JP2005260597A - Digital camera and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To save on work of a user by accurately determining image quality in document electronization using a digital camera. <P>SOLUTION: A control part 110 of a digital camera 100 segments a character part to specify a character size and specifies an edge part of the character part. The control part 110 calculates a rate of gradational change in the edge part and specifies the readable level of the character on the basis of the calculated rate of gradational change and the character size. Further, the control part 110 determines whether the document is illegible according to a relation of the specified readable level to a prescribed threshold. The control part 110 controls an outputting part 140 to output prescribed notification information corresponding to a determination result, notifying the user of the notification information. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a digital camera, and more particularly to a digital camera suitable for digitizing a document.

  Conventionally, in order to digitize documents on paper etc., it was common to read them with a flatbed scanner, etc., but in recent years, high-pixel digital cameras have become available at low prices. Digital cameras are being used instead of scanners and memos. Taking advantage of the features of a digital camera that can record the captured image as digital data, document information can be obtained by shooting a document printed on paper or characters on a signboard in the city with a digital camera. Can be easily digitized. In addition, unlike a flatbed scanner or the like, a digital camera is excellent in portability, so that it is possible to digitize a document anytime and anywhere regardless of the form of the target document. Further, since it can be recorded as digital data, it can be easily converted into convenient text data by a predetermined character recognition process or the like.

  However, unlike flatbed scanners where the distance to the target is always constant, shooting with digital cameras has different shooting conditions for each shooting, so it is always possible to obtain the image quality required for character recognition processing. Have difficulty. In particular, when shooting a document printed on paper, macro shooting is often performed, and camera shake and blurring tend to occur. An image in which camera shake, blurring, or the like occurs has low image quality, resulting in a decrease in recognition accuracy of the character recognition process.

  Therefore, when document information is recorded using a digital camera, conventionally, (1) a plurality of images of the same object to be imaged are selected and the best image quality is selected after shooting, or (2) each image is captured. In other cases, the image quality was confirmed on the screen, and if the image quality was poor, it was retaken. In the method (1), even image data with low image quality must be recorded, so that the memory area is occupied unnecessarily, which may hinder the recording of a really necessary image. In addition, a screen (such as a liquid crystal display device) provided in a digital camera often has a low resolution, and an enlarged display or scrolling must be performed in order to check the image quality. Therefore, in the method (2), the work time for confirmation becomes long and the photographing efficiency is lowered. In addition, since both methods are operations performed by the user (photographer), the user's labor is required and is a burden.

Techniques (for example, Patent Documents 1 and 2) for automatically determining whether or not a predetermined image quality is obtained by image processing have also been proposed, but in the conventional method, the determination criterion is only “character size”. Or, only “image blurring” is available. In such a method, for example, if the characters are sufficiently large even though the image is blurred, it may be determined that the image quality is low and not recorded. That is, when the character is sufficiently large, it is determined that “image quality is low” even though there is a case where the image quality is sufficient for character recognition even if there is a blur. On the other hand, if the characters are small even if there is no image blurring, the image quality required for character recognition may not be achieved, but based on the blurring only, the image quality is judged to be high and an image with low image quality is recorded. There is also a case. That is, depending on the prior art, the image quality cannot be accurately determined. As a result, even though a useful image can be captured, it is forced to re-shoot, or an image with low image quality is recorded. Inconveniences such as endlessness occurred, and it was not efficient.
JP-A-10-247220 Japanese Patent Laid-Open No. 08-063547

  The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a digital camera and a program capable of accurately determining and notifying image quality when photographing a document.

In order to achieve the above object, a digital camera according to the present invention provides:
In a digital camera that acquires image data by converting incident light into digital data,
Feature detection means for detecting the features of the image data;
Determination means for determining the readability level of the character indicated in the image data based on the feature detected by the feature detection means;
Execution means for executing a predetermined operation based on the readability level determined by the determination means,
The feature detection means further comprises a gradation detection means for detecting a gradation change rate in the image data,
The determination means determines the readability level of the character based on the gradation change rate detected by the gradation detection means,
The execution unit further includes an output unit that outputs predetermined notification information based on the readability level determined by the determination unit.
It is characterized by that.

The above digital camera
It is desirable to further comprise an edge detection means for detecting the edge of the character shown in the image data, in this case,
The gradation detecting means preferably detects the gradation change rate of the edge portion detected by the edge detecting means.

In the above digital camera,
The feature detection unit preferably further includes a character size specifying unit that specifies the size of the character. In this case,
The determination unit preferably determines the readability level of the character based on the gradation change rate detected by the gradation detection unit and the character size specified by the character size specification unit.

The above digital camera
It is desirable to further include a determination unit that determines whether or not the character can be read based on a predetermined threshold and the readability level determined by the determination unit.
Preferably, the output means outputs notification information indicating that when it is determined by the determination means that reading is impossible.

In the above digital camera,
The determination means determines a readability level at a plurality of locations on the image data,
The determination means determines whether or not the plurality of places can be read,
The output means may output the notification information when it is determined that reading is impossible at any of the plurality of locations.

The above digital camera
It is desirable to further comprise storage means for storing determination information in which the information indicating the feature detected by the feature detection means is associated with the readability level.
It is desirable that the determination unit determines a readability level based on determination information stored in the storage unit.

The above digital camera
It is desirable to further include an input means for inputting an instruction from the user.
The storage unit may update the determination information in accordance with a user instruction input to the input unit.

Also,
The character size specifying unit may specify a character size in accordance with a user instruction input to the input unit.

In the above digital camera,
The gradation detection means may further comprise gradation distribution detection means for detecting a distribution of gradation change rate in the image data.
The determination unit may determine the readability level of the character based on a distribution of gradation change rates detected by the gradation distribution detection unit.

in this case,
It is preferable that the determination unit determines the readability level based on the distribution of the gradation change rate, with the gradation change rate having the highest appearance frequency as the gradation change rate in the character.

The above digital camera
It is desirable to further comprise a character area extraction means for extracting the character area shown in the image data, in this case,
It is desirable that the feature detection means detects a feature in the character area extracted by the character area extraction means.

In order to achieve the above object, a program according to the second aspect of the present invention is:
To a computer that controls a digital camera that converts incident light into digital data and acquires image data,
Detecting characters shown in the acquired image data;
Detecting a gradation change rate in at least the detected character;
Identifying the size of the detected characters;
Determining a readability level of the character based on the detected gradation change rate and / or the specified character size;
Outputting predetermined notification information according to the determined readability level;
Is executed.

  According to the present invention, since the readability of the character shown in the image is determined and notified based on the gradation change rate in the captured image, the image quality can be accurately determined and notified. As a result, it is possible to save labor at the time of shooting and improve shooting efficiency, and to prevent the memory area from being wasted.

  Embodiments according to the present invention will be described below with reference to the drawings. In this embodiment, when a character or document is photographed by a digital camera (imaging device) that acquires image data using an imaging device such as a CCD (Charge Coupled Device) (hereinafter referred to as “document photographing”). ) As an example. In this embodiment, “document shooting” refers to, for example, shooting characters written on paper or a signboard with a digital camera and recording character information as a substitute for taking in by a scanner or a memo. To do. In addition, an image acquired for the purpose of taking a document is referred to as a “document image”. Such a document image is converted into text data by being subjected to a predetermined character recognition process. That is, document information is digitized by photographing a document with a digital camera.

  The configuration of the digital camera according to this embodiment will be described with reference to FIGS. FIG. 1 is a block diagram showing a system configuration (internal configuration) of a digital camera 100 according to the present embodiment, FIG. 2 is a diagram showing an example of the appearance of the digital camera 100 (FIG. 2A is a front view, FIG. b) is a rear view). As illustrated, the digital camera 100 according to the present embodiment includes a control unit 110, an imaging unit 120, an input unit 130, an output unit 140, an image recording unit 150, and a storage unit 160.

  The control unit 110 includes, for example, a CPU (Central Processing Unit) and controls each unit of the digital camera 100. Here, the control part 110 implement | achieves each process mentioned later by running the operation | movement program stored in the memory | storage part 160. FIG. Note that the control unit 110 includes a storage area for developing (loading) data and operation programs necessary for operation. This storage area (hereinafter referred to as “work area”) is configured by, for example, a register, a cache memory, and a RAM (Random Access Memory).

  The imaging unit 120 performs an imaging operation under the control of the control unit 110. In the present embodiment, as illustrated in FIG. 2, the imaging unit 120 includes a lens unit 121, a distance measurement / photometry unit 122, an imaging element 123, and the like.

  The lens unit 121 includes an optical member including a lens group and diaphragm blades, and a drive unit that drives the optical member. The drive unit drives the optical member under the control of the control unit 110. For example, the diaphragm blades are driven to achieve the set diaphragm. In addition, when the lens unit 121 has an AF (Auto Focus) function, the optical member is driven to focus on an object to be photographed. The optical member is driven accordingly.

  The distance measurement / photometry unit 122 includes, for example, an infrared irradiation unit, a distance sensor, and an optical sensor, and measures a distance from the digital camera 100 to a shooting target (subject), a light amount in a shooting space, and the like.

  The image pickup device 123 is configured by, for example, a CCD, and converts visible light (incident light) incident through the lens unit 121 into electric charge by photoelectric conversion and accumulates it. Here, the imaging device 123 is provided with a predetermined AD conversion circuit, and converts it into digital data based on the amount of the converted charge. That is, the converted digital data constitutes “image data” obtained from incident light.

  The input unit 130 includes buttons that accept instructions from the user and is operated by the user. The input unit 130 inputs a user instruction to the control unit 110 by sending a predetermined signal corresponding to the operation to the control unit 110. In the present embodiment, the input unit 130 includes at least a shutter button 131 and an operation input unit 132.

  The shutter button 131 includes a button that can be moved up and down in response to a press by the user, and a signal that instructs the start of shooting (hereinafter referred to as a “shutter signal”) when pressed by the user. 110. Note that the shutter button 131 sends different signals to the control unit 110 depending on two states, a so-called “half-pressed” state and a “full-pressed” state. Here, “half-pressed” refers to a state where the shutter button 131 is pressed halfway, and “full-pressed” refers to a state where the shutter button 131 is fully pressed.

  As with a normal camera, the digital camera 100 according to this embodiment performs distance measurement / photometry and focusing (focusing) (hereinafter referred to as “imaging preparation operation”) by “half-pressing”. “Press” to capture and record (hereinafter referred to as “imaging operation”). Therefore, when the shutter button 131 is half-pressed, a shutter signal indicating that effect (hereinafter referred to as “half-press signal”) is sent to the control unit 110, and the control unit 110 starts the focusing operation accordingly. And so on. When the shutter button 131 is fully pressed, a shutter signal indicating that effect (hereinafter referred to as a “full press signal”) is sent to the control unit 110, and the control unit 110 executes the start of an imaging / recording operation. To do. That is, the control unit 110 starts each operation for imaging upon reception of a signal transmitted from the shutter button 131.

  The operation input unit 132 includes, for example, predetermined operation buttons, dials, cursor keys, and the like, and instructions for changing various settings are input by user operations. In the present embodiment, it is used for changing / setting various modes including a photographing mode, and for setting a readability level (details will be described later) as a reference for determining whether or not characters can be read in a document image. In this embodiment, it is assumed that “document shooting mode” is prepared as the shooting mode. When the digital camera 100 performs shooting of a document, the operation input unit 132 is operated to set the document shooting mode.

  The output unit 140 outputs various types of information notified to the user. In the present embodiment, the display unit 141 that outputs an image, the sound report unit 142 that outputs sound, lighting, flashing, and the like. The indicator part 143 which performs this light emission is provided. The display unit 141, the sound report unit 142, and the indicator unit 143 are configured on the back side of the digital camera 100 as shown in FIG. In other words, it is desirable that the information output at the time of photographing using the digital camera 100 be configured to be transmitted to the photographer (user).

  The display unit 141 is composed of, for example, a liquid crystal display device, and displays and outputs image information. In the present embodiment, in addition to displaying an image obtained by the image sensor 123 and an image recorded in the image recording unit 150, a menu screen for changing the setting, and information indicating a setting state and a warning to the user (for example, , Characters and icons) are displayed. In the present embodiment, an image or character information indicating a warning or guidance to the user according to the readability level is displayed. Note that when the display unit 141 includes a touch panel or the like, the display unit 141 may function as an input device equivalent to the operation input unit 132.

  The sound report unit 142 includes, for example, a speaker and outputs predetermined sound information. Here, sound (alarm sound, etc.) corresponding to various operations (for example, focusing, shutter pressing) at the time of shooting by the digital camera 100 is output, and in this embodiment, a warning to the user according to the readability level is provided. And voices indicating guidance are output.

  The indicator unit 143 includes, for example, an LED (Light Emitting Diode) and emits light with a predetermined color or blinking pattern according to various operations (for example, focusing, shutter pressing) at the time of shooting with the digital camera 100. In addition, in the present embodiment, light is emitted also at the time of warning to the user according to the readability level.

  Note that the output of the warning or guidance according to the readability level may be performed by all of the display unit 141, the sound report unit 142, and the indicator unit 143, or may be performed by any of these. It is assumed that which output device is used can be arbitrarily selected and set by the user.

  The image recording unit 150 is configured by, for example, a memory card including a storage device such as a flash memory, and is obtained by an imaging operation of the imaging unit 120 by recording digital data converted by the AD conversion circuit of the imaging element 123. The recorded image data is recorded.

The storage unit 160 includes, for example, a storage device such as a flash memory, and stores a program executed by the control unit 110, data necessary for each process (hereinafter referred to as “process data”), and the like. Here, in addition to a program for executing various controls for an imaging preparation operation and an imaging operation (hereinafter referred to as an “imaging program”), in this embodiment, features based on image data obtained by the imaging element 123 are included. A program for detecting and determining the readability of the characters shown in the image data is stored. More specifically, the following program is stored.
(1) “Character detection program”: a program for determining whether or not a character is shown in image data and extracting a character area (2) “Edge detection program”: a program for detecting an edge of a character indicated in image data ( 3) “Character size identification program”: a program that identifies the size of the character indicated in the image data (4) “Gradation change rate detection program”: a program that detects the gradation change rate in the image data (5) “Readability level determination program”: a program for determining a character readability level (details will be described later) based on the character size and gradation change rate indicated in the image data (6) “readability determination program”: determined (7) “Notification information output program for determining whether or not the characters shown in the image data are legible based on the readability level : Program that outputs a predetermined notification information based on the readability levels determined

When the control unit 110 executes such a program, the control unit 110 realizes the following functions.
(1) “Imaging function”: a function for operating the imaging unit 120 and acquiring image data based on incident light incident from the lens unit 121. (2) “Character detecting function”: a character indicated in the acquired image data. Function for detecting part (3) “Feature detection function”: Function for detecting feature of acquired image data (4) “Readability level determination function”: indicated in image data based on detected feature of image data (5) “Readability determination function”: a function for determining whether or not the characters shown in the image data can be read based on the determined readability level (6) ) "Output function": A function to output notification information to the user according to the determined readability level

  The “feature detection function” detects a gradation change rate on image data as a “feature” of the image data. In addition, by the “output function”, predetermined “notification information” (sound, image, etc.) corresponding to the readability level of the character that the user is about to capture is output from the output unit 140.

  In the present embodiment, each function is realized by software processing by the control unit 110 executing a program. For example, a circuit that specially processes each function (so-called ASIC (Application Specific Integrated)). By configuring the circuit)) in the digital camera 100, the above functions may be realized by hardware processing.

  In the storage unit 160, information indicating various settings necessary for shooting such as a combination of an appropriate exposure value (aperture value), a shutter speed, and the like (hereinafter referred to as “shooting parameter”) as “process data”. In this embodiment, a “readability level determination table” for determining the readability level (image quality) of characters indicated in the captured image data is stored. An example of this “readability level determination table” is shown in FIG.

  As shown in the drawing, in the “readability level determination table”, “readability level” corresponding to “character size” and “gradation change rate” is recorded in association with each other. Here, the “character size” indicates the size of the character indicated by the image data obtained by the image sensor 123, and is defined based on the number of vertical and horizontal dots of the character. That is, since the image data obtained by the image sensor 123 is a raster image composed of a plurality of pixels (dots), the character size is defined by the number of dots constituting the character shown in the image. In the present embodiment, the number of horizontal dots of each character is used.

  Further, depending on the range of the number of dots indicating the character size, the character size is classified into five levels of “A”, “B”, “C”, “D”, and “E” (hereinafter referred to as “character size level”). Divide into In the present embodiment, the case where the dot number range is “˜8” is referred to as “level A”, and “level B” = “9-16”, “level C” = “17-32”, “level D” hereinafter. = “33 to 64” and “Level E” = “65 to”. In other words, the size relationship between the character sizes is expressed by “A <B <C <D <E”. Note that the setting of the number of divisions and the range of the number of dots is arbitrary and may be set by the user. Further, any unit may be adopted as long as the character size can be defined. For example, the number of vertical dots of each character may be used, or the sum or product of the number of horizontal dots and the number of vertical dots may be used.

  The “gradation change rate” is an index for determining the degree of blurring of an image, and is represented by “gradation change amount / number of dots”. In the present embodiment, the digital camera 100 shoots a document that captures characters, but if the image is out of focus (so-called “out-of-focus”) or camera shake occurs, the captured image becomes unclear and the character portion Is difficult to interpret. In other words, in the case of out-of-focus or camera shake, the image may be blurred or blurred. At this time, so-called “blurred feet” occur at the edge of the character on the image (the boundary between the character and the background). To do. In the present embodiment, “gradation change rate” is used as an index indicating the degree of “blurred foot” (image quality).

  For example, when a character (“C”) expressed in black on a white background as shown in FIG. 4A is displayed as an image, the edge portion of the character (rectangular region in the figure) without any blur or camera shake. When (RA) is enlarged, as shown in FIG. 4 (b), the edge portion is composed of “black” pixels (dots), and no blurring occurs. On the other hand, when out-of-focus or camera shake occurs, as shown in FIG. 4C, an edge portion is formed by pixels of intermediate gradation between the white background and the black character. This halftone pixel becomes blurred and becomes a character with a blurred outline, resulting in poor legibility. And the wider the blur foot, the lower the legibility.

  The gradation change in such a case will be described with reference to FIG. Here, the one-dimensional pixel when the character “C” as described above is displayed as an image with 256 gradations (the gradation of “black” is “0” and the gradation of “white” is “255”). A change in gradation is observed (for example, a pixel on “line segment XX ′” in FIG. 5A).

  First, FIG. 5B shows a gradation change in a state where there is no blurring or camera shake. Here, the horizontal axis represents dots (pixels) on the line segment X-X ′, and the vertical axis represents the gradation of each dot (the same applies to FIG. 5C). When no blurring occurs, the gradation change at the edge portion changes abruptly from “255” in the background to “0” in the character portion. That is, at the edge portion of the character, when one dot is shifted from the background pixel in the direction of the character portion, 256 gradation changes (gradation change amount). The gradation change rate in this case is “256 (tone change amount) / 1 (number of dots) = 256”. That is, the gradation change rate in a state where there is no out-of-focus or camera shake coincides with the gradation difference between the background and the character portion.

  On the other hand, when the blur is generated at the edge portion, as shown in FIG. 5C, the gradation gradually changes from white in the background to black in the character. In other words, there is an intermediate gradation pixel between the white background and the black character, resulting in blurring (the hatched portion in FIG. 5C). Here, the number of pixels (number of dots) constituting the intermediate gradation is the “blurred foot width”. For example, the gradation change rate when the number of dots of the intermediate gradation pixel is “8” is “256/8 = 32”. Further, the gradation change rate when the number of dots is “32” is “256/32 = 8”.

  That is, the “gradation change rate” in the present embodiment is a numerical value that decreases with an increase in the width of the blurred foot, with the gradation difference between the background and the character portion being the maximum value. Therefore, it is possible to determine the degree of blur of the image by relatively viewing the magnitude of the gradation change rate. That is, the smaller the gradation change rate value, the greater the degree of image blur. As described above, since the width of the blur foot affects the quality of the legibility of the characters, the degree of blur of the image can be seen to some extent only from the number of dots of the blur foot. However, there are cases where the gradation difference between the background and the character portion is small (for example, when the background or characters are in color) depending on the object to be photographed. There may be a case where a small part is mixed (for example, a case where a background or a character part is composed of a plurality of colors). Even in such a case, the gradation change rate derived from the relationship between the number of dots on the blurred foot and the gradation change amount (gradation difference) is used in order to appropriately determine the degree of blur of the entire image.

  In the “readability level determination table” shown in FIG. 3A, the gradation change rates are set to “A ′”, “B ′”, “C ′”, “D” according to the range of numerical values indicating the gradation change rate. It is divided into five levels of “′” and “E” (hereinafter referred to as “gradation change rate level”). The gradation change rate level is set for each predetermined numerical range of the gradation change rate. In the present embodiment, the case where the numerical value range is “˜5” is “level A ′”, and hereinafter, “level B ′” = “6-10”, “level C ′” = “11-20”, “level” D ′ ”=“ 21-40 ”and“ level E ′ ”=“ 41- ”. That is, the level change of the gradation change rate is represented by each level, “A ′ <B ′ <C ′ <D ′ <E ′”. Note that since the “gradation change rate” is based on the number of dots on the image, the performance of the digital camera 100 (for example, the resolution, the number of pixels, etc., hereinafter referred to as “image performance”) is affected. Therefore, a numerical value range that defines each stage is set in the “readability level determination table” according to the image performance of the digital camera 100 and the like. However, the setting of the number of sections and the numerical value range is arbitrary and may be set by the user.

  In the “readability level determination table”, the “readability level” associated with the character size and the gradation change rate is recorded (FIG. 3A). This “readability level” is information indicating the legibility of the characters shown in the document image data acquired by document shooting by the digital camera 100. In the present embodiment, the level is indicated by numerical values “1” to “9”, and the lower the numerical value, the lower the readability, and the higher the numerical value, the higher the readability. Note that “reading” in the present embodiment means that a user (human) recognizes a character by visual observation when the acquired document image is displayed, and the computer recognizes the character in a predetermined character recognition process. It means to do.

  Here, the readability of characters is classified into levels according to the size of characters and the degree of image blur. That is, the smaller the character size and the greater the degree of blur (when the gradation change rate value is smaller), the lower the readability. Therefore, the character size level including the smallest character size is “character size level: A”. The tone change rate level including the smallest tone change rate is expressed as “tone change rate level: A ′” (hereinafter referred to as “AA ′ level”. In this case, the hyphen The readability level corresponding to the character size level before and the tone change rate level after the hyphen is set to “1” which is the lowest level. On the other hand, as the character size is larger and the degree of blur is smaller (when the gradation change rate value is larger), the readability is higher. Therefore, “EE” includes the largest character size and the largest gradation change rate. The readability level corresponding to “level” is set to “9” which is the highest level.

  Here, a predetermined threshold is set for the readability level. In the present embodiment, “5” that is an intermediate value between “1” to “9” is set as a threshold value. This threshold value is used as a reference for determining whether or not the characters shown in the document image are legible. In this embodiment, when the numerical value of the readability level is less than the threshold value, it is not readable, and when it is equal to or higher than the threshold value, it is readable. That is, the readability level “5” indicates the minimum level at which characters can be read. Information indicating which readability level is set as such a threshold is stored in the storage unit 160 together with the “readability level determination table”.

  For example, the readability level “5” serving as the threshold is set to “C-C ′ level” in which both the character size level and the gradation change rate level are medium. Then, with the “CC ′ level” as a base point, the readability level is decremented by “1” every time the gradation change rate level is lowered by one level, and the readability level is incremented by “1” every time the gray level change rate level is raised by “C−”. Readability levels corresponding to “A ′ level”, “CB ′ level”, “CC ′ level”, “CD ′ level”, and “CE ′ level” are set. The readability level set in this way is, for example, “CA ′ level: 3”, “CB ′ level: 4”, “CC ′ level: 5”, “CD ′ level: 6”. "CE 'level: 7" (FIG. 3A).

  Then, with each of the readability levels set in this way as a base point, every time the character size level decreases by one step, the readability level is decremented by “1”, and every time the character size level increases by one, the readability level is incremented by “1”. Is set (FIG. 3A). With this setting, even if the character size is small, it may be legible if the degree of blur is small, and even if the degree of blur is large, it may be legible if the text size is large. . That is, the legibility is determined by the relative evaluation of the character size and the degree of blur. Therefore, it is possible to determine whether or not reading can be performed more accurately and flexibly than when determining whether or not reading is possible based on only the character size or only the degree of blur.

  Such a setting method is an example, and the readability level can be set by any method. Further, the readability level set by the method as described above may be a default value and may be changed as appropriate by the user. In this case, for example, it may be changed according to the user's visual acuity or preference, or may be changed according to the use of the document image. For example, when characters in a document image are extracted as text data by a predetermined character recognition program or the like, the readability level may be set according to the recognition accuracy of the character recognition program. It is assumed that the user's readability level is set by operating the operation input unit 132 of the digital camera 100 or the like. Further, in the case of setting according to the character recognition program as described above, the readability level may be automatically set by cooperation (cooperation) between the digital camera 100 and the program. .

  The storage unit 160 stores notification information output from the output unit 140. In this embodiment, notification information according to the readability level determined by the process described later and the legibility is output. Therefore, in the present embodiment, for example, there is a “notification information table” as shown in FIG. 3B in which information indicating the determination result of readability or information indicating the readability level is associated with the output notification information. It is stored in the storage unit 160. In “notification information”, text data indicating a predetermined message is recorded, and image information such as an icon is recorded as necessary.

  In the “notification information table”, information indicating which notification method is set is also recorded. Here, the “notification method” is an output form of notification information, and is prepared according to the configuration of the output unit 140 of the digital camera 100. In this embodiment, for example, “message display”, “icon display”, “indicator emission”, “voice notification”, “alarm notification”, and the like are prepared. These notification methods are arbitrarily selected by the user, and the notification method desired by the user is selected and set by the operation of the operation input unit 132 or the like. Details of each notification method will be described later.

  In addition to the above-described configuration, the digital camera 100 is provided with a configuration and functions necessary as a digital camera and other additional configurations and functions as necessary.

  Next, the operation of the digital camera 100 configured as described above will be described below with reference to the drawings. Each process to be described later is realized by the control unit 110 executing an operation program stored in the storage unit 160.

  The “document photographing process” according to the embodiment will be described with reference to the flowchart shown in FIG. Here, characters (hereinafter referred to as “document”) written on a sheet of paper or a signboard are photographed by the digital camera 100 and stored as image data indicating the contents of the document. For example, the stored image data is processed by a predetermined character recognition program to be converted into text data or the like and extracted, thereby being used as a substitute for capturing with a conventional memo or scanner. Shooting a document for such a purpose is referred to as “document shooting”, and an image obtained by document shooting is referred to as a “document image”. This “document shooting process” is started when the control unit 110 receives a shutter signal from the shutter button 131 in a state where the “document shooting mode” is selected as the shooting mode.

  First, a half-press signal is sent from the shutter button 131 to the control unit 110. The control unit 110 starts a shooting preparation operation when receiving a half-press signal from the shutter button 131 (step S101). Here, an appropriate shutter speed and aperture are determined based on the distance measurement / photometry operation by the distance measurement / photometry unit 122, and the lens unit 121 is driven to focus (focus) on the document. . It is assumed that the half-press signal continues to be sent from the shutter button 131 to the control unit 110 while the user is pressing the shutter button 131 halfway.

  When the document is focused by the shooting preparation operation, the image pickup device 123 receives incident light that enters through the lens unit 121, and accumulates (charges) the charge of the received device according to the determined shutter speed or the like. ) The image sensor 123 performs photoelectric conversion on the accumulated charge, thereby converting the accumulated charge into digital data corresponding to the amount of charge, and obtains image data (document image) (step S102). An image based on the image data acquired here is displayed on the display unit 141, for example.

  Based on the image data acquired in step S102, the control unit 110 executes “readability determination processing” for determining whether or not the characters shown in the document image are legible (step S200). Here, the readability level (image quality) is specified by detecting the size and the degree of blurring of the characters shown in the document image, and the document recorded as the document image according to the specified readability level. Judge whether or not interpretation is possible. Details of the “readability determination process” (step S200) will be described later.

  The control unit 110 executes “notification information output processing” for outputting predetermined notification information according to the determination result of “readability determination processing” (step S200) (step S500). Here, the notification information corresponding to the determination result of the “readability determination process” (step S200) is output from the output unit 140. In other words, when it is “unreadable”, information indicating that it is also displayed and information that prompts the user to change the shooting conditions is output. When it is “readable”, information indicating that the image can be taken is displayed. Output. Details of the “notification information output process” (step S500) will be described later.

  When the predetermined notification information is output by the notification information output process, the control unit 110 determines whether there is a photographing instruction from the user (step S103). Here, the shooting instruction is determined based on whether or not the full-press signal has arrived from the shutter button 131. For example, when the notification information output in step S500 indicates “readable”, the user decides to capture the document and presses the shutter button 131 fully. As a result, a full press signal is sent from the shutter button 131 to the control unit 110. That is, the half-press signal is switched to the full-press signal. When the control unit 110 receives a full-press signal from the shutter button 131, the control unit 110 determines that the instruction is a shooting instruction from the user (step S103: Yes), and records the image data acquired in step S102 in the image recording unit 150. (Step S104), the process ends.

  On the other hand, for example, when the notification information output in step S500 indicates “unreadable”, the user stops pressing the shutter button 131 to change the shooting condition, and the half-pressed state is released. As a result, the half-press signal sent to the control unit 110 is interrupted. Based on this, the control unit 110 determines that there is no shooting instruction from the user (step S103: No), ends the processing as it is, and waits for the next shooting preparation operation, for example.

  According to the “document shooting process” as described above, whether or not the document that is the target of document shooting is legible is determined based on the character size and the gradation change rate. Can be determined and notified. Therefore, the user does not wastefully record an image with low image quality. As a result, it is possible to prevent useless shooting operations and consumption of storage areas, and to save labor and improve shooting efficiency during shooting or post-processing.

  That is, it is not necessary to re-shoot many times at the time of shooting, and it is not necessary to select a legible one from the images that have been shot, and the user's labor can be reduced. In addition, although it is difficult to determine the image quality with the display device provided in the digital camera, in the above process, whether or not the control unit 110 can read using the image data acquired by the image sensor 123. Therefore, the image quality can be accurately determined.

  In the present embodiment, in the “readability determination process” (step S200) for determining whether or not it is legible, it is determined whether or not it is legible based on the character size of the document indicated in the image data and the degree of blur of the image. By doing so, an accurate and appropriate determination is made. Details of the “readability determination processing” (step S200) will be described below with reference to the drawings. First, “readability determination processing (1)”, which is an example of a method for determining readability, will be described with reference to a flowchart shown in FIG.

  In “readability determination processing (1)”, “feature detection processing” (step S210) for detecting the characteristics of the acquired document image is executed. In this “feature detection process”, the character size and the gradation change rate in the image shown in the document image are detected as features of the document image. The details of the “feature detection process” will be described with reference to the flowchart shown in FIG.

  The control unit 110 cuts out a character portion from the image data acquired in step S102. Here, characters can be cut out by a known technique. For example, after a background color is first detected, a portion composed of pixels having a color different from the background color is cut out as a character portion.

  For example, when the image data acquired in step S102 is a document image in which black characters are represented on a white background as shown in FIG. 9A, the control unit 110, for example, the four corners or edges of the image After the background color is specified as “white” based on the color of the pixels constituting the vicinity, a portion where “black” pixels are densely specified is specified as a character portion (step S211). In this case, the control unit 110 detects the gradation of the background color and the gradation of the character portion, and holds information indicating each detected gradation in the work area (step S212).

  In addition, the control unit 110 detects the position (coordinates) of each pixel constituting the character, and holds information indicating the detected position (hereinafter referred to as “position information”) in the work area (step S213). ). Further, the control unit 110 specifies a character range CA based on the detected position information (FIG. 9B), and cuts out a region including the range (step S214, FIG. 9C). Here, one of the characters shown in the document image may be cut out, or a plurality of characters may be cut out. In order to facilitate understanding, an example in which one character is cut out will be described below.

  The control unit 110 specifies the size of the cut out character (step S215). Here, the number of dots in the horizontal direction of the character is detected based on the coordinates of the pixels constituting the character detected at the time of clipping, and the detected number of dots is set as the size of the character. The control unit 110 holds information indicating the number of detected dots (hereinafter referred to as “character size information”) in the work area (step S216).

  Next, the control unit 110 detects an edge portion of the character based on the position information of the extracted character (step S217). Here, as in the XX ′ line segment shown in FIG. 5A, pixels are extracted in a one-dimensional direction from the cut out region, and the gradation of the background changes to the gradation of the character gradation. Is detected as an edge portion of the character. That is, among the extracted one-dimensional pixels, the pixel adjacent to the pixel indicating the background color gradation has a gradation different from the gradation of the background color, and the character color gradation. A location where a pixel adjacent to the pixel to be shown has a gradation different from the gradation of the character color is specified. And let the pixel between the specified two places be the edge part of the said character. The edge part detected here may be one place or a plurality of places. In order to facilitate understanding, a case where one edge portion is detected will be described below as an example.

  The control unit 110 obtains the number of dots in the edge part and the gradation change amount in order to obtain the gradation change rate at the detected edge part. Here, the number of pixels between the two locations specified in the edge detection in step S217 is detected as the number of dots in the edge portion, and information indicating the detected number of dots (hereinafter referred to as “dot number information”) is held in the work area. To do. The control unit 110 also obtains a difference between the gradation of the background color and the gradation of the character color and sets it as the gradation change amount. The control unit 110 holds information indicating the obtained gradation change amount (hereinafter referred to as “gradation change amount information”) in the work area (step S218).

  The control unit 110 uses the “dot number information” and the “gradation change amount information” acquired in step S218 to obtain the gradation change rate of the edge portion detected in step S217 (step S219). Here, the control unit 110 calculates “gradation change amount / number of dots” to obtain a gradation change rate at the edge portion, and information indicating the obtained gradation change rate (hereinafter, “tone change”). Rate information ”) in the work area. When the gradation change rate is obtained, the control unit 110 ends the “feature detection process” and returns to the flow of the “readability determination process (1)” illustrated in FIG.

  Next, the control unit 110 executes a “readability level specifying process” for specifying the readability level of the characters shown in the document image in accordance with the detection result in the “feature detection process” (step S230). The details of the “readability level specifying process” will be described with reference to the flowchart shown in FIG.

  The control unit 110 refers to the “readability level determination table” (FIG. 3A) in the storage unit 160, and the character size corresponding to the character size indicated in the character size information acquired in step S216 of the “feature detection process”. A level is specified (step S231). Similarly, a gradation change rate level corresponding to the gradation change rate indicated in the gradation change rate information acquired in step S219 of the “feature detection process” is specified (step S232).

  The control unit 110 specifies the readability level corresponding to the character size level specified in step S231 and the gradation change rate level specified in step S232 using the “readability level determination table” (step S233). When specifying the readability level, the control unit 110 stores information indicating the specified readability level (hereinafter referred to as “readability level information”) in the work area (step S234), and ends the “readability level specifying process”. Returning to the flow of “readability determination processing (1)” shown in FIG.

  The control unit 110 refers to the “readability level determination table” stored in the storage unit 160, and the readability level indicated in the readability level information acquired in step S234 of “readability level specifying process” (FIG. 10). By comparing with the threshold value, it is determined whether or not the character shown in the document image can be determined (step S250). Here, it is determined whether or not the readability level specified in the “readability level specifying process” is equal to or greater than a threshold value. In the example of the present embodiment, since the readability level serving as the threshold is “5”, it is determined whether or not the numerical value indicated by the specified readability level is “5” or more.

  If the identified readability level is equal to or higher than the threshold (step S250: Yes), the control unit 110 determines that the characters shown in the document image can be read (step S251). On the other hand, when the specified readability level is less than the threshold (step S250: No), the control unit 110 determines that the characters shown in the document image cannot be read (step S252).

  The control unit 110 holds information indicating the determination result in step S251 or step S252 (hereinafter referred to as “readability information”) in the work area (step S253), and the flow of “document photographing process” illustrated in FIG. Return to.

  In the “document photographing process”, predetermined notification information is output according to the determination result indicated in the legibility information acquired in step S253 (step S500). Here, if the determination result is “readable”, notification information indicating that is output from the output unit 140, and if it is “unreadable”, notification information that prompts the user to change the shooting condition is output.

  According to the “readability determination processing (1)” as described above, a character portion is cut out from a document image obtained by photographing a document, the size of the cut-out character is detected, and the gradation change rate of the edge portion of the character is obtained. Then, the degree of blur of the document image is determined. Then, based on the detected character size and gradation change rate, the readability level of the document image is specified, and the specified readability level is compared with a predetermined threshold value to thereby determine the character of the character indicated in the document image. Judgment of legibility is made.

  In the “readability determination process (1)”, the degree of blurring of the image is determined based on the gradation change rate at the edge portion of the cut out character, so that the legibility of the character can be determined accurately. However, in the “readability determination process (1)”, an appropriate determination can be made for a document having a large gradation difference between the background and the character portion (hereinafter referred to as “monochrome document”), such as a black and white document. When the characters in the document are composed of multiple colors or the background is composed of multiple colors (hereinafter referred to as “color document”), the gradation difference between the characters and the background color varies depending on the location. For this reason, it may not be possible to accurately determine whether or not the entire image can be read only by the readability level determined for a specific character. Alternatively, even when the size differs depending on the characters constituting the document, it is not possible to accurately determine whether or not the characters can be read only with the readability level determined for the specific characters. That is, even if it is determined as “readable” based on the readability level of a certain character, there may occur a case where the character can be read but other characters cannot be read. A process example for appropriately determining the legibility even when the background and characters have different gradation differences or when a plurality of different character sizes are mixed is shown as “readability determination process (2)”. This will be described below with reference to the flowchart shown in FIG.

  In this “readability determination process (2)”, a document image (color document image) obtained by photographing a document (color document) composed of characters of different colors as shown in FIG. Explained as an example Also, a plurality of areas are set in the acquired document image, and the character size and gradation change rate are detected for each area. Here, for example, a plurality (nine) areas as shown in FIG. 12B are set in the document image, and predetermined indexes (here, “1” to “9” are assigned consecutive numbers) in each area. ).

  First, the control unit 110 selects one of the set areas based on the assigned serial number (step S261). That is, the counter indicating the serial number assigned to the area is set to the initial value “1”, and the area corresponding to this value is selected.

  The control unit 110 executes the above-described “feature detection process” (step S210, FIG. 8) for the region selected in step S261 (step S262), “character segmentation”, “character size specification”, “edge” Part detection "," detection of gradation change rate "," determination of readability level "," determination of readability ", and" recording of determination result (holding in work area) ".

  The control unit 110 determines whether or not all areas have been determined to be legible (step S263). If there is an undetermined area (step S263: No), the control unit 110 increments the counter by “+1” (step S264). ) The “feature detection process” is executed for the area corresponding to the counter value (step S262).

  On the other hand, when the determination of readability for all areas is completed (step S263: Yes), the determination result for all areas is referred to, and whether there is a determination result of “unreadable” in any area. It is determined whether or not (step S265).

  When it becomes “illegible” in any region (step S265: Yes), the controller 110 determines that the entire document image is “illegible” (step S26). On the other hand, if “readable” is obtained in all the areas (step S265: No), it is determined that the entire document image is “readable” (step S267), and the flow of “document photographing process” shown in FIG. Return to.

  In the “document photographing process”, predetermined notification information is output based on the determination result in step S266 or S267 (step S500).

  According to the “readability determination process (2)” as described above, the document image is divided into a plurality of areas, and whether or not the entire image is readable is determined after determining whether or not each area is readable. Even when the gradation difference between the image and the background is not uniform or when different character sizes are mixed, it is possible to accurately determine whether or not the entire image can be read.

  The character size in the above “readability determination process (1)” and “readability determination process (2)” is determined based on the number of dots in the character portion of the document image. May be. For example, when shooting a document by shooting a document printed on paper of a prescribed size (JIS standard “A4 size”, etc.), the character size of the printed character (eg, “point”, etc.) When the character size indicated in a predetermined unit (hereinafter referred to as “default size”) is known in advance, the user inputs the default size by operating the operation input unit 132 or the like.

  In this case, it is assumed that shooting is performed so that the entire angle of view of the digital camera 100 matches the entire sheet to be shot. That is, shooting is performed so that the entire sheet can be accommodated by adjusting the distance between the digital camera 100 and the document or using a zoom function or the like. Here, the image performance (such as resolution and the number of pixels), the paper size (such as “A4 size”), and the default size unit (such as “point”) of the digital camera 100 are specified, so Based on the correspondence, the input specified size can be converted into the number of dots in the raster image.

  That is, a “character size correspondence table” as shown in FIG. 13 in which the paper size, the prescribed size, the image size, and the like are associated in advance with the character size (number of dots) is stored in the storage unit 160. Here, the “character size correspondence table” is set according to the image performance of the digital camera 100. The user operates the operation input unit 132 and the like to input a default size and a paper size before shooting. The control unit 110 acquires the number of dots corresponding to the input default size and paper size from the “character size correspondence table”, and specifies it as the “character size” in the document image obtained by photographing the document. The control unit 110 specifies the readability level from the “readability level determination table” based on the character size specified in this way and the gradation change rate acquired separately.

  In the “readability determination process (1)” and “readability determination process (2)”, the readability level is determined by specifying the readability level from the relationship between the gradation change rate and the character size. It may be determined whether or not it is legible. A processing example in this case will be described as “readability determination processing (3)” with reference to the flowchart shown in FIG.

  In this “readability determination process (3)”, the gradation change rate is detected at a plurality of locations in the document image, and the degree of blurring of the entire image is determined based on the appearance frequency of the gradation change rate to determine whether or not the image can be read. . In this “readability determination process (3)”, it is assumed that a plurality of areas are assigned to the document image as in the case of “readability determination process (2)” (see FIG. 12B).

  The control unit 110 selects one of the set areas based on the assigned serial number (step S271). That is, the counter indicating the assigned serial number is set to the initial value “1”, and an area corresponding to this value is selected.

  The control unit 110 executes the same process as the above-described “feature detection process” (step S210, FIG. 8) for the region selected in step S271, and performs “character segmentation”, “edge detection”, Detection of key change rate ".

  That is, the pixel of the portion where the gradation is changed is extracted from the pixel group in the one-dimensional direction in each region, and the gradation change rate (tone change amount / number of dots) of the portion is acquired (step S272). ). Here, it is assumed that the portion of the gradation change portion obtained in each region is arbitrary.

  The control unit 110 determines whether or not the gradation change rate has been acquired for all the regions (step S273). If there is an undetermined region (step S273: No), the control unit 110 increments the counter by “+1” (step S273: No). In step S274, the gradation change rate is acquired for the area corresponding to the counter value (step S272).

  On the other hand, when the gradation change rate is acquired in all regions (step S273: Yes), the control unit 110 creates a histogram indicating the appearance rate of the obtained gradation change rate (step S275). An example of the histogram created here is shown in FIG. The histogram shown in FIG. 15 shows an approximate distribution of gradation change rates in the entire document image by adding the gradation change rates acquired from each of a plurality of regions set in the document image. Here, the horizontal axis of the histogram indicates the acquired gradation change rate, and the vertical axis indicates the appearance frequency of the gradation change rate. The value that the gradation change rate on the horizontal axis can take is “0 to ∞”, and a value corresponding to the gradation change rate level set in the “readability level determination table” (FIG. 15), for example, is used as a standard index. It is set (in the example of FIG. 15, “5”, “10”, “20”, “40”).

  In such a histogram, an appearance frequency peak appears according to the degree of blur of the entire image. Here, a peak portion also appears near the minimum value and the maximum value (that is, “0” and “∞”) of the gradation change rate, and this peak is a portion where the same gradation value exists uniformly and widely, That is, since it corresponds to the background portion and the character portion, the peak portion excluding the peaks near “0” and “∞” is detected (step S276).

  Since such a peak portion indicates the most frequent gradation change rate in the image, the degree of blur of the entire image can be determined by checking the degree of the gradation change rate of the peak portion. Can do. For example, when the degree of blur of the entire image is high, the histogram is as shown in FIG. 15A, and when the degree of blur is low, the histogram is as shown in FIG. As described above, the gradation change rate becomes smaller as the width of the blur foot becomes wider (that is, the higher the degree of blur), so the higher the gradation change rate corresponding to the peak portion, the more the degree of blur of the image. The lower the gradation change rate corresponding to the peak portion, the higher the degree of image blur. Therefore, the degree of blurring of the image can be determined by looking at where the position where the peak portion appears on the horizontal axis of the histogram. In this case, a threshold is set on the horizontal axis of the histogram, and it is determined whether or not interpretation is possible by detecting whether the appearance position of the peak portion is right or left of the threshold. More specifically, it is determined whether or not the gradation change rate corresponding to the peak portion is greater than or equal to a threshold value (step S277).

  Here, in this processing, the gradation change rate serving as a threshold is set to “40”, for example. Since the character size is not considered in this process, even if the character size is small by setting a numerical value higher than the threshold value set in the “readability determination process (1)” and “readability determination process (2)” above. Appropriately determine whether or not it can be read. The threshold value can be set arbitrarily.

  When the gradation change rate corresponding to the detected peak portion is greater than or equal to the threshold (step S277: Yes), the control unit 110 determines that the character in the document image is “readable” (step S278), and the peak portion. If the gradation change rate corresponding to is less than the threshold value (step S277: No), it is determined as “unreadable” (step S279), and the flow returns to the “document photographing process” flow shown in FIG.

  In the “document photographing process”, predetermined notification information is output based on the determination result of step S279 or S280 (step S500).

  According to the “readability determination process (3)” as described above, it is possible to accurately determine the image quality from the degree of blur of the entire image without considering the character size, and determine whether or not the image can be read.

  The control unit 110 may be configured to execute any one of the “readability determination processes (1) to (3)”, but selects any one of the “readability determination processes (1) to (3)”. May be executed automatically. In this case, for example, the process can be selected by an instruction from the user. That is, since the “readability determination process (1)” is suitable for determining whether or not a monochrome document can be read, if it is clear that the document to be photographed is a monochrome document, the user can use the operation input unit 132 or the like. It is specified that the “readability determination process (1)” is executed by the operation. Similarly, if it is clear that the subject to be photographed is a color document, the “readability determination process (2)” is designated to be executed. For example, when it is clear that the character size is sufficiently large, such as when a large character written on a signboard is photographed, “readability determination processing (3)” that does not consider the character size is designated.

  The control unit 110 may determine which one of the “readability determination processes (1) to (3)” is applied by determining the type and character size of the document. That is, a suitable “readability determination process” is automatically selected and applied. In this case, by recognizing the background color and character color of the entire document image, it is determined whether the document is a “monochrome document” or “color document”, and the “readability determination process (1)” or “readability determination process” is performed. (2) "applies. In addition, when character extraction and size identification are first performed and the character size level is the maximum level (“level E” in the “readability level determination table” in FIG. 3A), “readability determination processing (3 ) "May be executed. As described above, the “readability determination processes (1) to (3)” may be selectively executed, and the control unit 110 executes all of the “readability determination processes (1) to (3)”. Based on the determination result, it may be possible to make a final determination as to whether or not readability is possible.

  When the readability is determined by the “readability determination process” as described above, predetermined notification information corresponding to the determination result is output by the “notification information output process” (step S500). Details of the “notification information output process” will be described with reference to the flowchart shown in FIG.

  The control unit 110 determines whether the determination result obtained by any of the above “readability determination processing” is “unreadable” or “readable” (step S501).

  When the determination result is “unreadable” (step S501: Yes), the control unit 110 accesses the storage unit 160, and the determination result is changed to “unreadable” from the “notification information table” (FIG. 3B). Corresponding notification information is acquired (step S502), and the notification method (output method) of the notification information is identified (step S503).

  The control unit 110 outputs the acquired notification information from the output unit 140 using the identified notification method (step S504). Here, for example, “message display”, “icon display”, “indicator light emission”, “voice notification”, “alarm notification”, and the like are prepared as notification methods. These notification methods are arbitrarily selected according to the user's desire. In other words, the notification method is set in advance by operating the operation input unit 132 or the like. One of these notification methods may be selected, or a plurality may be selected in combination.

  When “message display” is selected, for example, a message indicating that characters cannot be read even if shooting is performed under the current shooting conditions, or a message prompting the change of shooting conditions is displayed as text information as shown in FIG. As notification information on the display unit 141. That is, text data indicating such a message is acquired from the “notification information table” in the storage unit 160 and displayed on the display unit 141.

  When “icon display” is selected, for example, a predetermined icon (pictogram) indicating that the character cannot be read is displayed as notification information on the display unit 141 as shown in FIG.

  When “indicator light emission” is selected, for example, the indicator portion 143 is lit or blinks with a predetermined light emission pattern indicating that the character cannot be read. Here, as the light emission pattern, for example, a light emission color or a blinking pattern is defined in advance. For example, in the case where the indicator unit 143 is composed of LEDs of two colors of “green” and “red”, which color is to be blinked and how (flashing speed, etc.) is defined. Various detection events (for example, focusing, battery exhaustion, etc.) are associated in advance. As a light emission pattern indicating that the character cannot be read, for example, “flashing red LED” can be used. In this case, the prescribed light emission pattern functions as notification information.

  When “sound notification” is selected, for example, a message that the text cannot be read even if shooting is performed under the current shooting conditions, or a message that prompts the user to change the shooting conditions is output from the sound report unit 142. In this case, the text data acquired in the case of “message display” is acquired, and the message indicated by the text data is output from the sound report unit 142 as if a person is speaking by a predetermined speech synthesis method. Is done. In this case, the output voice functions as notification information.

  When “alarm notification” is selected, for example, an alarm sound is output from the sound report unit 142 with a predetermined sound pattern indicating that the character cannot be read. In this case, the report sound pattern is defined by, for example, tone color, pitch, or sound length, and is associated with various detection events (for example, in-focus, battery exhaustion, etc.) in the digital camera 100 in advance. Therefore, the alarm sound is output from the sound report unit 142 in the sound sound pattern associated with the event “character cannot be read”. In this case, a prescribed sound report pattern functions as notification information.

  When the notification unit outputs the notification information as described above, the control unit 110 returns to the “document photographing process” flow shown in FIG.

  On the other hand, when the determination result is “readable” (step S501: No), the process returns to the “document photographing process” (FIG. 6) without outputting the notification information as described above.

  According to the “notification information output process” as described above, during the shooting preparation operation at the time of document shooting, whether or not the characters after shooting can be read and whether or not the shooting conditions should be changed by various methods (shooting by the user). Person). Therefore, the user can determine on the spot whether or not to continue shooting under the current shooting conditions. As a result, useless shooting can be avoided, the waste of the memory area can be prevented, the user's labor can be reduced, and shooting efficiency can be improved.

  In the above “notification information output process”, the notification information is not output when the determination result is “readable”, but the notification information indicating that is output when the determination result is “readable”. You may do it.

  Further, in the case of determining using the “readability level determination table”, advice information corresponding to the readability level that is the basis of the determination may be output as notification information. For example, if the readability level is determined to be “unreadable” because it is one step lower than the threshold value, (1) “get close to the shooting target or use the zoom function to increase the character size”, (2 ) Advice indicating that it can be “readable” by taking measures such as “Use a tripod to prevent camera shake”, (3) “Use a flash to increase the amount of light”, etc. Information may be output.

  Further, the selection of the countermeasure may be determined by the control unit 110 based on the “character size level” and the “gradation change rate level” that are the basis of the readability level. In the case of the “readability level determination table” shown in FIG. 3A, the readability level of “AD ′ level” is “4”, which is one step lower than the threshold (“5”). In the case of “−D ′ level”, although the gradation change rate level is sufficient, the readability level is “4” because the character size is small, and as a result, it is determined as “unreadable”. In such a case, a countermeasure for increasing the character size (such as (1) above) is output as advice information. On the other hand, if the character size is sufficiently large, but the gradation change rate is low (the degree of blurring is large) and it is determined as “unreadable”, a countermeasure for suppressing blurring and camera shake ( (2) or (3)) is output as advice information.

  Further, even when the determination result is “readable”, for example, when the readability level is “one legible” because the readability level is one level higher than the threshold value, it is more reliable by changing the shooting condition. Advice information indicating that a legible image can be obtained may be output.

  As described above, according to the above-described embodiment of the present invention, it is possible to accurately determine the image quality of the document image and notify the user whether or not the document image can be read before recording the document image. As a result, useless shooting operations can be avoided, and shooting efficiency can be improved and user effort can be reduced. Moreover, since useless recording can be avoided, it is possible to prevent waste of photographing resources such as a memory area.

  The document image acquired in the above embodiment is preferably converted into text data or the like by character recognition processing, but the execution form of such character recognition processing is arbitrary. For example, the character recognition process can be executed by a computer device such as a personal computer. In this case, the document image data recorded in the image recording unit 150 is transferred to the computer apparatus, but the transfer method is arbitrary. For example, a predetermined interface (for example, USB (Universal Serial Bus) provided in the digital camera 100 is used. The document image data can be transferred by connecting to a computer device via a) or the like. Further, when the image recording unit 150 is constituted by a removable memory card, the computer device can transfer the document image data by reading it from the memory card.

  Such a character recognition process may be executed by the digital camera 100. In this case, a program (character recognition program) for performing character recognition processing is stored in the storage unit 160 of the digital camera 100, and the control unit 110 executes this character recognition program to perform character recognition processing. When character recognition processing is executed by the digital camera 100, conversion to text data can be performed on the spot where the image is taken. Text data converted by the digital camera 100 can be recorded in the image recording unit 150 or transferred to a computer device or the like.

  In the above embodiment, the notification information is output according to the determination result of legibility. However, the photographing operation of the digital camera 100 is automatically controlled according to the determination result to assist the user. Good. For example, if the gradation change rate is sufficient but the character size is small and it is determined that the character cannot be read, the control unit 110 drives the lens unit 121 to control the character size to be increased by the zoom function. Or when the camera shake occurs due to insufficient light quantity, the flash may be automatically emitted.

  In addition, by applying a program for realizing each process described above to a digital camera whose functions can be expanded by updating the operation program, the existing digital camera can function as the digital camera 100 according to the embodiment. it can. In this case, the operation program according to the above embodiment is installed in the digital camera, and is executed by the control unit of the digital camera. As a result, the above-described processes are executed, and it is possible to accurately determine the image quality at the time of document shooting and to determine and notify whether or not the document can be read. A method for distributing such an operation program is arbitrary. For example, the operation program can be distributed by being stored in a recording medium such as a CD-ROM or a memory card, and can also be distributed via a communication medium such as the Internet. it can.

  As described above, by applying the present invention as in the above embodiment, when a document is photographed, it is possible to accurately determine and notify the image quality based on the gradation change rate in the image. As a result, it is possible to save labor at the time of shooting and improve shooting efficiency, and to prevent the memory area from being wasted.

1 is a diagram illustrating a system configuration of a digital camera according to an embodiment of the present invention. It is a figure which shows the external appearance of the digital camera concerning embodiment of this invention, (a) is a front view, (b) is a rear view. FIG. 2 is a diagram illustrating an example of information recorded in a storage unit illustrated in FIG. 1, where (a) illustrates an example of a “readability level determination table” and (b) illustrates an example of a “notification information table”. It is a figure for demonstrating the image quality of the image imaged with the digital camera concerning embodiment of this invention, (a) shows the example of a document image, (b) Document image when there is no blur and camera shake (C) shows the state of the pixels in the document image when there is a blur or camera shake. It is a figure for demonstrating the gradation change rate in the image imaged with the digital camera concerning embodiment of this invention, (a) shows the example of a document image, (b) When there is no blurring or camera shake (C) shows the state of gradation change when there is a blur or camera shake. It is a flowchart for demonstrating the "document photographing process" concerning embodiment of this invention. FIG. 7 is a flowchart for explaining an example of a “readability determination process” executed in the “document photographing process” shown in FIG. 6. It is a flowchart for demonstrating the "feature detection process" performed by the "readability determination process (1)" shown in FIG. FIGS. 9A and 9B are diagrams for explaining character region segmentation executed in the “feature detection process” illustrated in FIG. 8, in which FIG. 8A illustrates an example of a target document image, and FIG. The range is shown, and (c) shows an example of the extracted character. It is a flowchart for demonstrating the "readability level specific process" performed by the "readability determination process (1)" shown in FIG. 7 is a flowchart for explaining another example of “readability determination processing” executed in “document photographing processing” shown in FIG. 6. FIG. 12 is a diagram for explaining the process of “readability determination process (2)” illustrated in FIG. 11, where (a) illustrates an example of a document image to be processed, and (b) illustrates an example of an area to be processed. . It is a figure which shows the example of the "character size corresponding | compatible table" used when specifying the character size in "readability determination processing (1)" or "readability determination processing (2)" by arbitrary input. 7 is a flowchart for explaining another example of “readability determination processing” executed in “document photographing processing” shown in FIG. 6. It is a figure which shows the example of the histogram created by the "readability determination process (3)" shown in FIG. 14, (a) shows the example of the histogram acquired with an image with high blur condition, (b) is the blur condition. The example of the histogram acquired with an image with low is shown. 7 is a flowchart for explaining a “notification information output process” executed in the “document photographing process” shown in FIG. 6. It is a figure which shows the example of the notification information output by the "notification information output process" shown in FIG. 16, (a) shows the example of a display of notification information in case a notification system is "message display", (b) Shows a display example of notification information when the notification method is “icon display”.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 ... Digital camera, 110 ... Control part, 120 ... Imaging part, 121 ... Lens unit, 122 ... Ranging and photometry part, 123 ... Imaging element, 130 ... Input part, 131 ... Shutter button, 132 ... Operation input part, 140 ... output unit, 141 ... display unit, 142 ... reporting unit, 143 ... indicator unit, 150 ... image recording unit, 160 ... storage unit

Claims (12)

In a digital camera that acquires image data by converting incident light into digital data,
Feature detection means for detecting the feature of the image data;
Determination means for determining the readability level of the character indicated in the image data based on the feature detected by the feature detection means;
Execution means for executing a predetermined operation based on the readability level determined by the determination means,
The feature detection means further comprises a gradation detection means for detecting a gradation change rate in the image data,
The determination means determines the readability level of the character based on the gradation change rate detected by the gradation detection means,
The execution unit further includes an output unit that outputs predetermined notification information based on the readability level determined by the determination unit.
A digital camera characterized by that. An edge detecting means for detecting an edge of the character indicated in the image data;
The gradation detection means detects a gradation change rate of the edge portion detected by the edge detection means;
The digital camera according to claim 1. The feature detecting means further comprises character size specifying means for specifying the size of the character,
The determination means determines the readability level of the character based on the gradation change rate detected by the gradation detection means and the character size specified by the character size specification means.
The digital camera according to claim 1, wherein the digital camera is a digital camera. A determination unit that determines whether or not the character can be read based on a predetermined threshold and the readability level determined by the determination unit;
The output means outputs notification information indicating that when it is determined by the determination means that it cannot be read,
The digital camera according to claim 1, wherein the digital camera is a digital camera. The determination means determines a readability level at a plurality of locations on the image data,
The determination means determines whether or not the plurality of places can be read,
The output means outputs the notification information when it is determined that it is unreadable at any of the plurality of locations.
The digital camera according to claim 4. A storage unit for storing determination information in which information indicating the feature detected by the feature detection unit is associated with a readability level;
The determination unit determines a readability level based on determination information stored in the storage unit.
The digital camera according to any one of claims 1 to 5, wherein: It further comprises an input means for inputting an instruction from the user,
The storage means updates the determination information in accordance with a user instruction input to the input means.
The digital camera according to claim 6. The character size specifying means specifies a character size in accordance with a user instruction input to the input means;
The digital camera according to claim 7. The gradation detection means further comprises gradation distribution detection means for detecting distribution of gradation change rate in the image data,
The determination unit determines a readability level of the character based on a distribution of gradation change rates detected by the gradation distribution detection unit;
The digital camera according to claim 1, wherein the digital camera is a digital camera. The determination means determines the readability level based on the distribution of the gradation change rate, with the gradation change rate having the highest appearance frequency as the gradation change rate in the character.
The digital camera according to claim 9. A character area extracting means for extracting a character area shown in the image data;
The feature detection means detects a feature in the character area extracted by the character area extraction means;
The digital camera according to claim 1, wherein the digital camera is a digital camera. To a computer that controls a digital camera that converts incident light into digital data and acquires image data,
Detecting characters shown in the acquired image data;
Detecting a gradation change rate in at least the detected character;
Identifying the size of the detected characters;
Determining a readability level of the character based on the detected gradation change rate and / or the specified character size;
Outputting predetermined notification information according to the determined readability level;
A program characterized by having executed.

Images