JP2010004300A - Image processing method, image processing apparatus, image processing program, and printer - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To perform image processing suitable for a variety of objects. <P>SOLUTION: When the image processing is performed to an image indicated by image data 13b, a human face as a first object and an animal face or a doubtful face as a second object, etc. are detected from the image first. Then, first image processing (steps S270, S280) is executed to the whole of the image on the basis of the detected human face as the first object, and second image processing (steps S300, S310) is performed to a region where the second object is detected. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an image processing method, an image processing apparatus, an image processing program, and a printing apparatus.

There has been proposed one that detects a human face from an image and sets image processing parameters in accordance with the detection result (see Patent Document 1). In this document, when a human face is detected, image processing parameters for portraits are set, and when no human face is detected, image processing parameters for person images are set. ing.
JP 2005-63406 A

However, an object included in a photograph taken with a digital still camera is not limited to a single type, and a plurality of types of objects may be included at the same time. In that case, the technique disclosed in the above-described literature has a problem that appropriate image processing cannot be executed according to the property of each object. In particular, in the case of a human photograph, shooting is performed while paying attention to a person's face, but an unintended subject (object) may appear in the background or the like. In many cases, the user wants to make corrections so as not to stand out for such unintended objects, but the problem is that image processing similar to that of a person's face is also performed on unintended objects. there were.
The present invention has been made in view of the above problems, and an object thereof is to provide an image processing method, an image processing apparatus, an image processing program, and a printing apparatus that can perform image processing suitable for various objects.

  In order to solve the above-mentioned problem, in the invention according to claim 1, when performing image processing on an image indicated by image data, first, a first object and a second object are detected from the image. Then, first image processing is executed on the entire image based on the detected first object. For example, the color and brightness of the entire image are corrected based on the color and brightness of the first object. On the other hand, for the area where the second object is detected, the second image processing is executed separately from the first image processing. Since the second image processing is performed on the area where the second object is detected, a local image processing effect can be obtained.

  Further, as a suitable example of the first image processing and the second image processing, the first image processing and the second image processing may be different types of image processing. For example, the color and brightness of the entire image may be corrected as the first image processing, and sharpness (enhancement) processing, blurring processing, and the like may be performed as the second image processing. By doing so, it is possible to emphasize the preferred second object and blur the undesirable second object while making the overall brightness and hue appropriate.

  In addition, when performing the second image processing locally on the second object, it is desirable that the user can confirm the influence range in advance. Therefore, display indicating the second object may be performed in the image. In this way, it is possible to know in advance in what area the second image processing is performed. For example, a region where the second object is detected may be surrounded by a frame. When such display is performed, an instruction for the second image processing may be received. Thus, it is possible to know what kind of the second object has been detected. For example, it is possible to select such that the preferable second object is emphasized and the unfavorable second object is blurred.

  As a specific example of the first object and the second object, the first object may be a human face, and the second object may be similar to a human face. By doing in this way, the said 1st image process of the whole can be performed based on the face of a person considered to be a main subject fundamentally.

  Furthermore, the technical idea of the present invention can be embodied not only in a specific image processing method but also in an image processing apparatus. That is, the present invention can be specified as an image processing apparatus having means corresponding to each step performed by the above-described image processing method. Of course, when the above-described image processing apparatus reads the program to realize each of the above-described means, the technical features of the present invention can be applied to a program for executing a function corresponding to each of the means and various recording media on which the program is recorded. Needless to say, the idea can be embodied. Needless to say, the image processing apparatus of the present invention can be distributed not only by a single apparatus but also by a plurality of apparatuses. Further, the image processing method of the present invention may be realized in a printing apparatus such as a printer or an image input apparatus such as a digital still camera.

Here, embodiments of the present invention will be described in the following order.
(1) Computer configuration:
(2) Object detection processing:
(3) Image processing / printing processing:
(4) Modification:

(1) Computer configuration:
FIG. 1 shows a schematic configuration of a computer that specifically realizes the object detection apparatus of the present invention. In FIG. 1, a computer 10 includes a CPU 11, a RAM 12, an HDD 13, a USB interface (I / F) 14, an input device interface (I / F) 15, and a video interface (I / F) 16 connected by an internal bus 18. The HDD 13 stores various program data 13a and a plurality of image data 13b, 13b, 13b. The CPU 11 reads the program data 13a and executes processing based on the program data 13a while using the RAM 12 as a work area. A printer 20 and a digital still camera 30 are connected to the USB I / F 14, and a mouse 40 and a keyboard 50 are connected to the input device interface 15. Further, a display 60 is connected to the video I / F 16.

  FIG. 2 shows a software configuration of a program executed on the computer 10. In the computer 10, an operating system (O / S) (not shown) is executed, and a printer driver P is executed on the O / S. The printer driver P generally includes an object detection unit P1, an image processing unit P2, and a print execution unit P3. The object detection unit P1 includes a determination condition setting unit P1a, a detection window setting unit P1b, a similarity calculation unit P1c, a threshold determination unit P1d, a color determination unit P1e, and a contrast determination unit P1f. The object detection unit P1 detects a human face as a first object, a suspicious face and an animal face as a second object, and the like. Note that a suspicious face is a suspicious face that should not be originally captured, mainly taken by a digital still camera, for example, a face that is unintentionally reflected by being reflected by a reflector around the subject. And a face reflected by an optical / electrical afterimage at the time of photographing. In addition, a case where the face looks like a face by accidental arrangement such as leaves or stones is not included in the category of a suspicious face. Note that the suspicious face and the animal's face are examples of what is detected as the second object, and other objects may be detected as the second object.

  The determination condition setting unit P1a designates various determination parameters to a detection window setting unit P1b, a similarity calculation unit P1c, a threshold determination unit P1d, a color determination unit P1e, and a contrast determination unit P1f, which will be described later. Different determination parameters are specified depending on whether the object is the first object or the second object. The detection window setting unit P1b generates QVGA (320 × 240) size detection image data whose size has been converted from the target image data 13b, and sets a detection window (detection region) SW in the detection image data. By sequentially shifting the size, position, and angle of the detection window SW, it is possible to detect the first object and the second object that exist in all sizes, positions, and angles in the detection image data. The range of the size of the detection window SW when detecting the first object and the second object is designated by the determination condition setting unit P1a, and the size of the detection window SW is shifted within the range according to the designation. In the present embodiment, the image data 13b is Exif format image data, and each pixel has color information of RGB values by decoding at the time of processing. Similarly, each pixel of the detection image data has color information of RGB values.

  The similarity calculation unit P1c acquires an image included in the detection window SW that is sequentially shifted, and acquires a plurality of feature amounts (color, brightness, contrast, and the like in various regions in the detection window SW) from the image. Then, a plurality of feature amounts are input to the similarity evaluation function 13c stored in the HDD 13, and the similarity is calculated as a result. The similarity evaluation function 13c according to the present embodiment is a neural network in which a plurality of feature amounts obtained from a large number of samples (templates) including faces are learned by a known learning method as learning data. The output value indicates the similarity. The similarity is normalized so as to have a value of 0 to 1000, and it can be determined that the larger the value is, the higher the similarity is with the sample including the face. That is, when the similarity is close to 1000, it can be determined that a face is included in the detection window SW. The threshold determination unit P1d compares the threshold specified by the determination condition setting unit P1a with the similarity, and determines whether the first object and the second object exist in the detection window SW according to the result. To do.

  When it is assumed that the face as the first object exists in the detection window SW, the color determination unit P1e determines whether or not the evaluation pixel that constitutes a part of the face shows an appropriate skin color. Specifically, the RGB value of the evaluation pixel is converted into an HSB value by a known conversion formula, and the detection window SW depends on whether the HSB value belongs to a predetermined color gamut designated by the determination condition setting unit P1a. It is determined whether there is a first object (human face) and a second object (animal face, suspicious face, etc.). The contrast determination unit P1f calculates a contrast value in the detection window SW, and determines whether the first object (person's human being) is in the detection window SW depending on whether or not the contrast value satisfies a predetermined threshold value specified by the determination condition setting unit P1a. Face) and the second object (animal face, suspicious face, etc.) are determined.

  On the other hand, the image processing unit P2 includes a UI unit P2a, a first image processing unit P2b, and a second image processing unit P2c. The UI unit P2a displays the target image data 13b on the display 60, and displays a frame indicating the area where the second object is detected in the display of the image data 13b. Then, designation of the type of image processing for the area where the frame is displayed is received by the mouse 40 and the keyboard 50. In the present embodiment, it is possible to specify, for each frame, either enhancement processing or blurring processing. The first image processing unit P2b analyzes the color and brightness indicated by the face as the first object detected from the image data 13b, and performs color and brightness correction processing on the entire image data 13b. Run as. The second image processing unit P2c executes the type of image processing accepted by the UI unit P2a only on the portion where the second object is detected in the image data 13b.

  The print execution unit P3 includes a size conversion unit P3a, a color conversion unit P3b, a halftone processing unit P3c, and a rasterization unit P3d. The size converter P3a converts the size of the image data 13b after the image processing described above so as to correspond to the print size. The color conversion unit P3b converts the image data 13b in which the color of each pixel is expressed by RGB values into the image data 13b expressed by the color of ink used by the printer 20. For example, when the printer 20 is an inkjet printer capable of ejecting CMYK ink, the color of each pixel is converted into image data 13b expressed by CMYK values. The halftone processing unit P3c performs halftone processing such as an error diffusion method and a dither method on the color-converted image data 13b, thereby generating halftone data that specifies whether or not to eject ink for each pixel. Convert. Next, the rasterizing unit P3d rearranges the halftone data in the printing order, and attaches a header that specifies the printing paper and the printing resolution, thereby generating print data that can be output by the printer 20. The print data is output to the printer 20 via the USB I / F 14, and the printer 20 performs printing based on the print data.

(2) Object Detection Processing FIG. 3 shows the flow of object detection processing. In step S100, the determination condition setting unit P1a determines whether the actual size of the subject in the image data 13b can be specified. Specifically, the target image data 13b acquired from the HDD 13 is decoded, and shooting information attached as tag information in the image data 13b is acquired. When the camera model, subject distance A [m], lens focal length B [mm], and digital zoom magnification D are all attached as shooting information, the actual size of the subject in the image data 13b can be specified. Is determined. Here, the reason why the actual size of the subject in the image data 13b can be specified based on the above information will be described.

  FIG. 4 schematically shows how to calculate the actual size of the subject in the image data 13b. In the figure, a cross section of an optical path when a certain subject is photographed by a digital still camera is schematically shown, and an image of the subject is formed on an image sensor provided in the digital still camera via a lens. First, since the camera model can be specified, the effective size S [mm] of the image sensor and the number of image pickup pixels P [pixel] can be specified. In the present embodiment, it is assumed that the image data 13b having the same number of pixels as the number of imaging pixels P is generated. In addition, when digital zoom is performed, image data having a number of pixels (P / D) that is a part of the number of pixels P (P / D) is generated in the first stage, and then D-times pixel interpolation is performed. Thus, the following description will be made assuming that the image data 13b having the same number of pixels as the number of imaging pixels P is finally generated.

First, the size of each image pickup pixel on the image sensor can be represented by S / P [mm], and the size of the subject having the number of pixels C [pixel] formed on the image sensor in the image data 13b is C. X S / P [mm]. When the digital zoom is performed, the size of the subject having the number of pixels C [pixels] formed on the image sensor in the image data 13b can be expressed by C × S / (P × D) [mm]. When the size of the upper subject of the image sensor can be specified as described above, the actual subject size X [mm] can be calculated by the following equation (1) from the positional relationship between the subject, the image sensor, and the lens shown in FIG. Can be represented.
X = (A × 1000) × C × S / (P × D × B) (1)
Thereby, when a subject having an arbitrary number of pixels C on the image data 13b is detected, the actual size of the subject can be specified. If it is determined in step S100 that the actual size of the subject can be specified, the determination condition setting unit P1a detects the detection window setting unit P1b, the similarity calculation unit P1c, and the threshold determination unit P1d in step S110. Various determination parameters are specified in the color determination unit P1e and the contrast determination unit P1f. In step S120, determination processing (first determination means) is executed.

  FIG. 5 schematically shows a flow of a series of first determination processes (step S120) executed by the detection window setting unit P1b, the threshold determination unit P1d, the color determination unit P1e, and the contrast determination unit P1f. In this embodiment, the first determination process (step S120), the first determination process (step S150), and the third determination process (step S180) are executed, but the determination condition setting unit P1a is different for each determination parameter. By designating, it is possible to obtain determination results with different personalities. Here, the first determination process (step S120) will be described first. In step S110, the determination condition setting unit P1 specifies the range of the size of the detection window SW to be shifted with respect to the detection window setting unit P1b as 10 to 40 cm.

  As a result, the detection window SW can be set in a range having a reasonable size as a human face. In addition, in order to set the size of the detection window SW to 10 to 40 cm, the relationship of the above-described formula (1) is used. By substituting 10 to 40 cm (100 to 400 mm) for the size X [mm] of the subject in the expression (1), the number C of pixels of the image data 13b corresponding to the size can be specified. Further, by converting the number of pixels C into the QVGA size, the size of the detection window SW to be set in the detection image data can be specified.

  The detection window setting unit P1b leaks in the detection image data in which the image data 13b is reduced by sequentially shifting the position and angle of the detection window SW while shifting the size of the detection window SW within the specified range. Detects a face without any noise. Each time the detection window setting unit P1b shifts the detection window SW, the similarity calculation unit P1c acquires an image belonging to the detection window SW, and acquires a plurality of feature amounts input to the similarity evaluation function 13c from the image. And the similarity (0-1000) as an output value of a neural network is calculated by inputting a feature-value into the similarity evaluation function 13c. This similarity is an index value indicating how similar an image belonging to the current detection window SW is to a human face.

  In step S110, the determination condition setting unit P1 designates the similarity threshold as 300 for the threshold determination unit P1d. The threshold determination unit P1d obtains the similarity from the similarity evaluation function 13c, and if the similarity is equal to or greater than the threshold, the color determination unit P1e and the contrast determination unit P1f have colors and contrasts that are appropriate for a human face. It is determined whether or not. On the other hand, if the similarity is smaller than the threshold, the determination of the color and contrast for the detection window SW is skipped, and the detection window setting unit P1b sets the next detection window SW. In step S120, the color determination unit P1e acquires the RGB values of the pixels belonging to the detection window SW, and determines whether or not the RGB values are appropriate as the skin color. Specifically, the RGB value is converted into an HSV value, and it is determined whether or not the HSV value is included in a predetermined color gamut in the HSV space. This color gamut is designated in advance by the determination condition setting unit P1.

  FIG. 6 shows a color gamut in which the color determination unit P1e performs determination. In the figure, an HSV space is shown. By setting a predetermined hue / saturation range (an elliptical area on the hue plane) in the HSV space, a color gamut appropriate for a human skin color is designated. is doing. In step S110, the determination condition setting unit P1 designates a color gamut corresponding to the skin color, and in the first determination process in step S120, it is determined whether or not the pixel belonging to the detection window SW is a strict skin color. If the skin color is determined, the contrast determination unit P1f continues to determine the detection window SW. On the other hand, when the skin color is not determined, it is similar to the face, but it is determined that the face color is not normal and the detection window SW includes an animal face or a suspicious face as the second object, The detection window setting unit P1b sets the next detection window SW, and performs a determination process for the next detection window SW.

The contrast determination unit P1f acquires the RGB values of the pixels belonging to the detection window SW, and calculates the luminance Y value of each pixel based on the RGB values. Then, the maximum RGBY value R _max G _max B _max Y _max of all pixels belonging to the detection window SW is detected, and the minimum RGBY value R _min G _min B _min Y _min of all pixels is detected. The maximum value among R _max / R _min , G _max / G _min , B _max / B _min , and Y _max / Y _min is set as the contrast value. In step S110, the determination condition setting unit P1 previously sets a threshold value for the contrast value for the color determination unit P1e.

  Therefore, in step S120, it is determined whether or not an image belonging to the detection window SW has a high contrast value. When an image belonging to the detection window SW has a high contrast value, it is determined that the detection window SW includes a face as the first object, and the detection window setting unit P1b determines that the next detection window SW Is set, and determination processing for the next detection window SW is performed. On the other hand, when the image belonging to the detection window SW does not have a high contrast value, it is determined that the detection window SW includes an animal face or a suspicious face as the second object, and the detection window setting unit P1b Sets the next detection window SW and performs a determination process for the next detection window SW. When the setting / determination is completed for the detection windows SW for all positions and angles in all designated size ranges, the first determination process is completed.

  Through the first determination process described above, each time the detection window setting unit P1b sequentially shifts the detection window SW, a series of determination processes for each detection window SW is executed. Whether or not there is an animal face or a suspicious face as the second object. Here, as a condition for determining that there is a human face as the first object, it is required that the similarity with the human face is high, the color indicates skin color, and the contrast is large. Become. In addition, as a condition for determining that there is an animal face or a suspicious face as the second object, the similarity to the human face is high but the color is not skin color, or the similarity to the human face However, it is required that the contrast is small. For example, in the case of a dog or a cat, the arrangement of facial organs is similar to some extent, but is often not a skin color. Further, in the case of a suspicious face due to an afterimage or reflection, the shape of the face is similar to some extent, but contrast is often not obtained.

  When the first determination process described above is completed, determination results for all detection windows SW in which the first or second object is detected are acquired in step S120, and a table describing the determination results is stored in the HDD 13 as detection data 13d. Store (step S130). In step S140, as in step S110, the determination condition setting unit P1a designates various determination parameters to the detection window setting unit P1b, the similarity calculation unit P1c, the threshold determination unit P1d, the color determination unit P1e, and the contrast determination unit P1f. . In step S150, a second determination process (second determination unit) is executed. The second determination process is a process similar to the first determination process described above, but since the determination condition setting unit P1a sets a determination parameter different from that in step S110, a determination result having a different property is obtained. Hereinafter, the second determination process in step S150 will be described.

  FIG. 7 schematically illustrates a flow of a series of second determination processes (step S150) executed by the detection window setting unit P1b, the threshold determination unit P1d, the color determination unit P1e, and the contrast determination unit P1f. The determination condition setting unit P1a specifies the range of the detection window SW to be set in the detection window setting unit P1b as a range smaller than 10 cm or larger than 40 cm. That is, a detection window SW having a size that is not detected in the determination process in step S120 is set. For the threshold determination unit P1d, the similarity threshold is designated as 100. By doing in this way, the threshold determination unit P1d can be detected even for an object that is less likely to be a human face than in step S120. In addition, the color gamut width and the contrast value threshold specified by the color determination unit P1e and the contrast determination unit P1f do not change, but the determination by the color determination unit P1e and the contrast determination unit P1f is negative. It is determined that there is an animal face or suspicious face as two objects.

  In other words, if the face is clearly different in size from the human face but is similar in image to the human face, if the contrast is low and the skin color is not shown, the animal face or suspicious face, etc. It is determined that Animal faces, suspicious faces, etc. are not very similar to human faces but can be considered to be somewhat similar. Therefore, by relaxing the similarity threshold to 100, it is possible to detect animal faces, suspicious faces, etc. it can. In addition, by not detecting objects with high contrast or skin color as animal faces or suspicious faces, etc., images and photographs of human faces on backgrounds, posters, clothes, etc. can be used for animal faces or suspicious faces. Can be prevented as detected. The determination of the skin color here is not intended to detect the human face as the first object, so the skin color gamut may be made larger than in the case of step S120.

  Similarly, the threshold value of the contrast value may be relaxed compared to the case of step S120. When the setting of the detection windows SW for all positions and angles is completed, the second determination process (step S150) is completed. When the above second determination process is completed, determination results for all detection windows SW in which the second object is detected in step S150 are acquired, and the determination results are added to the detection data 13d (step S160).

  In the above, the first object and the second object are detected by sequentially executing the first determination process and the second determination process based on the actual size of the subject. However, in step S100, the actual size of the subject is determined. If it is determined that it cannot be specified, the determination condition setting unit P1a sets a determination parameter in step S170, and executes a third determination process (step S180). The third determination process is similar to the first and second determination processes described above. However, since the determination condition setting unit P1a sets determination parameters different from those in steps S110 and S140, determination results having different properties are obtained. Will be. Hereinafter, the third determination process in step S180 will be described.

  FIG. 8 schematically shows a flow of a series of third determination processes (step S180) executed by the detection window setting unit P1b, the threshold determination unit P1d, the color determination unit P1e, and the contrast determination unit P1f. The determination condition setting unit P1a designates the range of the detection window SW to be set in the detection window setting unit P1b as the entire range. For the threshold determination unit P1d, the similarity threshold is designated as 500. Thus, by setting the largest threshold value, it is possible to strictly detect only the human face. Further, when the similarity satisfies the threshold value, the determination by the contrast determination unit P1f is not performed, and only the determination by the color determination unit P1e is performed. If the image belonging to the detection window SW has a similarity of 500 or more and the color indicates skin color, it is determined that a human face as the first object exists in the detection window SW. In other cases, it is determined that no object exists. When the setting of the detection windows SW for all positions and angles is completed, the third determination process (step S180) is completed. When the third determination process described above is completed, determination results for all detection windows SW in which the first object is detected are acquired in step S180, and a table describing the determination results is stored in the HDD 13 as detection data 13d ( Step S190).

  As described above, in this embodiment, in the first determination process (step S120), the second determination process (step S150), and the third determination process (step S180), different threshold values are set to 300, Set to 100,500. In the third determination process that requires the highest degree of similarity, only the face of a person as the first object can be accurately detected. In the first determination process that requires the next highest similarity, a human face as the first object, an animal face and a suspicious face as the second object, and the like are detected. The second determination process in which the degree of similarity is most relaxed is used to detect an animal face, a suspicious face, or the like as the second object. In this way, by using a plurality of threshold values (300, 100, 500) properly, only the similarity evaluation function 13c learned using a human face sample is used, and not only a human face but also an animal face. And suspicious faces can also be detected. Therefore, it is not necessary to prepare the similarity evaluation function 13c for detecting an animal face, a suspicious face, or the like.

(3) Image Processing / Printing Processing FIG. 9 shows the flow of image processing. First, in step S200, the UI unit P2a acquires the detection data 13d stored in steps S130, S160, and S190. In step S210, the UI unit P2a searches for the position, size, and angle of the detection window SW that is described in the detection data 13d as having been determined that an animal face, a suspicious face, or the like as the second object exists. In step S230, a preview screen is displayed on the display 60. On this preview screen, the image indicated by the image data 13b subjected to the object detection process is displayed with a predetermined resolution, and a rectangular frame indicating the outer edge of the detected detection window SW is displayed.

  FIG. 10 shows an example of a preview screen. In the figure, an image in which a person and a dog are photographed is displayed as an image indicated by the image data 13b. Compared with the face of the photographed person, the face of the dog is within the range of 10 to 40 cm, but does not show skin color, and therefore is determined as the second object as a result of the first determination process. Therefore, a frame is displayed so as to include the dog's face. Below the preview screen, there is provided a check box for receiving an instruction for image processing (second image processing) with respect to the area displayed by the frame, and any of printing, enhancement processing, and blurring processing is directly performed by the mouse 40. It is possible to select. If no animal face or suspicious face is detected as the second object, these check boxes are invalidated. A check box for receiving an image processing instruction for the entire area is also provided, and it is possible to select either printing or automatic correction processing as it is. Then, it is possible to click the print execution button with any of the check boxes.

  In step S230, the UI unit P2a receives a click on the print execution button, and acquires the selection status of each check box. It is determined whether or not the automatic correction process is checked (step S240). If the automatic correction process is checked, the first image processing unit P2b refers to the detection data 13d in step S250 and has detected the first object. Determine whether or not. When the human face as the first object is not detected, the first image processing unit P2b performs sampling without deviation from the pixels of the entire image indicated by the image data 13b, and based on the color and brightness indicated by the sampled pixels. The entire correction parameters are set (step S260). Specifically, when the entire screen is dark / bright, a correction parameter may be set so that the overall brightness is corrected upward / downward.

  Also, a correction parameter is set to correct the color balance (color fog correction) when the RGB value distribution balance is poor. On the other hand, when a human face as the first object is detected, the first image processing unit P2b performs sampling from pixels belonging to the detection window SW in which the face is detected, and the color or brightness indicated by the sampled pixels. Based on this, the entire correction parameter is set (step S270). Here too, correction parameters for brightness and color balance are set. However, since correction parameters are set by paying attention to the detected face area, it is possible to set appropriate correction parameters even when the face of a person is dark as a whole, for example, during backlight shooting. it can. As described above, when the correction parameter can be set, in step S280, the first image processing unit (first image processing means) P2b acquires each pixel of the image data 13b, and sequentially corrects the RGB value of each pixel. Go. If the automatic correction process is not checked, steps S260 to 280 are skipped.

  For example, when a correction parameter indicating that correction is to be performed brightly is set, a correction value is set such that the RGB value after correction of a predetermined gradation (for example, 64 gradations) is larger than the original RGB value, and the spline is set. A correction curve connecting the whole (maximum value, minimum value and the correction value) is generated by the curve. An automatic correction process can be performed by creating a correction LUT (lookup table) describing the relationship before and after such correction, and converting it to a corrected RGB value with reference to the LUT. When a human face is detected, a correction curve is set based on the brightness of the face in step S270, so that a correction result with an appropriate face brightness can be obtained. In step S280, all the pixels constituting the image data 13b are sequentially selected, and the above-described correction curve is applied. Thereby, the entire image indicated by the image data 13b is corrected.

  In the next step S290, the second image processing unit P2c determines whether printing, emphasis processing, or blurring processing is selected as it is for the area displayed by the frame. When printing is selected as it is (or when the check box is invalidated), the image processing ends. On the other hand, if either the enhancement process or the blurring process is selected, the image processing (second image processing means) designated for the animal's face or suspicious face as the second object in steps S300 and 310 is performed. Execute. For example, enhancement processing and blurring processing may be performed using an unsharp mask. In steps S300 and 310, only the pixels belonging to the detection window SW in which the face of the animal or the suspicious face as the second object is detected among the pixels constituting the image data 13b are sequentially corrected. Thereby, it is possible to correct only a part of the image where an animal face or a suspicious face exists. When the second image process is completed, the process proceeds to the next print process.

  As described above, the second object is somewhat similar to a human face, but is different in size, color, and contrast from a normal human face. As a typical example, the second object is an animal face or a suspicious face. And so on. However, what is specifically detected as the second object cannot be specified by the first and second determination processes described above. For example, if the object detected as the second object is the face of an animal such as a pet, the user is expected to emphasize it, but the object detected as the second object looks like a suspicious face or the like. If it is something that is not known, the user is expected to blur it to prevent it being misunderstood as a psychic photograph. Of course, it may be desired to print without any correction. As in the present embodiment, the second image processing suitable for the actual second object is executed by receiving an image processing instruction from the user while displaying a frame in an area where it is determined that the second object exists. Can do.

  FIG. 11 shows the flow of the printing process. In step S400, the size converter P3a converts the size of the image data 13b after image processing so as to correspond to the print size. In step S410, the color conversion unit P3b performs color conversion of the image data 13b in which the color of each pixel is expressed by RGB values into image data 13b of the amount of CMYK ink used by the printer 20. For example, color conversion is performed with reference to a color conversion LUT that defines the correspondence between the RGB color space and the CMYK color space. In step S420, the halftone processing unit P3c performs halftone processing such as an error diffusion method and a dither method on the image data 13b, thereby converting into halftone data that specifies whether or not to eject ink for each pixel. To do.

  Further, when the printer 20 ejects ink droplets of a plurality of sizes, it may be converted into halftone data that also specifies the size. In step S430, the rasterization unit P3d generates print data by performing rasterization processing for allocating halftone data to each main scanning pass and nozzles that eject ink. Furthermore, print data that can control the printer 20 is generated by attaching a header that specifies print paper and print resolution to the print data. In step S440, the print data is output to the printer 20 via the USB I / F 14, and the printer 20 actually performs printing based on the print data.

(4) Modifications In the above, the computer 10 executes object detection and image processing according to the present invention, but the functions of the computer 10 are executed by hardware / software resources provided in the printer 20. Also good. Of course, object detection and image processing on the image data 13b may be performed at the stage of image input. For example, the above-described processing may be executed by a digital still camera, a scanner, or the like. In the embodiment described above, whether to execute the first determination process, the second determination process, or the third determination process is switched depending on whether or not the size of the subject can be specified. You may make it switch by selection of.

It is a hardware block diagram of a computer. It is a software block diagram of a computer. It is a flowchart which shows the flow of an object detection process. It is a schematic diagram which shows a mode that the real size of a to-be-photographed object is calculated. It is a schematic diagram which shows the flow of a 1st determination process. It is a figure which shows the color gamut which a color determination part performs determination. It is a schematic diagram which shows the flow of a 2nd determination process. It is a schematic diagram which shows the flow of a 3rd determination process. It is a flowchart which shows the flow of an image process. It is a figure which shows a preview screen. 6 is a flowchart illustrating a flow of printing processing.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Computer, 18 ... Bus, 11 ... CPU, 12 ... RAM, 13 ... HDD, 13a ... Program data, 13b ... Image data, 14 ... USB I / F, 15 ... Input device I / F, 16 ... Video I / F 20 ... printer, 30 ... digital still camera, 40 ... mouse, 50 ... keyboard, 60 ... display, P ... printer driver, P1 ... object detection unit, P1a ... judgment condition setting unit, P1b ... detection window setting unit, P1c ... Similarity calculation unit, P1d ... threshold judgment unit, P1e ... color judgment unit, P1f ... contrast judgment unit, P2 ... image processing unit, P2a ... UI unit, P2b ... first image processing unit, P2c ... second image processing unit, P3: Print execution unit, P3a: Size conversion unit, P3b: Color conversion unit, P3c: Halftone processing unit, P3d: Rasterization unit

Claims (8)

An image processing method for performing image processing on an image indicated by image data,
Detecting a first object and a second object from the image;
Performing first image processing on the entire image based on the detected first object;
An image processing method, wherein second image processing is executed on an area where the second object is detected.   The image processing method according to claim 1, wherein the first image processing and the second image processing are different types of image processing.   The image processing method according to claim 2, wherein a display indicating the second object is performed in the image.   The image processing method according to claim 3, wherein an instruction for the second image processing is received in the display.   The image processing method according to claim 1, wherein the first object is a human face, and the second object is an object similar to a human face. An image processing apparatus that performs image processing on an image indicated by image data,
Detecting means for detecting the first object and the second object from the image;
First image processing means for performing first image processing on the entire image based on the detected first object;
An image processing apparatus comprising: second image processing means for executing second image processing on the detected second object. A computer-readable image processing program for causing a computer to execute a function of executing image processing on an image indicated by image data,
A detection function for detecting the first object and the second object from the image;
A first image processing function for performing first image processing on the entire image based on the detected first object;
A computer-readable image processing program causing a computer to execute a second image processing function for executing second image processing on an area where the second object is detected. A printing apparatus for printing an image indicated by image data,
Detecting means for detecting the first object and the second object from the image;
First image processing means for performing first image processing on the entire image based on the detected first object;
Second image processing means for performing second image processing on an area where the second object is detected;
A printing apparatus comprising: a printing execution unit that executes printing of the image data.

Images