JP2009288942A - Image processing system, image processing method, image processing program, and recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processing system, an image processing method, an image processing program and a recording medium for selecting proper correction information, and for correcting an image on the basis of information based on the manual correction of a user to an original image. <P>SOLUTION: A manual correction input means 100 inputs a first correction value. An image processing means 101 performs image processing, and transmits an image processing result to an image display means 102. An image featured value calculation means 103 calculates the featured value of an original image by a correction item. A correction similarity decision means 104 decides the similarity of correction from first and second correction values and the second featured value for each correction time, and specifies the second correction value b close to the manual correction of a user from a post-stage second correction information DB 105. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an image processing system, an image processing method, an image processing program, and a recording medium that perform image correction based on information based on manual correction of an original image by a user.

  An image taken with a digital camera, a mobile phone or the like is usually displayed on its own display screen or a screen of a personal computer or the like, and the quality of the image is often judged. At that time, the intended image may not be taken due to various conditions such as daytime, nighttime, indoor / outdoor, and retrograde.

  In such a case, image correction software for photo retouching software is started up via a personal computer or the like, and image correction is performed. There are a variety of image correction items, but depending on the adjustment, the image quality can be greatly improved. Therefore, several devices shown below have been developed.

  Patent Document 1 is a patent for searching an image database for auction for an image similar to a desired image via a network.

  Patent Document 2 was developed to efficiently correct a large amount of images. In summary, the feature amount of the original image is calculated, compared with the feature amount of the image in the DB prepared in advance, and the most similar image is extracted. A single image correction is linked to the extracted image, and the linked image correction is applied. A desired image is obtained by repeating this process a plurality of times.

Patent Document 3 divides an image into items such as a face, a flower, and a blue sky, and counts and records what kind of processing each user performs. When an image to be corrected is input next time, processing with a high frequency of the counting result is performed.
JP 2003-303188 A JP 2004-54751 A JP 2006-323580 A

  However, Patent Document 1 searches for similar images, and there is no means for performing image correction based on the search results.

  Since Patent Document 2 is an application that focuses on automatically correcting a large amount of images, there is a problem that it is difficult to perform image correction reflecting individual preferences when used by a plurality of users.

  Patent Document 3 is an application in which an image is divided into items, each user's favorite processing count is counted for each item, priority is given, and image correction is performed. Therefore, one user needs to repeat the number of processes and form a favorite procedure. New users of this system will not get proper correction. In addition, the number of processing times is greatly affected by the image to be processed. In other words, there is a problem that the priority changes drastically when the opportunity to process an image deviating from normal increases.

  Accordingly, an object of the present invention is to determine an similarity with correction information registered in advance based on information based on a user's manual correction with respect to an original image, select appropriate correction information, and perform image correction A processing system, an image processing method, an image processing program, and a recording medium are provided.

  In order to achieve the above object, the image processing system according to claim 1, an image display unit that displays an image and an image correction item, a unit that inputs or calculates first correction information according to the correction item, A second correction information storage unit that stores or holds second correction information in advance for each correction item; a correction similarity determination unit that determines a similarity between the first correction information and the second correction information; Features.

  The image processing system according to claim 2 is the image processing system according to claim 1, wherein the correction similarity determination unit is configured to calculate a first similarity based on a first feature amount and a second feature amount included in the first correction information. The second similarity determination is performed from the first correction amount and the second correction amount included in the second correction information, and the two determination results of the first similarity determination and the second similarity determination are performed. The second correction amount is specified.

  The image processing system according to claim 3 is the image processing system according to claim 2, wherein the first similarity determination has a minimum difference between the first feature amount and the second feature amount; and In the second similarity determination, the difference between the first correction amount and the second correction amount is minimum.

  The image processing system according to claim 4 further includes user ID detection means for performing user recognition in the image processing system according to any one of claims 1 to 3, wherein the detected user ID, The second correction information specified by the correction similarity determination unit is recorded by the second correction information storage unit, and the second correction information recorded from the user ID, the correction item, and the first feature amount of the image is stored. It is characterized by specifying.

  The image processing method according to claim 5 includes an image display step of displaying an image and an image correction item, a step of inputting or calculating first correction information in accordance with the correction item, and a second for each correction item in advance. It has the 2nd correction information storage process which memorize | stores or hold | maintains correction information, 1st correction information, and the correction | amendment similarity determination process which determines the similarity of 2nd correction information, It is characterized by the above-mentioned.

  The image processing program according to claim 6, an image display process for displaying an image and an image correction item, a process for inputting or calculating first correction information in accordance with the correction item, and a second for each correction item in advance. A second correction information storage process for storing or holding correction information, a first correction information, and a correction similarity determination process for determining a similarity between the second correction information are caused to be executed by an image processing system. .

  A recording medium according to a seventh aspect stores the image processing program according to the sixth aspect.

  According to the present invention, based on information based on a user's manual correction with respect to an original image, an image processing system that determines similarity with previously registered correction information, selects appropriate correction information, and performs image correction. It is possible to provide an image processing method, an image processing program, and a recording medium.

The present invention will be described with reference to embodiments described below.
The embodiment described below is an embodiment that suitably or optimally implements the present invention, and is not limited thereto, and can be modified and modified within a range that can be easily conceived by those skilled in the art.

(First embodiment)
FIG. 1 is a block diagram showing an example of an image processing system having an image processing apparatus (1 in a broken line frame) and image display means 102 according to the present embodiment.

  The image processing apparatus (1 in a broken line frame) includes a manual correction input unit 100, an image processing unit 101, an image display unit 102, an image feature amount calculation unit 103, a correction similarity determination unit 104, and a second correction information DB (data list). Source) 105.

  The manual correction input unit 100 inputs the first correction amount manually corrected by the user according to the correction item for the original image to be corrected.

  The image processing unit 101 performs image processing based on the correction item and the first correction amount, sends the image processing result to the image display unit 102, and displays an image.

  The image display means 102 displays various information for supporting the input of the correction item and the first correction amount.

  The image feature amount calculating unit 103 is a unit that calculates the feature amount of the original image for each correction item.

  The correction similarity determination unit 104 determines the similarity of correction from the first and second correction amounts and the first and second feature amounts for each correction item, and the user's manual operation is performed from the second correction information DB 105 at the subsequent stage. A second correction amount b close to the correction is specified.

  The second correction amount b is fed back to the image processing means 101, the image processing is executed on the original image with the second correction amount b corresponding to the correction item, and the corrected image is displayed on the image display device 102.

The processing of the image processing system having the configuration of the image processing apparatus and the image display means described above is shown in FIGS.
FIG. 2 is a flowchart showing an example of processing performed in the image processing system of the present embodiment.
FIG. 3 is a diagram showing information handled in the image processing system of the present embodiment. (A) is an example of an original image input to the image display means, (b) is a slide bar that corrects the brightness of the original image and the image displayed on the image display means, and (c) is an image It is a figure which shows the one result which produced | generated the input-output brightness conversion characteristic according to the direction and movement amount which slid the knob in the process part, (d) is the direction and movement amount which slid the knob in the image processing part. FIG. 10 is a diagram showing one result of generating input / output brightness conversion characteristics as shown in FIG. 5C and performing brightness conversion on the original image.
FIG. 4 is a diagram illustrating an example of information recorded in the second correction information DB in the image processing system according to the present embodiment.

<S100>
First, the process of S100 will be described. FIG. 3A is an example of an original image input to the image processing unit 101. When the original image is input to the image processing apparatus 101, the image processing apparatus 101 detects this and requests the correction item number from the manual correction input means 100.

  The correction items include brightness correction, contrast correction, skin color hue correction, skin color saturation correction, gray hue correction, gray saturation correction, blue sky and sea hue correction, blue sky and sea saturation correction, and smoothing. There are various types such as correction and edge enhancement.

  These correction items are given unique numbers, and the correction items can be easily specified by the correction item numbers.

  In this processing, correction item number = 1 will be described below as brightness correction. The manual correction input unit 100 sends correction item number = 1 to the image processing unit 101.

<S101>
In the process of S101, the image display unit 102 displays information that prompts the input of the first correction amount.

<S102>
In the process of S102, the first correction amount is input from the manual correction input unit 100 while referring to the image and information displayed on the image display unit 102. The first correction amount with the correction item number = 1 will be described later. The image processing unit 101 corrects the original image with the first correction amount and displays the corrected image on the image display unit 102.

(Brightness correction of correction item number = 1 and its first feature value)
The image display means 102 displays an original image (FIG. 3A) and a slide bar (FIG. 3B) for correcting the brightness of the image. The user slides the knob within the range from “dark” on the left side of the slide bar to “bright” on the right side so that the user feels desirable brightness while viewing the image.

  The image processing unit 101 generates input / output brightness conversion characteristics as shown in FIG. 3C according to the direction and amount of movement of the knob, and performs brightness conversion on the original image. The “as is” in the slide bar is a point having the lightness L1, and the input / output lightness conversion characteristics are 2. It is. When the knob is moved to the right, the lightness L1 moves toward the lightness L2, and as an input / output lightness conversion characteristic with (0, 0) and (100, 100) fixed as 1. Generate the curve.

The direction in which this knob is slid and the amount of movement are the first correction amount. Here, the lightness L will be supplemented. The description will be made assuming that the original image of this processing conforms to the international standard sRGB.
sRGB
IEC (International Electrotechnical Commission) international standard for color space established in 1998 sRGB data is converted to tristimulus XYZ data in a 3x3 matrix determined in advance based on appropriate gamma correction and reference white . RGB after γ conversion can be converted into tristimulus values XYZ by the following equation (1).
X = α00 × R + α01 × G + α02 × B
Y = α10 × R + α11 × G + α12 × B
Z = α20 × R + α21 × G + α22 × B (1)

Further, the tristimulus values XYZ are converted into L * a * b * data using a defined conversion formula as a CIELAB uniform color space. It is shown in the following formula (2).
CIELAB is an international standard for uniform color space defined by the International Lighting Association (CIE) and JIS.
L * = 116 * (Y / Y0) ^ (1/3) -16 (however, Y / Y0> 0.008856)
a * = 500 × [(X / X0) ^ (1/3) − (Y / Y0) ^ (1/3)]
b * = 200 × [(Y / Y0) ^ (1/3)-(Z / Z0) ^ (1/3)
(X0, Y0, Z0 are the values of the reference reflecting surface) ..... Expression (2)
This L * represents the perceptual lightness L.

<S103>
The process of S103 will be described. Information indicating whether or not the input of the first correction amount related to the correction item has been completed is displayed on the image display unit 102. When the input is completed, completion is input from the manual correction input unit 100. If completed, the process proceeds to the next step S104. If not completed, the process returns to S102, and the first correction amount is input.

<S104>
The process of S104 will be described. Upon receipt of the completion, the image processing unit 101 sends the original image to the image feature amount calculation unit 103, receives the correction item number from the manual correction input unit 100, and calculates the first feature amount based on the correction item.

Hereinafter, the first feature amount will be described. The above-described conversion from RGB to L * a * b * may be performed on all pixels of the original image, or may be performed by reducing the resolution of the image. In any case, N * of all the pixels to be converted are obtained, divided by the total number of pixels N, and the average brightness level of the image is obtained by the following equation (5).
Lave = ΣL * / N (5)
This “Lave” represents the first feature amount in FIG. 3A, that is, the average brightness of the image. A dark image has a small level, and a bright image has a large level.

  As another first feature amount calculation method, the brightness range of L * is equally divided, and the brightness distribution of the image is obtained by counting the number of pixels corresponding to each divided brightness, and the distribution shape is utilized. Also good. For example, when the lightness range is divided into three parts, and each of the lightness ranges is high lightness, medium lightness, and low lightness, the ratio of each lightness area to the entire image and its dispersion are obtained, and the ratio and dispersion of the lightness area having the highest ratio are calculated One feature amount is set.

<S105>
The process of S105 will be described. The correction similarity determination unit 104 receives the correction item number and the first correction amount from the manual correction input unit 100. Further, the first feature amount is received from the image feature amount calculation unit 103. The correction similarity determination unit 104 accesses the second correction information DB 105 and limits the search area from the second correction information recorded by the correction item number.

  A first similarity determination is performed using the first feature amount and the second feature amount in the limited range, and the second feature amount is limited.

  Further, the second similarity determination is performed by using the first correction amount and the second correction amount in the limited range, and the second correction amount b is specified. The second correction amount b is sent to the image processing unit 101 and corrected by the image processing unit 101 to obtain a corrected image. The corrected image and the next correction item are displayed on the image display means 102.

  An example of information recorded in the second correction information DB 105 is shown in FIG. Each correction item includes a combination of the second feature amount and the second correction amounts a and b.

  Since it is known which correction item is the first correction amount input from the manual correction input means 100 (correction item number = 1: brightness correction in this process), the same correction item from the second correction information DB 105 is used. Number = 1 can be specified. In the second correction information DB 105, a second feature amount calculated for a certain image for each correction item and a second correction amount a and b by a hand of an image processing expert are paired and recorded in advance. Yes.

  The second feature amount with the correction item number = 1 is divided into M1, (1,1) to (1, M1). Further, the second feature amount (1, 1) is set as five second correction amounts from (1, 1, 1) a, b to (1, 1, 5) a, b. This indicates that T types of correction items are recorded with the same structure.

FIG. 3D shows an example of the correction item number = 1 and the second correction amount (1,1,1) a, b to (1,1,5) a, b of the second feature amount (1,1). ). The second correction amount b sent to the image processing means 101 indicates that the input / output conversion characteristics are numbers 1 to 5. In response to this, in the image processing means 101, the image processing is specified by the following equation (6), and the processing is executed.
f (correction item number, second correction amount b) (6)

Next, a similarity determination method for specifying the second correction amounts a and b will be described.
The first feature amount in the original image FIG. 3A is the average brightness level of the image calculated from the equation (5). First, a first similarity determination is made as to which second feature value corresponds. In the first similarity determination, the absolute value of the difference between “Lave” and all the second feature values is obtained from the following equation (7), and the second feature value having the minimum value is selected.
Δ1 = abs (Lave−second feature amount (1, 1))
Δ2 = abs (Lave−second feature amount (1, 2))
・
・
・
ΔM1 = abs (Lave−second feature quantity (1, M1)) (7)

In this process, it is assumed that Δ1 is the minimum value. Second in second feature (1, 1)
It can be seen that there are five correction amounts (1,1,1) a, b to (1,1,5) a, b. Next, one of them is specified by the second similarity determination. It is assumed that the first correction amount is input from the manual correction input unit 100 as L2 in FIG.

As shown in FIG. 3D, the second correction amount a from Ls4 to Ls0 is recorded in the second correction amount a. Therefore, using the equation (7), the following equation (8) is obtained.
Δt1 = abs (L2−Ls4)
Δt2 = abs (L2-Ls3)
Δt3 = abs (L2−Ls2)
Δt4 = abs (L2−Ls1)
Δt5 = abs (L2−Ls0) (8)

After obtaining Equation (8), the second correction amount b indicating the minimum value of the difference is specified. If Δt2 is the minimum value in the example, “2” in the second correction value (1, 1, 2) b is output from the second correction information DB 105 as the second correction value b.

In the image processing means 101, the following equation (9) is obtained, and the original image is corrected by the input / output brightness conversion characteristic “2” in FIG.
f (1,2) ... Formula (9)

  In FIG. 3, item number = 1: brightness correction is taken as an example, and the first correction amount, the second correction amount a, and the second correction amount b have been described. The brightness correction is correction based on the input / output brightness conversion characteristics. Needless to say, the first correction amount, the second correction amount a, and the second correction amount b are different depending on the correction item.

<S106>
In the processing of S106, if the next correction item is executed, S107 is executed, and if not executed, the correction is completed.

<S107>
In the process of S107, the image corrected in S105 is set as an original image in the image processing means 101, and the manual correction input means 100 moves to the next correction item, and returns to S101.

  For users who are unfamiliar with image processing, it is very difficult to perform advanced processing such as specialization accustomed to image processing. Therefore, in the image processing system of the present embodiment, the user roughly performs manual correction according to the displayed correction items, selects correction information having a similar correction similarity from the correction amount and the feature amount of the image, and is registered in advance. By applying special corrections that are familiar to image processing, advanced corrections can be made while taking advantage of the user's own preferences.

(Second Embodiment)
FIG. 5 is a block diagram showing an example of an image processing system having the image processing apparatus (1 in a broken line frame) and image display means 102 of the present embodiment.
The difference from the image processing apparatus according to the first embodiment is that it has a user ID detection unit 106 that can identify a user.

  The image processing apparatus (1 in a broken line frame) includes a manual correction input unit 100, an image processing unit 101, an image display unit 102, an image feature amount calculation unit 103, a correction similarity determination unit 104, and a second correction information DB (data list). ) 105 and user ID detection means 106.

  The manual correction input unit 100 inputs the first correction amount manually corrected by the user according to the correction item for the original image to be corrected.

  The image processing unit 101 performs image processing based on the correction item and the first correction amount, sends the image processing result to the image display unit 102, and displays an image.

  The image display means 102 displays various information for supporting the input of the correction item and the first correction amount.

  The image feature amount calculating unit 103 is a unit that calculates the feature amount of the original image for each correction item.

  The correction similarity determination unit 104 determines the similarity of correction from the first and second correction amounts and the first and second feature amounts for each correction item, and the user's manual operation is performed from the second correction information DB 105 at the subsequent stage. A second correction amount b close to the correction is specified.

  The second correction amount b is fed back to the image processing means 101, the image processing is executed on the original image with the second correction amount b corresponding to the correction item, and the corrected image is displayed on the image display device 102.

  For each correction item, the second correction information specified by the first feature amount of the original image, that is, the second feature amount and the second correction amounts a and b are stored. When the image to be corrected next time has the same first feature amount, the use of the stored information can save time and effort to repeatedly input the first correction amount, thereby improving efficiency.

  The user ID detection means 106 can specify a user. The second correction information DB 105 is accessed with the detected user ID. Information for identifying a user includes, but is not limited to, input of an ID number, an RFID tag, and an IC card.

Processing of the image processing system having the above-described configuration of the image processing apparatus and the image display means will be described with reference to FIGS. 6, 7, and 8. FIG.
FIG. 6 shows an example of information registered in the user area in the image processing system of the present embodiment.
7 and 8 are flowcharts showing an example of processing performed in the image processing system of this embodiment.
In the flowcharts of FIGS. 7 and 8, the processing from S101 to S107 is the same as that of the first embodiment.

<S200>
First, in the process of S200, the user ID detection unit 106 detects the user ID.

<S201>
The process of S201 will be described. The user ID detection unit 106 accesses the second correction information DB 105 and searches whether the second correction information that matches the user ID is recorded in the user area. FIG. 6 shows an example of information registered in the user area. User A's correction item number = 1 includes the second feature amount (1, 1), the second correction amount (1, 1, 3) a, b, and the second correction amount (1, 1, 4) a, b. In the two correction item numbers = 2, one of the second feature amount (2, 2) and the second correction amount (2, 2, 1) a and b is registered.

<S202>
The process of S202 will be described. As a result of the search, if there is registered second correction information, S203 is skipped, and if not, S203 is skipped.

<S203>
In the process of S203, the user ID detection means 106 sends the correction item numbers to the manual correction input means 100 in the order of the recorded correction items.

<S204>
In the process of S204, if there is registered second correction information, the process jumps to S104, and if not, S101 is executed.

<S205>
In the process of S205, if there is registered second correction information, S207 is executed, and if not, S105 and S206 are executed.

<S206>
In the process of S206, the second correction information is recorded in the same correction item in the area that matches the user ID in the user area.

<S207>
The process of S207 will be described. The corrected similarity determination unit 104 receives the first feature amount from the image feature amount calculation unit 103. Further, a location that matches the user ID in the user area of the second correction information DB 105 is accessed, and the search area is limited by the correction item number. The first feature amount and the limited second feature amount are used to determine the first similarity, the second feature amount is limited, and all the second correction amounts b in the limited range are image processing means 101. To send.

  The image corrected by the second correction amount b by the image processing unit 101 is displayed on the image display unit 102. With reference to the displayed image, the image is selected from the manual correction input means 100.

  In the image processing system of the present embodiment, in addition to the effects of the first image processing system, correction information registered in the past can be registered, and the effect of reducing the input of manual correction can be reduced. Also have.

  The image processing systems of the first and second embodiments can also be applied to an image correction method (image processing method). The image correction method (image processing method) can be realized by executing a prepared program on a computer such as a personal computer or a workstation. This program is recorded on a computer-readable recording medium such as a hard disk, a flexible disk, a CD-ROM, an MO, and a DVD, and is executed by being read from the recording medium by the computer. Further, this program may be a transmission medium that can be distributed via a network such as the Internet.

1 is a block diagram illustrating an example of an image processing system according to a first embodiment. It is a flowchart figure which shows the process of the image processing system concerning 1st Embodiment. It is a figure which shows the information handled in the image processing system concerning 1st Embodiment. (A) is an example of an original image input to the image display means, (b) is a slide bar that corrects the brightness of the original image and the image displayed on the image display means, and (c) is an image It is a figure which shows the one result which produced | generated the input-output brightness conversion characteristic according to the direction and movement amount which slid the knob in the process part, (d) is the direction and movement amount which slid the knob in the image processing part. FIG. 10 is a diagram showing one result of generating input / output brightness conversion characteristics as shown in FIG. 5C and performing brightness conversion on the original image. It is a figure which shows the example of information currently recorded on 2nd correction information DB in the image processing system concerning 1st Embodiment. It is a block diagram which shows an example of the image processing system concerning 2nd Embodiment. An example of the information registered in the user area in the image processing system according to the second embodiment is shown. 7 is a flowchart illustrating an example of processing performed in an image processing system according to a second embodiment. 7 is a flowchart illustrating an example of processing performed in an image processing system according to a second embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Image processing apparatus 100 Manual correction | amendment input means 101 Image processing means 102 Image display means 103 Image feature-value calculation means 104 Correction | amendment similarity determination means 105 2nd correction information DB
106 User ID detection means

Claims (7)

Image display means for displaying images and image correction items;
Means for inputting or calculating first correction information according to the correction item;
Second correction information storage means for storing or holding second correction information in advance for each correction item;
An image processing system comprising correction similarity determination means for determining similarity between first correction information and second correction information. The correction similarity determination unit performs the first similarity determination from the first feature amount and the second feature amount included in the first correction information, and the first correction amount and the second correction amount included in the second correction information. To determine the second similarity,
The image processing system according to claim 1, wherein a second correction amount is specified based on two determination results of the first similarity determination and the second similarity determination.   In the first similarity determination, the difference between the first feature amount and the second feature amount is minimum, and in the second similarity determination, the difference between the first correction amount and the second correction amount is minimum. The image processing system according to claim 2, wherein: A user ID detecting means for performing user recognition;
The detected user ID and the second correction information specified by the correction similarity determination unit are recorded by the second correction information storage unit,
4. The image processing system according to claim 1, wherein second correction information recorded from a user ID, a correction item, and a first feature amount of an image is specified. 5. An image display process for displaying an image and image correction items;
Inputting or calculating first correction information according to the correction item;
A second correction information storing step for storing or holding second correction information in advance for each correction item;
An image processing method comprising: a correction similarity determination step for determining a similarity between the first correction information and the second correction information. Image display processing for displaying images and image correction items;
A process of inputting or calculating first correction information according to the correction item;
A second correction information storing process for storing or holding second correction information in advance for each correction item;
An image processing program that causes an image processing system to execute correction similarity determination processing for determining similarity between first correction information and second correction information.   A recording medium on which the image processing program according to claim 6 is recorded.

Images