JP4556602B2 - Image processing apparatus, printer including the same, image processing method, and image processing program - Google Patents

Description

  The present invention relates to an image processing apparatus, a printer including the same, an image processing method, and an image processing program. More specifically, the present invention relates to an image processing apparatus that performs predetermined image correction processing on an image and outputs the image, and a printer including the image processing apparatus. The present invention relates to an image processing program.

Conventionally, this type of image processing apparatus detects whether or not an input image includes an object such as a human face, and based on the detection result, the image content (such as a person image or a landscape image) is detected. There has been proposed one that corrects an image by determining and setting a correction parameter (see, for example, Patent Document 1). In this apparatus, when detecting the face of a person, a skin color region is detected or pixels corresponding to eyes and mouth are detected.
Japanese Patent Laid-Open No. 2004-236110

  However, in the above-described apparatus, when detecting an object such as a human face, the positional relationship of pixels corresponding to eyes and mouth and the arrangement of the object with respect to the entire image are taken into consideration. In some cases, object detection and image content determination cannot be performed correctly.

  An object of the image processing apparatus, the printer including the image processing apparatus, the image processing method, and the image processing program according to the present invention is to more reliably perform object detection and image content determination. Another object of the image processing apparatus, the printer including the image processing apparatus, the image processing method, and the image processing program according to the present invention is to more appropriately perform image correction processing according to the image content.

  The image processing apparatus, the printer including the same, the image processing method, and the image processing program of the present invention employ the following means in order to achieve at least a part of the above object.

The image processing apparatus of the present invention
An image processing apparatus that performs predetermined image correction processing on an image and outputs the image,
When an instruction to perform a predetermined image correction process is given to the designated image, an object region included in the image is extracted based on the pixel information of the image, and the designated image includes a plurality of degrees including 0 degrees. Image content determination means for determining the image content based on the arrangement of the extracted object region in each rotated image rotated at a rotation angle;
Image processing means for performing predetermined image correction processing on the designated image based on the determined image content, and outputting the processed image;
It is a summary to provide.

  In the image processing apparatus of the present invention, an object region is extracted based on the pixel information of the image, and the image content is determined based on the arrangement of the object region in each rotated image obtained by rotating the image at a plurality of rotation angles. Based on the determined image content, an image correction process is performed and output. Therefore, since the image content is determined based on the arrangement of the object areas by rotating the image at a plurality of rotation angles, the image content can be determined more reliably. As a result, the image correction process based on the image content can be performed more appropriately. Here, examples of the plurality of rotation angles include “0 degree, 90 degrees, 180 degrees, and 270 degrees”.

  In such an image processing apparatus of the present invention, the image content determination means extracts a person prediction area predicted to be a human face as one of the object areas, and arranges the extracted person prediction area in each rotated image. It is also possible to determine that the image content is a person image based on the above. In this way, the person image can be determined more reliably. In this case, the image content determination unit may be a unit that extracts a skin color region constituted by pixels of a predetermined skin color as the person prediction region.

  In this aspect of the image processing apparatus of the present invention, the image content determination means determines that the image content is a person image when the arrangement of the person prediction area in any of the rotated images satisfies a predetermined condition. It can also be a means to do. In this case, the image content determination means is a person image by using, as one of the predetermined conditions, that the amount of the person prediction area contacting the upper side and the left and right sides of each rotated image is equal to or less than a predetermined amount. Or a region aligned in the horizontal direction corresponding to both eyes of the person and a region corresponding to the mouth of the person located below the region in the person prediction region in each rotated image. It may be a means for determining that the image content is a person image by using existence as one of the predetermined conditions. Here, the “predetermined amount” includes the value 0. In other words, it includes one that determines that the content of the image is a person image using one of the predetermined conditions that the person prediction area does not touch the upper side and the left and right sides of the image.

  In the image processing apparatus of the present invention, the image content determination means may be means for determining that the image content is a landscape image when the image content is not determined to be a person image.

  Furthermore, in the image processing apparatus of the present invention, the image content determination unit extracts a sky blue area constituted by a predetermined sky blue pixel as one of the object areas, and One of the conditions is that the sky blue area is not in contact with the lower side of the image, and the image content may be determined as a landscape image with the sky blue area as the sky. In this way, it is possible to more reliably determine a landscape image having the sky.

  In such an image processing apparatus of the present invention, when the designated image is accompanied by shooting rotation information indicating a rotation angle at the time of shooting, the image content determining means rotates an image rotated at the rotation angle indicated by the shooting rotation information. Can be used as a means for preferentially using the image to determine the image content. In this way, since the image rotated at the rotation angle at the time of shooting is preferentially used, the image content can be determined more quickly. In this case, the image content determination unit determines the image content related to the person image using the rotated images, and determines the image content related to the landscape image using the image rotated at the rotation angle indicated by the shooting rotation information. It can also be assumed. In this way, since the determination regarding the landscape image is performed using the image rotated at the rotation angle at the time of shooting, the image content can be determined more efficiently. Here, the determination of the image content related to the person image is performed using each rotated image. In the case of a person image, it is assumed that the person is photographed in a lying state or standing upside down. This is based on the fact that it is considered that the determination of the image content cannot be performed reliably even if only the image rotated at the rotation angle is used.

  Further, in the image processing apparatus of the present invention, the image processing means, when it is determined that the image content is a person image, as a predetermined image correction process, a skin color correction process for correcting a skin color, a soft focus process, a lightness correction process Of the image, or when it is determined that the image content is a landscape image, a sky blue correction process for correcting a sky color and a new green color correction process for correcting a new green color as the predetermined image correction process. , And means for performing at least one of the saturation correction processing.

  The gist of the printer of the present invention is that it includes any one of the above-described image processing apparatuses of the present invention and a print execution unit that prints an image output from the image processing apparatus.

  Since the printer of the present invention includes the image processing apparatus of the present invention according to any one of the above-described aspects, the effect of the image processing apparatus of the present invention, for example, the effect that the image content can be more reliably determined, An effect that image correction processing based on image contents can be more appropriately performed can be achieved.

The image processing method of the present invention includes:
An image processing method for performing a predetermined image correction process on an image and outputting the image,
(A) When an instruction to perform a predetermined image correction process is given to a specified image, an object area included in the image is extracted based on pixel information of the image, and the specified image is set to 0 degrees. Determining the image content based on the arrangement of the extracted object regions in each rotated image rotated at a plurality of rotation angles including,
(B) A predetermined image correction process is performed on the designated image based on the determined image content and output.
This is the gist.

  In the image processing method of the present invention, an object region is extracted based on pixel information of an image, the image content is determined based on the arrangement of the object region in each rotated image obtained by rotating the image at a plurality of rotation angles, Based on the determined image content, an image correction process is performed and output. Therefore, since the image content is determined based on the arrangement of the object areas by rotating the image at a plurality of rotation angles, the image content can be determined more reliably. As a result, the image correction process based on the image content can be performed more appropriately. Here, examples of the plurality of rotation angles include “0 degree, 90 degrees, 180 degrees, and 270 degrees”.

The image processing program of the present invention includes:
An image processing program for performing predetermined image correction processing on an image and outputting the image,
When an instruction to perform a predetermined image correction process is given to the designated image, an object region included in the image is extracted based on the pixel information of the image, and the designated image includes a plurality of degrees including 0 degrees. An image content determination module that determines the image content based on the arrangement of the extracted object regions in each rotated image rotated at a rotation angle;
An image processing module that performs predetermined image correction processing on the specified image based on the determined image content, and outputs the image processing module;
It is a summary to provide.

  In the image processing program of the present invention, an object area is extracted based on pixel information of an image, and the image content is determined based on the arrangement of the object area in each rotated image obtained by rotating the image at a plurality of rotation angles. Based on the determined image content, an image correction process is performed and output. Therefore, since the image content is determined based on the arrangement of the object areas by rotating the image at a plurality of rotation angles, the image content can be determined more reliably. As a result, the image correction process based on the image content can be performed more appropriately. Here, examples of the plurality of rotation angles include “0 degree, 90 degrees, 180 degrees, and 270 degrees”.

  Next, the best mode for carrying out the present invention will be described using examples.

  FIG. 1 is a configuration diagram showing an outline of the configuration of a printer 20 equipped with an image processing apparatus as an embodiment of the present invention. As shown in the figure, the printer 20 of the embodiment includes a control unit 21 that controls the entire apparatus, an interface unit 22 that manages connection with a storage medium 30 such as a memory card, a digital still camera 31, a personal computer 32, and the like, and a storage unit. An image processing unit 23 that performs various types of image processing on an image file read from the medium 30 and the like, a data buffer 24 that temporarily stores data, and color conversion processing and binarization processing are performed on the read image file. A print data generation unit 25 that performs print data generation, an image buffer 26 that stores the generated print data, and a printer engine 27 that executes printing based on the print data stored in the image buffer 26. It is configured as an ink jet printer that performs printing by ejecting ink of each color onto paper.The image processing unit 23 also performs processing for restoring an image file stored in the storage medium 30 or the like and compressed by the JPEG method or the like. A JPEG image file is compressed through steps such as block division, discrete cosine transform (hereinafter abbreviated as DCT) operation, quantization processing, and Huffman coding. Such a JPEG image file is a general image file and does not form the core of the present invention.

  Next, the operation of the printer 20 configured as described above, particularly the operation when performing image correction processing on an image file read from the storage medium 30 and printing will be described. FIG. 2 is a flowchart illustrating an example of image printing processing executed by the control unit 21, the image processing unit 23, the print data generation unit 25, and the like when an image file printing instruction is issued. In the image printing process, first, as shown in the figure, the data of one access unit of the image file related to the print instruction is read from the storage medium 30 or the like (step S100), and the process of decompressing the read data is executed (step S110). ). Here, the access unit is a data reading unit predetermined according to the specifications of the storage medium 30 and the like, and corresponds to a sector in the flash memory, for example. In addition, as illustrated in FIG. 3, the data size of each block of the image file compressed by the JPEG method does not necessarily match the access unit, and the data size between the blocks is different. Across multiple access units.

  Then, the process of reading the data in the access unit and decompressing the Huffman is repeatedly executed until the data for one block is decompressed. When the data for one block is decompressed (step S120), the decompressed data (block Inverse quantization processing is performed on the (quantized DCT coefficients) (step S130). By executing the inverse quantization process, the DCT coefficients of the block are obtained. Here, in the JPEG image file, since the difference value between blocks is quantized for the quantized DC component of the quantized DCT coefficient, the difference value of the quantized DC component obtained by Huffman decompression is accumulated. As a result, a quantized DC component is acquired, and a DC component is acquired by applying an inverse quantization process to the quantized DC component.

  Subsequently, inverse DCT calculation processing is performed on the DCT coefficient acquired by the inverse quantization processing in step S130 to acquire pixel information of the block (step S140), and color conversion for converting the color space of the acquired pixel information Processing is executed (step S150). Here, the color conversion process is a process of converting the YCC color space used in the JPEG image file into the RGB color space.

  When the pixel information converted into the RGB color space is obtained in this way, this pixel information is stored in the predetermined area of the data buffer 24 as image data of the entire image (step S160), and a histogram indicating the distribution of RGB values of the entire image is obtained. Are accumulated and stored in a predetermined area of the data buffer 24 (step S170). Note that the pixels for which histograms are to be accumulated may be pixels sampled according to an appropriate rule.

  When the current block is not the last block of the image file, the process returns to step S100 (step S180), and the processes of steps S100 to S170 are repeatedly executed for the next block. Thus, by restoring the image file in order from the top, the image data of the entire image and the histogram of RGB values are acquired and stored in the data buffer 24.

  When the processes in steps S100 to S170 are repeatedly executed up to the last block of the image file in this way, next, an object recognition process for determining the image content based on the image data of the entire image is executed (step S190). Here, the description of the image printing process is interrupted, and the object recognition process illustrated in FIG. 4 will be described.

  In the object recognition process, first, as shown in FIG. 4, a skin color region including skin color pixels is extracted based on the image data of the entire image stored in the predetermined region of the data buffer 24 (step S400). The extraction of the skin color region is performed by previously determining a range of RGB values corresponding to the skin color and extracting pixels within the range of RGB values.

  Then, in order to determine whether the extracted skin color area is a person's face, whether or not the skin color area touches the upper side or the left and right sides of the image, and a pixel area corresponding to both eyes and mouth of the person is in the skin color area. It is determined whether or not it exists (steps S410 and S420). Here, in order to determine whether or not there is a pixel area corresponding to both eyes and mouth, a pixel area having a lower brightness than the surrounding pixels is extracted from the pixels in the skin color area, and the pixel area having a lower brightness is shown in FIG. As shown in FIG. 5, when they exist at the positions of the apexes of a substantially inverted triangle (that is, two positions aligned in the horizontal direction and one position below the two positions), these pixels correspond to both eyes and mouth. The area was determined.

  When the skin color area is not in contact with either the upper side or the left and right sides and there is a pixel area corresponding to both eyes and mouth, the skin color area is determined to be a person's face, and the image content is defined as a person image. A determination is made (step S430), and the object recognition process is terminated.

  On the other hand, when the skin color area is in contact with either the upper side or the left and right sides, or when there is no pixel area corresponding to both eyes or mouth, the image is rotated 90 degrees to the right, and the skin color area is a person for the rotated image. The process of determining whether or not the face is the same is repeated until the rotation angle in the right direction reaches 270 degrees (steps S440 and S450). FIG. 6 is an explanatory diagram showing an example of a state in which the process of determining whether the skin color area is a human face is repeatedly executed. In this example, as shown in the figure, since the extracted skin color area is in contact with the right side in the first image, it is not determined to be a human face, and when the image is rotated 90 degrees to the right, the skin color area is Since there is also a pixel area corresponding to both eyes and mouth that is not in contact with any of the sides, it is determined as a human face. In this way, the process of determining whether the skin color area is a human face is repeatedly executed while rotating by 90 degrees until the rotation angle of the image in the right direction reaches 270 degrees.

  If the flesh-color area is not determined to be a human face even if the rotation angle of the image in the right direction is 270 degrees, then a sky-blue area composed of sky-blue pixels is extracted based on the image data (step S460). . As in the case of extracting the skin color area, the sky blue area is extracted by setting a range of RGB values corresponding to the sky blue in advance and extracting pixels within the range of RGB values.

  Next, it is determined whether or not the extracted sky blue area is in contact with the lower side of the image (step S470). When the extracted sky blue area is not in contact with the lower side of the image, it is determined that the sky blue area is empty and the image content is set to empty. It is determined that the image has a landscape image (step S480), and the object recognition process is terminated.

  On the other hand, when the sky blue area is in contact with the lower side of the image, the image is rotated 90 degrees to the left, and processing for determining whether the sky blue area is in contact with the lower side of the image is performed on the rotated image. The process is repeatedly executed until the rotation angle in the right direction becomes 0 degree (steps S490 and S500). Here, when step S470 is executed for the first time, the rotation angle of the image in the right direction is 270 degrees. Therefore, by repeating the process of rotating 90 degrees in the left direction, the rotation angle in the right direction is changed. The images are sequentially rotated so as to be 270 degrees, 180 degrees, 90 degrees, and 0 degrees. FIG. 7 is an explanatory diagram illustrating an example of a state in which the process of determining whether or not the sky blue region is in contact with the lower side of the image is repeatedly executed. In this example, as shown in the figure, in the first image or an image that is once rotated 90 degrees to the left, the extracted sky blue area is in contact with the lower side, so it is not determined that it is empty. When the angle is rotated, the sky blue area is determined to be empty because it does not touch the lower side.

  If the sky blue area is in contact with the lower side even if the rotation angle of the image in the right direction is 0 degree, the image content is determined to be a landscape image having no sky (step S510), and this object recognition processing is performed. Exit. When the sky blue area is extracted in step S460, if there is no sky blue area in the image data, the image content is determined to be a landscape image having no sky.

  When the image content is determined by the object recognition processing in this way, the image printing processing is next performed for automatic correction processing for performing automatic correction processing on the image file based on the determined image content, a histogram of accumulated RGB values, and the like. The parameter is calculated and stored in a predetermined area of the data buffer 24 (step S200). The calculation of the parameters for automatic correction processing can be performed by applying various rules. In the embodiment, when the image content is a human image, the parameters are set so as to perform a skin color correction process and a soft focus process for correcting the skin color to be a better skin color, a brightness correction process for correcting the brightness to be slightly brighter, and the like. When the image content is a landscape image, a new green correction process that corrects the new green color to be a better new green color, a saturation correction process that corrects the saturation to be vivid, etc. The parameters were calculated. Further, when the image content is a landscape image having sky, the parameters are further calculated so as to perform sky blue correction processing for correcting the sky blue to be a better sky blue. Regardless of whether the image content is a person image or a landscape image, parameters are calculated such that various image correction processing (contrast correction processing, γ correction processing, etc.) is performed according to the RGB value histogram. It is.

  When the parameters for automatic correction processing are calculated in this way, the process returns to the top of the image file, reads data for one access unit, decompresses Huffman (steps S210 and S220), and decompresses data for one block (step S230). Inverse quantization processing, inverse DCT calculation processing, and color conversion processing are executed (steps S240 to S260).

  Subsequently, the block is subjected to an automatic correction process according to the calculated automatic correction process parameter (step S270) and output to the data buffer 24 (step S280). When performing the automatic correction process, specifically, a process of converting the pixel information is performed using a lookup table that is a conversion table of the pixel information generated based on the parameters for the automatic correction process.

  Then, the processing of steps S210 to S280 is repeatedly executed until the blocks output to the data buffer 24 are accumulated and become a band unit which is a unit for executing printing by the printer engine 27 (step S290). Becomes the band unit, the print data generation unit 25 generates print data based on the band unit data and outputs the print data to the image buffer 26 (see FIG. 8), and the printer engine 27 performs printing based on the print data. Execute (step S300). When the last block of the image file is reached (step S310), the image printing process is terminated.

  According to the printer 20 of the embodiment described above, when an instruction to print an image file is given, a skin color area and a sky blue area are extracted from the image data, and it is determined whether the skin color area is a person's face or a sky blue area. The image content can be determined by determining whether it is empty or not with respect to an image rotated by 90 degrees. That is, since the image content is determined by rotating the image at a plurality of rotation angles, the image content can be determined more reliably. As a result, the image correction process based on the image content can be performed more appropriately.

  Here, in the printer 20 according to the embodiment, the control unit 21 and the image processing unit 23 that execute the object recognition process correspond to an image content determination unit, and the control unit 21 and the image processing unit that execute the processes in steps S200 to S290. 23 corresponds to an image processing unit, and the control unit 21, the print data generation unit 25, and the printer engine 27 that execute the process of step S 300 correspond to a print execution unit. The skin color area and sky blue area correspond to the object area.

  In the printer 20 of the embodiment, the skin color area composed of skin color pixels is extracted to determine whether the face is a person's face. However, it is sufficient that an area predicted to be a person's face can be extracted. It is good also as what is extracted on the basis. For example, it is good also as what judges the face of a person by extracting the area | region which consists of an edge pixel whose edge degree based on a luminance difference with an adjacent pixel is larger than a predetermined threshold value.

  In the printer 20 of the embodiment, the image content is determined to be a person image on the condition that the skin color area is not in contact with either the upper side or the left and right sides of the image. It does not matter.

  In the printer 20 of the embodiment, the image content is determined based on whether the skin color area is in contact with the upper side or the left and right sides of the image, and whether the pixel area corresponding to both eyes and mouth of the person exists in the skin color area. Of course, it is determined that the image is an image. However, the present invention is not limited to this, and when one of these conditions is satisfied, it may be determined that the image is a person image, or another condition may be used. I do not care.

  In the printer 20 of the embodiment, it is determined whether or not the sky blue area is in contact with the lower side of the image. However, the determination of such a sky blue area may be omitted. In this case, when the image content is not determined to be a person image, the image content may be determined to be a landscape image. Furthermore, it is possible to determine image contents other than a person image and a landscape image.

  In the printer 20 of the embodiment, the determination as to whether the skin color area is a human face or the determination as to whether the sky blue area is empty is performed on an image rotated by 90 degrees. Needless to say, it is not limited to degrees, and for example, it may be rotated in units of 30 degrees or 45 degrees. Further, an image rotated at a plurality of rotation angles including 0 degrees may be targeted, and an image rotated at other rotation angles may be targeted.

  In the printer 20 according to the embodiment, it is determined whether the skin color area is a human face while rotating 90 degrees at a time until the rotation angle of the image in the right direction reaches 270 degrees, and then the rotation angle in the right direction is 0. It is determined whether or not the sky blue area is empty while rotating 90 degrees to the left until it reaches the degree. However, if the image file includes shooting rotation information indicating the rotation angle at the time of shooting, this shooting is performed. The rotation angle of the image may be determined according to the rotation information. In this case, for example, the object recognition process illustrated in FIG. 9 may be executed instead of FIG. That is, as shown in FIG. 9, the rotation angle at the time of shooting is acquired based on the shooting rotation information of the image file (step S400a), and the image is rotated in advance so that the acquired rotation angle at the time of shooting is obtained. (Step S400b) The process of determining whether the skin color area is a human face while rotating by 90 degrees until the subsequent rotation angle of the image to the right reaches 270 degrees is repeatedly executed (Steps S400 to S450). . Then, when determining whether or not the sky blue area is empty, it is determined whether or not the sky blue area is in contact with the lower side of the image using the image rotated so as to have the rotation angle at the time of shooting (step S460a). , S460 to S510). In this way, the image rotated to the rotation angle at the time of shooting is preferentially used to determine whether the skin color area is a human face, and only the image rotated to the rotation angle at the time of shooting is used. Therefore, since it is determined whether the sky blue area is empty, the image content can be determined more efficiently. Here, when determining whether the skin color region is a person's face, an image that is rotated by 90 degrees from the rotation angle at the time of shooting is also targeted. In the case of a person image, the person lies down or stands upright. This is based on the fact that it is considered that reliable image content cannot be determined using only the image rotated at the rotation angle at the time of shooting.

  In the embodiment, the present invention for determining the image content for an image rotated at a plurality of rotation angles has been described as an embodiment of the printer 20, but an image processing apparatus that performs image correction processing by determining such image content. It is good also as a form. Further, it can be in the form of such an image processing method or an image processing program.

  The best mode for carrying out the present invention has been described with reference to the embodiments. However, the present invention is not limited to these embodiments, and various modifications can be made without departing from the gist of the present invention. Of course, it can be implemented in the form.

1 is a configuration diagram showing an outline of the configuration of a printer 20 equipped with an image processing apparatus of the present invention. 6 is a flowchart illustrating an example of image printing processing. Explanatory drawing which shows the relationship between the block of an image file, and an access unit. The flowchart which shows an example of an object recognition process. Explanatory drawing which shows an example of the pixel area | region corresponded to both eyes and an opening | mouth. Explanatory drawing which shows an example of a mode which judges whether a skin color area | region is a person's face. Explanatory drawing which shows an example of a mode that it is judged whether a sky blue area | region is empty. Explanatory drawing which shows a mode that the accumulate | stored block becomes a band unit. The flowchart which shows an example of the object recognition process of a modification.

Explanation of symbols

  20 printer, 21 control unit, 22 interface unit, 23 image processing unit, 24 data buffer, 25 print data generation unit, 26 image buffer, 27 printer engine, 30 storage medium, 31 digital still camera, 32 personal computer.

Claims (4)

An image processing apparatus that performs predetermined image correction processing on an image and outputs the image,
When an instruction to perform a predetermined image correction process is given to the designated image, an object region included in the image is extracted based on the pixel information of the image, and the designated image includes a plurality of degrees including 0 degrees. Image content determination means for determining the image content based on the arrangement of the extracted object region in each rotated image rotated at a rotation angle;
Image processing means for performing predetermined image correction processing on the designated image based on the determined image content, and outputting the processed image;
Equipped with a,
The image content determination means extracts a person prediction area predicted as a human face as one of the object areas, and the arrangement of the extracted person prediction areas in any one of the rotated images satisfies a predetermined condition. One of the predetermined conditions is a means for determining that the content of the image is a person image when satisfying, and that the amount of the person prediction area contacting the upper side and the left and right sides of each rotated image is equal to or less than a predetermined amount. An image processing apparatus as means for determining that the image content is a person image . An image processing apparatus according to claim 1 Symbol placement,
Print execution means for printing an image output from the image processing apparatus;
Printer with. An image processing method for performing a predetermined image correction process on an image and outputting the image,
(A) When an instruction to perform a predetermined image correction process is given to a specified image, an object area included in the image is extracted based on pixel information of the image, and the specified image is set to 0 degrees. Determining the image content based on the arrangement of the extracted object regions in each rotated image rotated at a plurality of rotation angles including,
(B) the output to the designated image by performing predetermined image correction processing based on the determination image content,
The step (a) extracts a person prediction area predicted as a person's face as one of the object areas, and the arrangement of the extracted person prediction areas in any one of the rotated images satisfies a predetermined condition. One of the predetermined conditions is a step of determining that the image content is a person image when satisfying, and that the amount of the person prediction area contacting the upper side and the left and right sides of each rotation image is equal to or less than a predetermined amount. An image processing method that is a step of determining that the image content is a person image . An image processing program for performing predetermined image correction processing on an image and outputting the image,
When an instruction to perform a predetermined image correction process is given to the designated image, an object region included in the image is extracted based on the pixel information of the image, and the designated image includes a plurality of degrees including 0 degrees. An image content determination module that determines the image content based on the arrangement of the extracted object regions in each rotated image rotated at a rotation angle;
An image processing module that performs predetermined image correction processing on the specified image based on the determined image content, and outputs the image processing module;
Equipped with a,
The image content determination module extracts a person prediction area predicted as a person's face as one of the object areas, and the arrangement of the extracted person prediction area in any one of the rotated images satisfies a predetermined condition. One of the predetermined conditions is a module for determining that the image content is a person image when satisfying, and that the amount of the person prediction area contacting the upper side and the left and right sides of each rotation image is equal to or less than a predetermined amount. An image processing program which is a module used to determine that an image content is a person image .

Images