under the situation by the image of taking such as the imageing sensor (charge coupled device (CCD) or complementary metal oxide semiconductors (CMOS) (CMOS)) of digital camera, to be included in all pixels in the frame be not a time input but import in order according to particular order.Particularly; The pixel that has matrix structure and be arranged in image has input form line by line: the pixel that is included in row is transfused to the order of row; The pixel that is included in the next column is transfused to capable order, and wherein, the input of pixel begins from the first capable pixel that is positioned at first row.Can use embodiments of the invention by input sequence with each image pixel of the input of input form line by line.
The input picture of having used the embodiment of the invention can be through normal image having been used the specific binary picture of only representing to have desired character that covers.For example; Extracting the red pixel of expression from the general pattern shown in Fig. 3 A to remove under the situation of seeing red; Cover if used redness, then can obtain such binary picture, wherein; Only comprise that other pixel is shown as white (having value 1) greater than other red pixel of a specific order is shown as black (having value 1), shown in Fig. 3 B.
In aforesaid embodiments of the invention; When the binary picture of the pixel that has customizing messages with the scheme input expression of lining by line scan; At operation S11, pixel is confirmed the coordinate of unit 12 definite current input pixels, and confirms whether the input pixel is the pixel that is used to cut apart.For example, pixel confirms that unit 12 can confirm in binary picture that the input pixel has value 1 or has value 0, and the pixel of confirming to have value 1 is a valid pixel.The pixel of not confirmed to confirm as valid pixel in unit 12 by pixel (promptly; Pixel with value 0) shall not be applied to a series of processing, and confirm that in operation S13 pixel unit 12 is in the holding state that receives follow-up input pixel according to the embodiment of the invention as described below.
When operation S12 pixel is confirmed that by pixel valid pixel is confirmed as in unit 12, use the information that is stored in the pixel groups in the group information memory cell 11 in advance to scan the group adjacent with current valid pixel in 14 groups of scanning elements of operation 13.
Group information memory cell 11 is parts of storing the information of the pixel groups that comprises a plurality of neighbors, and group information memory cell 11 can be stored the information of each pixel groups by register form.Fig. 4 is the exemplary diagram of a divided pixel groups.Fig. 5 illustrates a plurality of registers of the group information of 32 groups of storages.As shown in Figure 4, the group information of a pixel groups can be represented as the horizontal min coordinates x_min that is included in the pixel groups, horizontal maximum coordinates x_max, vertical min coordinates y_min is with vertical maximum coordinates y_max and be included in the pixel quantity grp_pixel_count in the pixel groups.In addition, as shown in Figure 5, can store all group information by register form.
As shown in Figure 5, group information memory cell 11 can comprise: sign grp_Flag, whether expression group information can be upgraded by current input pixel.Sign grp_Flag representes whether corresponding pixel groups is in the state of activation that group information can be upgraded by current input pixel.For example, at the vertical maximum coordinates y_max of corresponding pixel groups not under the situation within the vertical scanning scope at the valid pixel of current input, because the information of corresponding pixel groups no longer is updated, so can in sign grp_Flag, represent illegal state.Can carry out and to confirm to handle in the validity for pixel groups of following description according to sign grp_Flag.
At operation S14, the sets of adjacent pixels of group scanning element 13 scanning valid pixels, group scanning element 13 can be in user's scanning neighbor group from the preset sweep limits of outside input.Can the processing of scanning neighbor group mainly be divided into two schemes according to the shape of the sweep limits that forms from current valid pixel.
Shown in Fig. 6 A; First scheme is such scheme: be applied to whole pixel groups (group 2, group 3) through the preset level sweep limits x_scan_range that will have identical size and come scanning neighbor pixel groups, at least a portion of said pixel groups to be included in from the preset vertical scanning scope y_scan_range of valid pixel P.Because will having the horizontal sweep scope x_scan_range of identical size, the first such scheme is applied to whole pixel groups; At least a portion of said pixel groups is included among the vertical scanning scope y_scan_range, therefore can implement this scheme very simply.As a result, shown in Fig. 6 B, first scheme is identical with the scheme that square window is expressed as sweep limits, and said window is represented as horizontal sweep scope and the vertical scanning scope in the current valid pixel.Yet; Because first scheme not Considering Vertical is used identical horizontal sweep scope apart from ground, the pixel groups that therefore on horizontal or vertical direction, is present in than is included on the diagonal of the farther square window of pixel groups in the square window possibly comprise sets of adjacent pixels.
Shown in Fig. 7 A or Fig. 7 B; Alternative plan is such scheme: the preset level sweep limits x_scan_range through the size that will change according to the vertical range between valid pixel P and each pixel groups is applied to each pixel groups and comes the scanning neighbor pixel groups, and at least a portion of said pixel groups is included in from the preset vertical scanning scope y_scan_range of valid pixel P.With reference to Fig. 7 A, 2 horizontal sweep scope x_scan_range are set to 5 pixels for the group in the pixel groups that is included among the vertical scanning scope y_scan_range, and wherein, the distance from valid pixel to group 2 is 3 pixels.3 horizontal sweep scope x_scan_range are set to 6 pixels for the group in the pixel groups that is included among the vertical scanning scope y_scan_range, and wherein, the distance from valid pixel to group 3 is 1 pixel.With reference to Fig. 8 A, 2 horizontal sweep scope x_scan_range are set to 3 pixels for the group in the pixel groups that is included among the vertical scanning scope y_scan_range, and wherein, the vertical range from valid pixel to group 2 is 3 pixels.3 horizontal sweep scope x_scan_range are set to 6 pixels for the group in the pixel groups that is included among the vertical scanning scope y_scan_range, and wherein, the vertical range from valid pixel to group 3 is 1 pixel.
Shown in Fig. 7 B, the sweeping scheme of Fig. 7 A is according to vertical range preset level sweep limits x_scan_range from the pixel groups to the valid pixel, make sweep limits become on the current valid pixel P convex window W1.Similarly, shown in Fig. 8 B, the sweeping scheme of Fig. 8 A makes sweep limits become the spill window W2 on the current valid pixel P according to the vertical range preset level sweep limits x_scan_range from the pixel groups to the valid pixel.
In order different horizontal sweep limits x_scan_range to be set according to the vertical range from the pixel groups to the valid pixel; The user can confirm to represent the index of the vertical range from the pixel groups to the valid pixel, and can be set to look-up table according to the size of the horizontal sweep scope x_scan_range of said index.
To S17, whether group information updating unit 14 is according to existing sets of adjacent pixels that scans and the pixel groups that comprises valid pixel input pixel before to update stored in the information of the pixel groups in the group information memory cell 11 at operation S15.
Especially; When there is not the group that is scanned by group scanning element 13 in the scanning result demonstration of operation S15; Group information updating unit 14 confirms that current valid pixels are first pixels that form new pixel groups, and the group information stores of new pixel groups that will comprise current valid pixel at operation S17 is in group information memory cell 11.
When showing, the scanning result of operating S15 has the group that scans by group scanning element 13; Then at operation S16, group information updating unit 14 can upgrade and proofread and correct with the group that comprises current valid pixel input pixel before and be stored in the group information of organizing in the information memory cell 11 in advance according to the quantity of the group that scans.
Fig. 9 is a kind of flow chart that is used for when scanning sets of adjacent pixels, upgrading the method that is stored in the group information of organizing information memory cell 11 in advance that illustrates according to the embodiment of the invention.With reference to Fig. 9, at operation S91, the quantity of the sets of adjacent pixels that group scanning element 13 scans is one.When the pixel groups that comprises the pixel before the current valid pixel and a pixel groups that scans at operation S91 not simultaneously; At operation S93; A sets of adjacent pixels that scans is merged in the pixel groups of the pixel that comprises previous input; With the deletion from group information memory cell 11 of the information of this sets of adjacent pixels that scans, and the information of upgrading the pixel groups that comprises previous input pixel is to be included in valid pixel in the said pixel groups that comprises previous input pixel.
In addition; When the quantity of the sets of adjacent pixels that scans in operation S91 group scanning element 13 is one; And when a sets of adjacent pixels scanning of operation S92 was identical with the pixel groups that comprises previous input pixel, the information of pixel groups that comprises previous input pixel in operation S94 renewal was to be included in valid pixel in the said pixel groups that comprises previous input pixel.
When the quantity of the sets of adjacent pixels that scans of group scanning element 13 is two; And when a pixel groups in two sets of adjacent pixels that operation S95 scans is identical with the pixel groups that comprises previous input pixel; At operation S96; With merging in the said pixel groups that comprises previous input pixel with the said pixel groups pixels with different group that comprises previous input pixel in two sets of adjacent pixels that scan; The information of the pixel groups that merges from the deletion of group information memory cell 11, and the information of upgrading the said pixel groups that comprises previous input pixel is to be included in valid pixel in the said pixel groups that comprises previous input pixel.
When the quantity of the sets of adjacent pixels that scans in operation S91 group scanning element 13 is two; And two sets of adjacent pixels scanning of operation S95 all with the pixel groups that comprises previous input pixel not simultaneously; At operation S97; Two sets of adjacent pixels that scan are merged to the said pixel groups that comprises previous input pixel; The group information of two pixel groups that merge from the deletion of group information memory cell 11, and the information of upgrading the said pixel groups that comprises previous input pixel is to be included in valid pixel in the said pixel groups that comprises previous input pixel.
At last; At operation S18; The quantity of the pixel that comprises in the pixel groups that stack features is confirmed to accomplish according to the renewal that is confirmed as group information in the pixel groups in being stored in group information memory cell 14 unit 15 confirms with the ratio of width to height of the pixel groups that the renewal that is confirmed as group information is accomplished whether corresponding pixel groups is effective, and deletes the group information of the pixel groups that is confirmed as the inactive pixels group from information memory cell 11.
Just; Stack features confirms that unit 15 inspections are stored in the sign grp_Flag (state of activation whether this sign expression exists the group information updating of pixel groups to accomplish) in the group information in the group information memory cell 11 with register form, and is included in the quantity of the pixel in the corresponding pixel groups on the pixel groups of forbidding and the ratio of width to height of corresponding pixel groups according to result's inspection of inspection.When the pixel quantity of the pixel groups of forbidding during less than the quantity of preset reference pixel, stack features confirms that unit 15 confirms that the pixel groups of forbidding is different with the pixel groups of the pattern of expression expectation, and deletes corresponding group of information from group information memory cell 11.In addition, stack features confirms that unit 15 uses horizontal minimum value, horizontal maximum, vertical minimum value and the vertical maximum of the pixel groups of forbidding to calculate the ratio of width to height of the pixel groups of forbidding.When the ratio that calculates was not in preset ratio ranges, stack features confirmed that unit 15 confirms that the pixel groups of forbidding is different with the pixel groups of the pattern of expression expectation, and deletes corresponding group of information from group information memory cell 11.For example; Shown in figure 10; In group information memory cell 11, stored a plurality of pixel groups of forbidding, and dividing method of the present invention has been applied to detect eye shape with under the situation that detects the system of seeing red, just; Stack features confirm unit 15 confirm with the pixel groups shown in Figure 10 in the ratio of width to height to have the group 5 of a lot of difference different with the pixel groups of the pattern with expectation with group 6, and delete the group information of corresponding pixel groups from group information memory cell 11.On the other hand, stack features confirms that the group 1 to 4 of unit 15 definite Figure 10 has suitable ranks ratio, and corresponding pixel groups is exported to be used for follow-up mode detection.
As stated, embodiments of the invention are confirmed to cut apart by importing pixel line by line, and through simply pixel and cutting apart of preset value comparison carries out image being need not complicated operations, thereby promptly the executed in real time image cuts apart.In addition, because embodiments of the invention need not be used for the microprocessor of complex operations, therefore can save time and cost.
Below, with describing a kind of equipment and method that removes the blood-shot eye illness in the 2D image by the area information of the definite pixel groups of above-mentioned splitting equipment and method of using in detail.
Figure 11 is the block diagram of equipment of blood-shot eye illness that is used for removing the 2D image according to the embodiment of the invention.
With reference to Figure 11, comprise reference length computing unit 110, length in pixels computing unit 120, intervally confirm unit 130 and image manipulation unit 140 according to the equipment 100 of blood-shot eye illness that is used for removing the 2D image of the embodiment of the invention.
Reference length computing unit 110 receives the coordinate information in the zone that comprises the blood-shot eye illness in the entire image, and confirms this regional centre coordinate and to the reference length of centre coordinate.
Length in pixels computing unit 120 orders receive each pixel of image, and calculate as the length in pixels of pixel with the distance at the center in zone.
Interval definite unit 130 compares reference length and length in pixels, and selects the pixel of length in pixels less than reference length.The interval interval of confirming unit 130 according to the pixel of in predetermined interval, confirming to comprise selection to the distance of centre coordinate.
Figure 12 is the block diagram of method of blood-shot eye illness that is used for removing the 2D image according to the embodiment of the invention.
With reference to Figure 12, comprise operation S100, S110, S120, S130 and S140 according to the method that is used to remove blood-shot eye illness of the embodiment of the invention.Operation S100 receives the coordinate information in the zone that comprises the blood-shot eye illness in the entire image.The centre coordinate in the zone that operation S110 confirm to receive and to the reference length of centre coordinate.Operation S120 receives each pixel of image in proper order, and calculates the length in pixels as the distance between each pixel and this regional center.Operation S130 compares reference length and length in pixels, and selects the pixel of length in pixels less than reference length.Operation S130 is according to the interval of the pixel of in predetermined interval, confirming to comprise selection to the distance of centre coordinate.Operation S140 is provided with the gain of each interval luminance gain and color, and uses luminance gain and the pixel value of color gain with the pixel of correction selection.
Figure 13 is the flow chart of operation S140 that illustrates in greater detail the correction pixel value of Figure 12.
With reference to Figure 13, the operation S140 that revises pixel value can comprise that operation S141 is to S143.Operation S141 is provided with each interval luminance gain distinctively according to the brightness value of the pixel of selecting.Operation S142 is provided with each interval color distinctively according to color of pixel value of selecting and difference between the color of object value to be revised and gains.Operation S143 is modified to Y, Cb and the Cr value of the pixel of selecting like following equality 1 and arrives shown in the equality 3:
Y’＝Ly×Y (1)
Cb’＝Cb+(Sb-Cb)×Jb (2)
Cr’＝Cr+(Sr-Cr)×Jr (3)
At equality 1 in equality 3, the Y value of the pixel that Y representes to select, the Cb value of the pixel that Cb representes to select; The Cr value of the pixel that Cr representes to select, Ly representes each interval Y gain, Sb representes target Cb value to be revised; Jb representes each interval Cb gain, and Sr representes target Cr value to be revised, and Jr representes each interval Cr gain; The Y value of the correction of the pixel that Y ' expression is selected, the Cb value of the correction of the pixel that Cb ' expression is selected, the Cr value of the correction of the pixel that Cr ' expression is selected.
Below, will describe the equipment and the method for blood-shot eye illness that is used for removing the 2D image in detail with reference to accompanying drawing according to the embodiment of the invention.
With reference to Figure 11 and Figure 12, begin from the operation S100 that area information is input to reference length computing unit 110 according to the method that is used to remove blood-shot eye illness of the embodiment of the invention.Area information is the zone that comprises the pixel of cutting apart, and the said pixel of cutting apart is used for cutting apart by expection in the entire image of waiting to remove blood-shot eye illness and is used to represent the pixel of seeing red.This zone is represented as minimum level coordinate, maximum horizontal coordinate, minimum vertical coordinate and the maximum perpendicular coordinate of the pixel of cutting apart.Therefore, the area information that is input to reference length computing unit 100 is minimum level coordinate, maximum horizontal coordinate, minimum vertical coordinate and the maximum perpendicular coordinate in zone that is used to cut apart the pixel of expression blood-shot eye illness.2D image segmentation apparatus through according to the embodiment of the invention has been described the technology relevant with pixel segmentation in detail with method.In brief; According to the embodiment of the invention be used for remove zone that equipment and the method for blood-shot eye illness of 2D image occur blood-shot eye illness through the image partition method of using according to the embodiment of the invention and carry out and cut apart; And the coordinate information of the pixel groups of confirming is input to reference length computing unit 110, thereby begin to cut apart.
Figure 14 is the concept map in said zone.Shown in figure 14, this zone is represented as the zone between minimum level coordinate x_min, maximum horizontal coordinate x_max, minimum vertical coordinate y_min and the maximum perpendicular coordinate y_max of cutting apart pixel.
At operation S110, reference length computing unit 110 uses the information of input area to confirm to be applied to the reference length len_max of this input area.In order to confirm reference length, reference length computing unit 110 calculates centre coordinate X_0 and the Y_0 of this regional minimum level coordinate x_min, maximum horizontal coordinate x_max, minimum vertical coordinate y_min and maximum perpendicular coordinate y_max.Here; The horizontal coordinate that the middle point coordinates X_0 of minimum level coordinate x_min that this is regional and maximum horizontal coordinate x_max is centre coordinate X_0 and Y_0, the vertical coordinate that the middle point coordinates Y_0 of minimum vertical coordinate y_min that this is regional and maximum perpendicular coordinate y_max is centre coordinate X_0 and Y_0.Next, the mean value ((a+b)/2) of 1/2nd (b=(y_max-y_min)/2) of distance is confirmed as reference length len_max between the minimum level coordinate that reference length computing unit 110 should the zone and 1/2nd (a=(x_max-x_min)/2) of the distance between the maximum horizontal coordinate and this regional minimum vertical coordinate and the maximum perpendicular coordinate.
At operation S120, length in pixels computing unit 120 orders receive each pixel of the image that is provided with the zone, and calculate the length in pixels as the distance between the centre coordinate in pixel and zone.With reference to Figure 14, false coordinate x_pos and y_pos are the coordinates of input pixel, and length in pixels len_xy is " (x_con)
2+ (y_con)
2" square root.Here, x_con be the input pixel with the centre coordinate in zone between horizontal range, y_con is the vertical range between input pixel and the regional centre coordinate.
Interval definite unit 130 compares reference length len_max and length in pixels len_xy, and selects the pixel of length in pixels len_xy less than reference length len_max.The interval interval of confirming unit 130 according to the pixel of in predetermined interval, confirming to comprise selection to the distance of centre coordinate X_0 and Y_0.
Just, shown in figure 14, interval confirm that unit 130 selects to be included in reference to the pixel in the circle, wherein, this center with reference to circle is coordinate X_0 and Y_0, is reference length len_max with reference to the radius of circle.
The interval change rate of confirming can use Figure 15 in unit 130 is to change color or brightness according to the relation between reference length len_max and the length in pixels len_xy (that is, being included in reference to the locations of pixels in the circle) smoothly.The change rate be with will be in the luminance gain of following description and color gain corresponding concept, can be understood that through natural color change being used and color is revised in the gain that changes gradually according to locations of pixels.
Shown in figure 15, preferably use the change rate that continuously changes according to the input locations of pixels, but the operation that on hardware or software, realizes the change rate possibly be very complicated.Therefore, shown in the dotted line of Figure 16 in Figure 19, embodiments of the invention can be selected such method: a plurality of intervals are set and intervally use identical gain to one according to the distance to centre coordinate.
Interval definite unit 130 is provided with a plurality of intervals according to the distance to centre coordinate, and confirms the input pixel is in which interval.
At operation S140, image manipulation unit 140 is provided with the gain of each interval luminance gain and color, and uses luminance gain and the pixel value of color gain with the pixel of correction selection.Particularly, at operation S141 and operation S142, gain is cut apart according to the brightness and the color application of input pixel in image manipulation unit 140.
Figure 16 and Figure 17 are the curve charts that illustrates in embodiments of the present invention according to the change of the luminance gain of the locations of pixels of selecting.In the curve chart shown in Figure 16 and Figure 17, the x axle representes that to the distance of centre coordinate (centre coordinate is set to 1), the y axle is represented luminance gain.For example, has under the situation of high luminance values the pixel that high-gain is applied to be positioned at outermost regions shown in figure 16 in the input pixel.On the other hand, in the input pixel when having low relatively brightness value, gain application that will be low relatively shown in figure 17 is in the pixel that is positioned at outermost regions, thereby can reduce and difference near the pixel of centre coordinate.
Figure 18 and Figure 19 are the curve charts that illustrates in embodiments of the present invention according to the change of the color gain of the locations of pixels of selecting.In the curve chart shown in Figure 18 and Figure 19, the x axle representes that to the distance of centre coordinate (centre coordinate is set to 1) the y axle is represented the color gain.Opposite with above-mentioned luminance gain, if pixel near centre coordinate, then color gain is set to have high-gain.Yet, similar with above-mentioned luminance gain, when the aberration between color of object to be changed and current input pixel is huge, the pixel that high-gain is applied to be positioned at outermost regions shown in figure 18.On the other hand, the aberration between color of object and current input pixel hour, color gain application that will be low relatively shown in figure 19 is in the pixel that is positioned at outermost regions, thereby can reduce the difference with the central area.
The brightness that can suitably select to be used for a luminance gain selecting to be employed from the luminance gain of Figure 16 and Figure 17 by the user with reference to and be used for the color reference that gains from the color that the color gain of Figure 18 and Figure 19 is selected to be employed.
In addition, under input color of pixel situation identical with skin color or very bright color (bright such as adding), image manipulation unit 140 gets around corresponding pixel and does not revise the shell degree and the color of this corresponding pixel.
Reflective becoming occurring at the image of expression eyes because in the central area of eyes does not have under the bright situation of adding of color, and image is being represented more perfect natural eyes aspect the meticulous visual angle.Image manipulation unit 140 can confirm luminance gain with through use such advantage make the interval that comprises centre coordinate in the interval comprise pixel nearly all be add bright.Just, shown in the dotted line of Figure 20 and Figure 21, image manipulation unit 140 can be through approach 1 gain and represent very brightly eye image using between the center that comprises centre coordinate.Under the situation of using such gain, shown in figure 22, can obtain the natural image of eyes, wherein, cornea deepening and glisten towards the center in the center.
For example, under the situation that the gain between the given zone of confirming according to above-mentioned processing is determined in the YCbCr color space, the brightness value Y of pixel, color value Cb and color value Cr are confirmed as shown in following equality 1 to 3:
Y’＝Ly×Y (1)
Cb’＝Cb+(Sb-Cb)×Jb (2)
Cr’＝Cr+(Sr-Cr)×Jr (3)
At equality 1 in equality 3, the Y value of the pixel that Y representes to select, the Cb value of the pixel that Cb representes to select; The Cr value of the pixel that Cr representes to select, Ly representes each interval Y gain, Sb representes target Cb value to be revised; Jb representes each interval Cb gain, and Sr representes target Cr value to be revised, and Jr representes each interval Cr gain; The Y value of the correction of the pixel that Y ' expression is selected, the Cb value of the correction of the pixel that Cb ' expression is selected, the Cr value of the correction of the pixel that Cr ' expression is selected.
As stated, embodiments of the invention according to eye center apart between cut section and use the gain that changes gradually according to the interval, thereby can remove the eye image that obtain nature after seeing red.In addition, because embodiments of the invention can be removed blood-shot eye illness in real time according to the pixel order of input picture, so they can be carried out fast processing and not need independent a plurality of storage space.
Embodiments of the invention can be cut apart the pixel executive real-time of input picture line by line, thereby have obviously reduced in the time of cutting apart the cost of 2D image.
Owing to do not need the complicated operation that is used to cut apart the 2D image, so embodiments of the invention can omit the other hardware resource such as microprocessor, thereby have reduced the cost that realization spent of hardware.
Because the image of removing after the blood-shot eye illness is very natural, so embodiments of the invention do not have the vestige of image manipulation, and can be used to see red storage space and the operational ton of removing operation thereby reduced according to the pixel cell executive real-time operation of input picture.
Though combined exemplary embodiment to illustrate and described the present invention, for those skilled in the art clearly, under the situation of the spirit and scope of the present invention that do not break away from the claim qualification, can make and revising and variation.
Claims (20)
1. 2D image segmentation apparatus that is used to cut apart the pixel of importing the 2D image line by line comprises:
The group information memory cell, storage comprises the information of the pixel groups of a plurality of neighbors;
Pixel is confirmed the unit, confirms the coordinate of input pixel, and confirms whether the input pixel is the valid pixel that is used to cut apart;
The group scanning element when the input pixel is confirmed that by pixel valid pixel is confirmed as in the unit, scans being arranged in from the sets of adjacent pixels of the preset sweep limits of valid pixel in the group information memory cell;
Whether the group information updating unit is according to existing sets of adjacent pixels that scans and the pixel groups that comprises valid pixel input pixel before to update stored in the information of the pixel groups in the group information memory cell.
2. equipment as claimed in claim 1 wherein, is input to pixel and confirms that the pixel of unit is the pixel of binary picture.
3. equipment as claimed in claim 1; Wherein, The group scanning element is applied to whole pixel groups through the preset level sweep limits that will have identical size and comes the scanning neighbor pixel groups, and at least a portion of said whole pixel groups is included in from the preset vertical scanning scope of valid pixel.
4. equipment as claimed in claim 1; Wherein, The group scanning element is applied to said each pixel groups through the preset level sweep limits that will have the size that changes according to the vertical range between valid pixel and each pixel groups and comes the scanning neighbor pixel groups, and at least a portion of said each pixel groups is included in from the preset vertical scanning scope of valid pixel.
5. equipment as claimed in claim 4, wherein, said preset level sweep limits with the size that changes according to the vertical range between valid pixel and each pixel groups is stored in the look-up table.
6. equipment as claimed in claim 1, wherein, when not having the sets of adjacent pixels that scans, the group information updating unit will comprise valid pixel new pixel groups information stores the group information memory cell in.
7. equipment as claimed in claim 1; Wherein, When the quantity of the sets of adjacent pixels that scans is one and a sets of adjacent pixels scanning when identical with the pixel groups that comprises previous input pixel; The group information updating unit upgrades the information of the pixel groups that comprises previous input pixel, thereby valid pixel is included in the pixel groups that comprises previous input pixel;
When the quantity of the sets of adjacent pixels that scans be one and sets of adjacent pixels that scans and the pixel groups that comprises previous input pixel not simultaneously; The group information updating unit merges to a said sets of adjacent pixels that scans in the pixel groups that comprises previous input pixel; And upgrade the information of the pixel groups comprise previous input pixel, thereby valid pixel is included in the pixel groups that comprises previous input pixel;
When the quantity of the sets of adjacent pixels that scans is two; And when a pixel groups in two sets of adjacent pixels that scan is identical with the pixel groups that comprises previous input pixel; In two sets of adjacent pixels that the group information updating unit will scan with comprise that the pixel groups pixels with different group of previous input pixel merges in the pixel groups that comprises previous input pixel; And upgrade the information of the pixel groups comprise previous input pixel, thereby valid pixel is included in the pixel groups that comprises previous input pixel;
When the quantity of the sets of adjacent pixels that scans is two; And two sets of adjacent pixels that scan and the pixel groups that comprises previous input pixel are not simultaneously; Two sets of adjacent pixels that the group information updating unit will scan merge in the pixel groups that comprises previous input pixel groups; And upgrade the information of the pixel groups comprise previous input pixel, thereby valid pixel is included in the pixel groups that comprises previous input pixel.
8. equipment as claimed in claim 1, wherein, the information that is stored in the pixel groups in the group information memory cell comprises:
Be included in the horizontal min coordinates of the pixel in the pixel groups, horizontal maximum coordinates, vertical min coordinates and vertical maximum coordinates;
Be included in the quantity of the pixel in the pixel groups; With
Whether expression group information can be by the sign of current input pixel renewal.
9. equipment as claimed in claim 8, wherein, the group information updating unit is represented in said sign: the renewal that is not in the group information of the pixel groups in the preset horizontal sweep scope of valid pixel is accomplished.
10. equipment as claimed in claim 1; Also comprise: stack features is confirmed the unit; The pixel quantity that comprises in the pixel groups according to the renewal completion that is confirmed as group information in the pixel groups that is stored in the group information memory cell confirms with the ratio of width to height of the pixel groups of the renewal completion that is confirmed as group information whether corresponding pixel groups is effective, and is confirmed as the group information of the pixel groups of inactive pixels group from the deletion of group information memory cell.
11. one kind is used to cut apart the 2D image partition method of importing the 2D image line by line, comprises:
Confirm the coordinate of input pixel, and confirm whether the input pixel is the valid pixel that is used to cut apart;
When input pixel when being confirmed as valid pixel, be arranged in from the sets of adjacent pixels of the preset sweep limits of valid pixel in the group information memory cell to the information of storage pixel group and scan;
According to whether existing sets of adjacent pixels that scans and the pixel groups that comprises valid pixel input pixel before to update stored in the information of the pixel groups in the group information memory cell.
12. method as claimed in claim 11, wherein, in the step of confirming coordinate, the input pixel is the pixel of binary picture.
13. method as claimed in claim 11; Wherein, The step that sets of adjacent pixels is scanned comprises: be applied to whole pixel groups through the preset level sweep limits that will have identical size and come the scanning neighbor pixel groups, at least a portion of said whole pixel groups is included in from the preset vertical scanning scope of valid pixel.
14. method as claimed in claim 11; Wherein, The step that sets of adjacent pixels is scanned comprises: be applied to said each pixel groups through the preset level sweep limits that will have the size that changes according to the vertical range between valid pixel and each pixel groups and come the scanning neighbor pixel groups, at least a portion of said each pixel groups is included in from the preset vertical scanning scope of valid pixel.
15. method as claimed in claim 14, wherein, said preset level sweep limits with the size that changes according to the vertical range between valid pixel and each pixel groups is stored in the look-up table.
16. method as claimed in claim 11, wherein, the step of lastest imformation comprises: when not having the sets of adjacent pixels that scans, will comprise valid pixel new pixel groups information stores the group information memory cell in.
17. method as claimed in claim 11, wherein, the step of lastest imformation comprises:
When the quantity of the sets of adjacent pixels that scans is one and a sets of adjacent pixels scanning when identical with the pixel groups that comprises previous input pixel; Renewal comprises the information of the pixel groups of previous input pixel, thereby valid pixel is included in the pixel groups that comprises previous input pixel;
When the quantity of the sets of adjacent pixels that scans be one and sets of adjacent pixels that scans and the pixel groups that comprises previous input pixel not simultaneously; A said sets of adjacent pixels that scans is merged in the pixel groups that comprises previous input pixel; And upgrade the information of the pixel groups comprise previous input pixel, thereby valid pixel is included in the pixel groups that comprises previous input pixel;
When the quantity of the sets of adjacent pixels that scans is two; And when a pixel groups in two sets of adjacent pixels that scan is identical with the pixel groups that comprises previous input pixel; With merging in the pixel groups that comprises previous input pixel with the pixel groups pixels with different group that comprises previous input pixel in two sets of adjacent pixels that scan and upgrade the information of the pixel groups that comprises previous input pixel, thereby valid pixel is included in the pixel groups that comprises previous input pixel;
When the quantity of the sets of adjacent pixels that scans is two; And two sets of adjacent pixels that scan and the pixel groups that comprises previous input pixel are not simultaneously; Two sets of adjacent pixels that scan are merged in the pixel groups that comprises previous input pixel groups; And upgrade the information of the pixel groups comprise previous input pixel, thereby valid pixel is included in the pixel groups that comprises previous input pixel.
18. method as claimed in claim 11, wherein, the information that is stored in the pixel groups in the group information memory cell comprises:
Be included in the horizontal min coordinates of the pixel in the pixel groups, horizontal maximum coordinates, vertical min coordinates and vertical maximum coordinates;
Be included in the quantity of the pixel in the pixel groups; With
Whether expression group information can be by the sign of current input pixel renewal.
19. method as claimed in claim 18, wherein, the step of lastest imformation also comprises: the renewal that expression is not in the group information of the pixel groups in the preset horizontal sweep scope of valid pixel in said sign is accomplished.
20. method as claimed in claim 11 also comprises:
The pixel quantity that comprises in the pixel groups according to the renewal completion that is confirmed as group information in the pixel groups that is stored in the group information memory cell confirms with the ratio of width to height of the pixel groups of the renewal completion that is confirmed as group information whether corresponding pixel groups is effective;
Be confirmed as the group information of the pixel groups of inactive pixels group from the deletion of group information memory cell.
Applications Claiming Priority (6)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| KR1020080021236A KR100961136B1 (en) | 2008-03-07 | 2008-03-07 | Segmentation apparatus and method for an object classification of 2d image |
| KR1020080021236 | 2008-03-07 | ||
| KR10-2008-0021236 | 2008-03-07 | ||
| KR1020080044444 | 2008-05-14 | ||
| KR1020080044444A KR100973012B1 (en) | 2008-05-14 | 2008-05-14 | Apparatus and method of removing red-eye in digital image |
| KR10-2008-0044444 | 2008-05-14 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101527780A CN101527780A (en) | 2009-09-09 |
| CN101527780B true CN101527780B (en) | 2012-11-07 |
Family
ID=41095456
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2008101815752A Expired - Fee Related CN101527780B (en) | 2008-03-07 | 2008-11-27 | 2D image segmentation apparatus and method, and apparatus and method for removing red eye from image |
Country Status (2)
| Country | Link |
|---|---|
| KR (1) | KR100961136B1 (en) |
| CN (1) | CN101527780B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1182333A (en) * | 1996-06-28 | 1998-05-20 | 松下电器产业株式会社 | Image coding apparatus, image decoding apparatus, image coding method, image decoding method, image coding program recording media, image decoding program recording media |
| CN1197251A (en) * | 1996-10-24 | 1998-10-28 | 松下电器产业株式会社 | Filling method of image signal, its coding device and decoding device |
| WO2004088587A1 (en) * | 2003-03-28 | 2004-10-14 | National Institute Of Information And Communications Technology, Independent Administrative Agency | Image processing method and image processing device |
| WO2007027847A2 (en) * | 2005-09-01 | 2007-03-08 | Geosim Systems Ltd. | System and method for cost-effective, high-fidelity 3d-modeling of large-scale urban environments |
-
2008
- 2008-03-07 KR KR1020080021236A patent/KR100961136B1/en not_active IP Right Cessation
- 2008-11-27 CN CN2008101815752A patent/CN101527780B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1182333A (en) * | 1996-06-28 | 1998-05-20 | 松下电器产业株式会社 | Image coding apparatus, image decoding apparatus, image coding method, image decoding method, image coding program recording media, image decoding program recording media |
| CN1197251A (en) * | 1996-10-24 | 1998-10-28 | 松下电器产业株式会社 | Filling method of image signal, its coding device and decoding device |
| WO2004088587A1 (en) * | 2003-03-28 | 2004-10-14 | National Institute Of Information And Communications Technology, Independent Administrative Agency | Image processing method and image processing device |
| WO2007027847A2 (en) * | 2005-09-01 | 2007-03-08 | Geosim Systems Ltd. | System and method for cost-effective, high-fidelity 3d-modeling of large-scale urban environments |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101527780A (en) | 2009-09-09 |
| KR20090095967A (en) | 2009-09-10 |
| KR100961136B1 (en) | 2010-06-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9239946B2 (en) | Method and apparatus for detecting and processing specific pattern from image | |
| CN100469099C (en) | System and method for connecting regions of image data having similar characteristics | |
| CN100517461C (en) | Image processing device and method | |
| JP5455038B2 (en) | Image processing apparatus, image processing method, and program | |
| US10275677B2 (en) | Image processing apparatus, image processing method and program | |
| US20100232680A1 (en) | Pattern detection on an simd processor | |
| CN102959582B (en) | Image processing apparatus | |
| US20110075950A1 (en) | Image retrieval device and computer program for image retrieval applicable to the image retrieval device | |
| CN108596908B (en) | LED display screen detection method and device and terminal | |
| CN103503031A (en) | Low resolution buffer based pixel culling | |
| US20140294301A1 (en) | Image processing apparatus and image processing method | |
| US20110229026A1 (en) | Image processing apparatus, image processing method, and storage medium of image processing method | |
| JP5864936B2 (en) | Image processing apparatus, image processing method, and program | |
| JP2007200246A (en) | Method for evaluating image processing algorithm, method, device, and program for generating same algorithm, and recording medium | |
| CN101527780B (en) | 2D image segmentation apparatus and method, and apparatus and method for removing red eye from image | |
| US20230394871A1 (en) | Method for verifying the identity of a user by identifying an object within an image that has a biometric characteristic of the user and separating a portion of the image comprising the biometric characteristic from other portions of the image | |
| CN115187593B (en) | Screen defect detection method and device | |
| US8515169B2 (en) | Apparatus and method for removing red-eye in a two-dimensional (2D) image | |
| US10424051B2 (en) | Image analyzing apparatus and non-transitory storage medium storing instructions executable by the image analyzing apparatus | |
| JP2019012424A (en) | Image processing device and computer program | |
| CN109815975A (en) | A kind of objective classification method and relevant apparatus based on robot | |
| JP7385192B2 (en) | Image processing device, image processing method, and computer program | |
| JP2010224612A (en) | Image processing device, image reading device, image forming device and program | |
| JP5653308B2 (en) | Region detection apparatus, region detection method, and program | |
| KR20020020941A (en) | Method, System and Apparatus |
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: 20121107 Termination date: 20151127 |