JP2009044405A - Image processing device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve confirmation work of a correction process by generating a list about correction points, in an image processing device which detects a defective region in digital image data for correcting. <P>SOLUTION: A list about correction points is generated based on defective regions set automatically or manually, and conformation work is done for a correction process using the list. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention performs correction processing, creates a list of correction points when an enlarged display of the correction result and the original image and confirms the correction result, and uses the list to efficiently check the correction result. The present invention relates to a processing apparatus.

  Today, various digital image data such as digital image data taken with a digital camera and digital image data taken with a film camera film taken with a scanner or the like are widely used.

  As an advantage of the digital image, when there is a defective area in the image data, it can be considered that a correction process can be performed to make the defective area inconspicuous by applying processing to the area. Correction processing is performed in a wide range of fields.

  The correction process is very time-consuming if the user finds a correction area and manually performs the correction process for each defective area. In view of this, application software has been developed that can automatically detect a defective area and easily and automatically correct the detected area.

  It is necessary to confirm whether or not the correction processing is automatically performed with respect to the correction processing automatically executed by application software or the like.

  When enlarging and confirming a correction result with respect to a correction portion that has been subjected to processing such as red-eye correction, it takes time if the correction region is enlarged while searching for the correction portion while scrolling and checking the color correction or the like. In view of this, Japanese Patent Laid-Open No. 2003-209683 has devised a method for improving work efficiency by enlarging the detected correction point as a center. In addition, if there are multiple correction locations, the display is enlarged with each correction location as the center, and the display order depends on at least one element such as the position, size, orientation, and image feature amount of the correction location. It is a method characterized by displaying.

  Japanese Patent Application Laid-Open No. 2003-179750 discloses a method of displaying an image after correcting a defective area automatically and comparing an image before correcting the defective area with an image after correcting the defective area. Is devised.

However, when a large number of subjects are included as in a group photo, the number of defective areas detected increases, and it is considered that the work cannot be performed efficiently if each correction point is simply confirmed by an enlarged comparison or the like. Further, since various correction processing methods are executed, it is considered that it is often difficult to verify correction results specialized for each process.
Japanese Patent Laid-Open No. 2003-209683 JP 2003-179750 A

  Therefore, in the present invention, a method for creating a list of detected correction points and efficiently checking the correction result of each correction point is devised.

  The present invention comprises the following means.

Means for automatically detecting a defective portion that needs correction from image data;
Means for adding correction points from the image data by the user;
Means for determining means for correcting the defective portion with respect to the detected illegal area;
Means for creating a list from correction points automatically added from image data or manually by a user;
Means for rearranging the order of the correction parts displayed in the correction part list in various orders;
Means capable of enlarging and comparing the image data before and after the correction processing of the correction portion designated using the correction portion list;
Means for determining the execution of the correction process after enlarging the image data before and after the correction process of the correction part selected from the correction part list;
Means for changing the correction process after enlarging the image data before and after the correction process of the correction part selected from the correction part list;
An image processing apparatus having means for performing a determined correction.

  The correction location list can enumerate not only correction locations of one image but also a plurality of correction locations included in a plurality of images.

  In the correction location list, correction locations can be rearranged for each of various items, and work efficiency can be improved by improving operability.

  The correction location list is arranged in such a way that the coordinate position is close to or far from the center, the focus is best or least focused, the correction area is large or the correction area is small, and the correction amount by correction is large. Alternatively, various orders such as a list arranged by correction method or a list arranged by parts corrected based on the result of face detection can be considered.

  When correction is performed on a defective part detected by the present invention and a part specified by the user, a list relating to the correction part is created so that the work can be efficiently performed when the correction result is confirmed. Is possible.

  Next, details of the present invention will be described in accordance with the description of the embodiments.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

(First embodiment)
An embodiment of the present invention will be described using an image processing apparatus as shown in FIG. FIG. 1 is an example of an image processing apparatus for realizing the present invention, and is a block diagram of a personal computer.

  In the figure, 100 is a CPU that performs processing operations, 101 is a ROM that stores processing programs, 102 is a RAM that temporarily stores processing results, 103 is a hard disk that stores image data and a program for correction, 104 is a keyboard for the user to input instructions, 105 is an input assisting means, such as a pointing device such as a mouse, 106 is a display for displaying processed images, and 107 is an I / O interface.

  The input interface of the I / O interface is to read image data taken with a digital camera, scan from a film taken with a silver salt camera and read it as image data, or image data such as photographic images on a website There are various input methods, such as for obtaining from a communication line.

  There are various output methods for the output interface of the I / O interface, such as when the correction result is printed out by a printer, or when the correction result is copied to an external memory.

  FIG. 2 is a block diagram showing a procedure of an embodiment for creating a correction location list relating to correction locations and performing correction efficiently based on the list. First, an outline of the operation will be described with reference to FIG. Image processing is realized by processing by the CPU according to a program stored in the ROM.

  In FIG. 2, S101 is a portion for reading image data, and inputs image data expressed in RGB. Since the input RGB image data usually has 256 gradations, each of R, G, and B is normalized from 0 to 1 here.

  In step S102, the normalized RGB image data is converted into HIS image data that displays hue, saturation, and intensity. This is because it is easier to set the detection criterion for the feature value by handling the image data converted to HSI than handling the image data in RGB. Conversion from RGB image data to HIS image data is performed using the following relational expression.

H = tan ⁻¹ {GB} / (2R−GB)} (1)
S = {(BR) ² + (RG) ² + (GB) ² } / 3 (2)
I = (R + G + B) (3)
However, normalize to 0 <= H <= 360 °, 0 <= S <= (2/3), 0 <= I <= 3.

  Next, as shown in S103, the following detection criteria are determined from the converted HIS image data, and the face area is specified. The face area is specified by using the following detection criteria and foresight information such as the size of a general face area.

10 ° <H <30 ° (4)
0.2 <S <0.35 (5)
1.8 <I (6)
Of course, it is obvious that any detection method may be used as the detection method of the face area, and it can be applied to the correction process of the face area.

  Here, correction points are set on the setting screen as shown in FIG. 3, and a correction point list is created. As a procedure for creating a correction location list, first, various defective areas to be corrected are automatically detected from the feature data from the converted image data in the face area, and sequentially added to the correction location list. Next, the user can arbitrarily specify a correction location that the user wants to execute correction and add it to the correction location list.

  Referring briefly to FIG. 3, reference numeral 301 denotes an automatic detection button. When this button is pressed, an operation until a correction location is automatically detected and added to the correction location list is performed. Reference numeral 302 denotes an enlargement button, which can be enlarged so that the user can easily work when adding a correction area, and is enlarged and displayed as much as the button is pressed. Reference numeral 303 denotes a reduction button, which is used when the user adds a correction area, and is reduced and displayed as much as the button is pressed. A correction area addition button 304 is used to select a correction area to be added by the user and press the button to add a specified area to the correction location list. A correction area deletion button 305 is used to select a correction location added by automatic detection or a correction location added by the user from the correction location list, and can be deleted from the correction location list by pressing this button. A list determination button 306 is a button used when determining a correction location list created by automatic detection and manual selection, and means completion of the correction location list. Press this button to move to the next process. Reference numeral 307 denotes a selection setting point. When the user wants to add a correction item, the user can select a correction area using a pointing device such as a mouse. Reference numeral 308 denotes a selection area, which is a correction area selected using 307 points. Reference numeral 309 denotes a display screen, which is used when a correction area added to the list is to be indicated to the user by a red circle or the like. A correction location list 310 displays correction locations set by automatic detection or manual detection by a user as a list. Reference numeral 311 indicates each correction point.

  In step S104, the correction target portion is automatically detected. A correction method for automatically performing correction on a defective area in the detected face area is automatically determined. The correction items relate to red-eye area correction, skin softening, blur correction, catch light processing, and the like. The correction of the red-eye region is to correct a phenomenon in which the eyes shine red by photographing with a flash in a dark place to a natural color. The skin beautification process is a process that can brightly correct the face area when the face area is dark, and at the same time smooth the skin of the face portion to convert it into a more natural image. Further, the blur correction is a process for making a face region more prominent by enhancing an edge when a face outline portion is blurred. Finally, the catchlight process is to perform a process of adding a highlight portion in the pupil area because highlights in the pupil area are important in an image including a person.

  Obviously, other correction methods may be used as the correction method.

  A correction method is automatically determined from the feature amount of the defective area. For example, when a red-eye area is detected as a defective area, the following red-eye detection reference is set, and the red-eye area is detected for the detected face area.

0 <H <30 ° (7)
S> 0.6 (8)
0.9 <I <1.2 (9)
It is obvious that the red-eye region may be detected by a standard other than the above or a method other than the above as a detection method of the red-eye region.

  In the same manner, appropriate detection criteria are set for the skin beautification processing target portion, the blur correction target portion, and the catch light processing target portion, respectively, and the target portion is detected. Each time these correction processing target portions are detected, they are added to the correction portion list in FIG. When all the correction methods and face areas have been detected, the automatic detection process ends.

  In addition, correction points added to the list are highlighted on the 309 display screen in FIG.

  Next, in S105, correction points are added manually. The user wishes to arbitrarily execute correction, and the user designates and adds to the correction position list. As an addition method, the point 307 in FIG. 3 is moved by a pointing device, the region added as a correction portion is dragged, and the range is set as 308. When the range is set, press the 304 correction area addition button. Then, the selected correction location can be added to the correction location list.

  Next, in step S106, correction is performed for correction points detected as defective areas automatically in the areas added to the list. By performing correction, the correction data is not overwritten on the original image data. The original image data is also stored, and if it is not corrected later, the image data of the correction processing portion is set in a state where it can be restored. At this stage, the correction result of the correction portion is not displayed yet. By performing the correction, it is possible to sequentially arrange the list according to the correction method, and the work efficiency can be further improved.

  In the correction process, a plurality of correction processes may be performed at one place. For example, red eye correction and highlight processing. Here, for simplification of description, the correction process will be described as one correction per one place.

  Next, in S107, a correction location list is created, and an enlarged display of the correction location and determination of correction processing are performed by using the interface as shown in FIG. Briefly describing FIG. 4, 401 is an enlarged view of image data before correction, and 402 is an enlarged view after correction. The enlarged data after the correction of 402 can be displayed as long as there is data that is automatically detected and corrected, but the correction point specified by the user is not corrected at this stage. ,Do not show. Reference numeral 403 denotes a correction location list created in S107. If the display order of the list is set in a part of the 404 “display order”, the list can be rearranged in various orders. For example, it can be rearranged depending on the correction amount or the correction method. Reference numeral 405 denotes a screen displaying the entire image data, and it can be grasped which part is currently being corrected. 406 is marked at the current correction location so that it can be easily understood by the user. This mark is added when the correction location list is selected. Reference numeral 407 denotes an enlargement / reduction button which is changed by the amount of enlargement / reduction of the screen before correction and the screen after correction being displayed, and is always displayed at the same magnification. Reference numeral 408 denotes a step button. When there are many correction points, the correction point can be moved by pressing this button when proceeding to the next correction point or returning to the next correction point. If the user decides to execute correction as a result of checking the corrected data, the user presses the “correction determination” button 409, and presses the cancel button when correction processing is not performed. 410 is used to set a correction method, and is used when changing the correction process or when setting a correction process to be performed on a correction portion added by the user.

  In S108, from the correction location list of FIG. 4, the display order of correction locations to be enlarged and displayed and the correction location to be enlarged are selected. When selecting a list, it is possible to select all the correction portions and check all the correction results, and it is possible to enlarge and display only the portions selected from the list. Alternatively, it is possible to perform enlargement comparison processing for each correction method, and work efficiency is improved by using a list.

  In S109, the parts selected from the list are enlarged and displayed in order.

  In S110, the user determines whether or not to execute correction from the correction result displayed on the enlarged comparison display. When executing correction, the correction execution button is pressed, and when correction is not performed, Cancel is pressed. When the correction method is changed or the correction location set by the user is selected.

  When changing the correction method, the process proceeds to S111, and the correction method is determined at 410 in FIG. When each correction method button is pressed, the process returns to S109, the reset correction method is executed, and the result is enlarged and displayed. Thereafter, the same procedure is followed.

  If it is determined that correction is to be executed, the process proceeds to S112, and if all correction points selected in the list have been corrected, the process ends. If correction of all the correction points selected in the list has not been completed, the process proceeds to S113, an enlargement comparison is performed on the next correction point from the list, and thereafter the same procedure is applied to all correction points selected in the list. Repeat until executed.

  When correction is performed only for the correction points specified in the created list, whether correction is to be executed is not determined for correction points not specified. Various measures such as setting whether or not to execute correction of correction points not designated by the user at the time of reaching S114 are made possible.

Image processor hardware configuration Image processing flowchart Correction point list creation example Example of comparison using a list

Claims (10)

Means for automatically detecting a defective portion that needs correction from image data;
Means for adding correction points from the image data by the user;
Means for determining means for correcting the defective portion with respect to the detected illegal area;
Means for creating a list from correction points automatically added from image data or manually by a user;
Means for rearranging the order of the correction parts displayed in the correction part list in various orders;
Means capable of enlarging and comparing the image data before and after the correction processing of the correction portion designated using the correction portion list;
Means for determining the execution of the correction process after enlarging the image data before and after the correction process of the correction part selected from the correction part list;
Means for changing the correction process after enlarging the image data before and after the correction process of the correction part selected from the correction part list;
An image processing apparatus having means for performing a determined correction.   2. The image processing apparatus according to claim 1, wherein the defective portion automatic detection means is defective area detection means for detecting a defective area from a feature amount of image data.   2. The image processing apparatus according to claim 1, wherein the correction location adding means by the user is a means for adding a correction location to the image data being displayed by using a pointing device input device.   2. The image processing apparatus according to claim 1, wherein the correction processing determination unit is a unit that determines correction based on a feature amount from image data of a correction portion determined to be a defective area.   2. The image processing apparatus according to claim 1, wherein the correction location list creation means is a means for creating a list that collectively manages correction locations detected automatically or correction locations designated by a user.   The list rearranging means can change the display order of the correction locations listed according to the correction-specific item or the correction-specific item based on the result of face detection according to the correction method having a large correction region or a small correction region. Item 12. The image processing device according to Item 1.   2. The image processing apparatus according to claim 1, wherein the correction location list displays all correction locations or correction locations designated by the user in an enlarged manner.   The correction part enlargement / comparison means managed in the correction part list is displayed so as to be easily compared with the user by enlarging mainly the image data of the correction result and the image data before correction. The image processing apparatus according to 1.   2. The image processing apparatus according to claim 1, wherein the correction processing determination unit executes the correction processing when the user confirms the correction portion displayed in the enlarged comparison display and determines that appropriate correction has been performed.   The correction process changing means can process the correction part being displayed again by another correction method when it is determined that the correction part displayed in the enlarged comparison display is not an appropriate correction process when the user confirms the correction part. The image processing apparatus according to claim 1.

Images