JP2001245139A - Digital image recording apparatus and method, and transmission method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain a threshold value for binarizing a captured color image correctly and in a fixed processing time. SOLUTION: To re-size a captured image into a predetermined size to store it on an area different from the for the captured images on a DRAM (step 22), before moving to step S23 where a threshold is derived after the captured image is pasted on DRAM at a step S21. At a step 23, an optimum threshold is derived based on color image data of the re-sized image. Since distribution of color information for the original image and that for the resized image are relatively the same, a same threshold can be obtained using any image. A threshold is derived after an image is re-sized to a predetermined sized image, so that the same algorism is applicable, irrespective of the size of the recorded image to be selected, thus making it possible to keep constant the processing time required for deriving a threshold.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a digital image recording apparatus and method applicable to, for example, a digital camera apparatus, and a transmission method.

[0002]

2. Description of the Related Art Recently, a digital image is recorded on a digital still camera or the like using a nonvolatile semiconductor memory device such as a flash memory or a recording medium such as a hard disk or a floppy disk, and recording an object image as image data on the recording medium. Recording devices are rapidly becoming widespread. The digital image recording apparatus is configured to convert a photographed subject image into a digital image signal, compress the digital image signal, and record the compressed image information on a recording medium. In a digital image recording apparatus, a natural image is captured as a color image, and J
What compresses with PEG is known.

When photographing a character document, a white board, or the like, it is preferable to record the photographed color image in binary form rather than recording it as a color image. That is, FIG. 16A shows a color image obtained by photographing a character document, and FIG. 16B shows an image obtained by converting the color image into a binary image. As can be seen from the figure, in the binary image, the character of the subject and the background are more clearly distinguished than in the color image, and the character becomes easier to read.

In order to binarize a color image, it is necessary to discriminate each pixel into white and black by a threshold value in a photographed color image. FIG. 16B shows a case where an optimal threshold value is used. If the threshold value is too high than the optimum threshold value, the binarized image becomes dark as a whole as shown in FIG. 16C, and if the threshold value is too low below the optimum threshold value, as shown in FIG. As shown in FIG. 16D, the binarized image becomes bright overall and characters are difficult to see.

[0005]

In particular, when an image is taken by a digital color image recording apparatus, one of a character and a background image does not always take a fixed value.
Since both values can take any value in the color space, a fixed threshold cannot be set in advance. Therefore, it is necessary to obtain a threshold value using all pixel data of a captured image. In that case, the following problem occurs.

When the number of pixels or the size of an image to be recorded is plural, the amount of data to be processed is different, the time required for deriving the threshold is different, and the total number of pixels is different. The derivation algorithm cannot be shared. The use of all pixel data causes a problem that the processing time becomes longer. Furthermore, when all pixel data is used, information such as noise, deterioration, and unnecessary unnecessary objects in the periphery of the image affects the derivation of the threshold value, and a correct threshold value cannot be obtained. For example, peripheral objects other than the subject, such as a character document or a white board to be binarized, may be included in the image, which hinders the derivation of a correct threshold value.

SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to provide a digital image recording apparatus and method, and a transmission method capable of solving the above-mentioned problems when obtaining a threshold value for binarization for each captured image. Is to provide.

[0008]

SUMMARY OF THE INVENTION In order to solve the above-mentioned problems, a first aspect of the present invention is a digital image recording apparatus for recording an image as a digital signal on a recording medium. Image processing means for converting a captured color image into a binary image, and recording means for recording the output of the image processing means on a recording medium, wherein the image processing means converts the captured color image into a fixed image. A digital image recording apparatus is characterized in that it is converted into an image having a size or the number of pixels, a threshold value is determined based on the converted image, and a binary image is generated based on the threshold value. The invention according to claim 9 is a method for converting a color image into a binary image and recording it.

According to a second aspect of the present invention, there is provided a digital image recording apparatus for recording an image as a digital signal on a recording medium, comprising: an image capturing means for capturing a color image;
Image processing means for converting the captured color image into a binary image, and recording means for recording the output of the image processing means on a recording medium, wherein the image processing means thins out the captured color image. A digital image recording apparatus characterized in that a thinned image is generated by the above method, a threshold value is determined based on the thinned image, and a binary image is generated by the threshold value. A tenth aspect of the present invention is a method of converting a color image into a binary image and recording it.

According to a third aspect of the present invention, there is provided a digital image recording apparatus for recording an image as a digital signal on a recording medium, comprising: an image capturing means for capturing a color image;
Image processing means for converting the captured color image into a binary image; and recording means for recording the output of the image processing means on a recording medium. A digital image recording apparatus characterized in that a threshold value is determined based on a partial image including a subject to be binarized, and a binary image is generated based on the threshold value. The invention according to claim 11 is a method for converting a color image into a binary image and recording the converted image.

According to a twelfth aspect of the present invention, there is provided a digital image transmission method for transmitting an image as a digital signal to a communication medium, comprising the steps of: converting a captured color image into a binary image by a threshold value; Transmitting to a communication medium.
A digital image transmission method characterized by converting an image into an image having a fixed size or the number of pixels, determining a threshold value based on the converted image, and generating a binary image based on the threshold value.

According to a thirteenth aspect of the present invention, there is provided a digital image transmission method for transmitting an image as a digital signal to a communication medium, comprising the steps of: converting a captured color image into a binary image by a threshold value; Transmitting to a communication medium.
A digital image transmission method characterized by generating a thinned image by thinning an image, determining a threshold value based on the thinned image, and generating a binary image by the threshold value.

According to a fourteenth aspect of the present invention, there is provided a digital image transmission method for transmitting an image as a digital signal to a communication medium, comprising the steps of: converting a captured color image into a binary image by a threshold value; Transmitting to a communication medium.
A digital image transmission method characterized in that a threshold value is determined based on a partial image including a subject to be binarized in an image, and a binary image is generated based on the threshold value.

According to the first, ninth and twelfth aspects of the present invention, the threshold value is obtained by using the resized image regardless of the size of the image to be recorded. And the processing time can be prevented from fluctuating.

According to the second, tenth, and thirteenth aspects of the present invention, since the threshold value is obtained by using the thinned image, the processing can be speeded up without any trouble in deriving the threshold value.
By changing the thinning process according to the subject,
It is possible to derive a correct threshold value with a minimum number of data.

According to the third, eleventh and fourteenth aspects of the present invention, the threshold value is derived using the information of the partial image instead of using the data of the entire image, so that it is affected by unnecessary information. It is possible to derive a correct threshold value for optimally binarizing an intended portion without any need. Further, by enabling the user to arbitrarily set the partial image to be used, the user can arbitrarily select a portion to be binarized.

[0017]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below. FIG. 1 shows a system configuration of the present embodiment, and 1 indicates a CCD (Charge Coupled Device). CC
The number of pixels of D1 (the number of horizontal pixels × the number of vertical pixels) is, for example, U
XGA (1600 × 1280) pixels. The number of pixels of the recorded image is UX relative to the number of pixels of the captured image (color image).
In addition to GA, SXGA (1280 x 1024) pixels, XGA (1
024 × 768) pixels and VGA (640 × 480) pixels can be selected. The CCD 1 outputs a subject image passed through a lens unit (not shown) as an imaging signal. In the lens unit, an automatic aperture control operation and an automatic focus control operation are performed.
An imaging signal is supplied to the camera block 2.

Note that the CCD 1 may be capable of performing an operation of reading a document in the same manner as an image scanner. Also, when processing digital color images received from other devices and communication media besides the CCD,
The present invention can be applied. Further, the present invention is not limited to recording the processed image on a recording medium, and may be applied to a case where the processed image is transmitted to a communication medium.

The camera block 2 includes a clamp circuit, a luminance signal processing circuit, a contour correction circuit, a defect compensation circuit, an automatic aperture control circuit, an automatic focus control circuit, an automatic white balance correction circuit, and the like. For example, RG from the camera block 2
A digital imaging signal is generated in the form of a component signal composed of a luminance signal and a color difference signal converted from the B signal. The digital imaging signal is supplied to the memory control block 3.

The memory control block 3 has a signal switching section, a display buffer memory, a D / A converter, and the like. The display device 4 and the data transmission path 5 are connected to the memory control block 3. In the memory control block 3, the generated RGB signals are supplied to the display device 4 via the D / A converter. The display device 4 is a liquid crystal display (LCD) provided integrally with the camera.
ay) and other display devices. When the image signal from the camera block 2 is supplied to the display device 4, the image being captured is displayed, and the read image of the recording medium 9 supplied via the data transmission path 5 is displayed. The display device 4 displays a VGA (640 × 480) image.

For the data transmission line 5, a DRAM (Dyn
amic Random Access Memory) 6 and an image processing block 7 composed of a microcomputer. The DRAM 6 is controlled by the memory control block 3 or the image processing block 7 and has an area for storing an original image via the memory control block 3 and an area for storing image data after image processing by the image processing block 7.

An operation input unit 8 and a recording medium 9 are connected to the image processing block 7 via interfaces. When control information is supplied from the image processing block 7 to each unit, image data is processed, data is written to and read from the DRAM 6, and writing and reading to and from the recording medium 9 are executed.

The operation input unit 8 has a shutter button, a mode designation switch, and other various switches operated by the photographer. For example, a first mode designated mainly when photographing a natural image and a second mode designated mainly when photographing an image mainly composed of characters such as a text document and a white board can be selected. An operation input from the operation input unit 8 is supplied to the image processing block 7. The recording medium 9 is a memory card (IC card), a floppy disk, a rewritable optical disk, or the like, and is detachable from the still camera body. In addition to the recording medium 9, a communication medium such as the Internet may be used.

In the image processing block 7, the image processing method differs between the first mode mainly used for photographing natural images and the second mode mainly used for photographing character images. In the first mode, for example, JPEG (Jo
int Photographic Experts Group) is used. Second
In the mode, the image is binarized, and the LZW (Lemp
(el Ziv Welch) method, and a process of adding necessary components to the compressed data to convert the binary image into a GIF file.

JPEG is a standard encoding method for compressing a color still image, and includes a lossless encoding method and an irreversible encoding method. As a lossless encoding method, an intra-space prediction encoding method is adopted, and as an irreversible encoding method, DCT (D
A compression method using iscrete cosine transform) is employed. Normally, it is considered that a slight deterioration in image quality due to the irreversible coding method does not cause any practical problem, and the coding method using DCT is used as JPEG. Also in this specification, the term JPEG refers to DCT and lossy encoding in which coefficient data generated by DCT is quantized and the quantized output is encoded by entropy encoding.

The image processing block 7 controls writing and reading of data to and from the recording medium 9.
The image data stored in the DRAM 6, that is, the JPEG file obtained in the first mode or the GIF file obtained in the second mode is output to the recording medium 9. These image files read from the recording medium 9 are stored in the DRAM 6 by the image processing block 7.

Further, a resolution conversion unit 10 is provided, and the resolution conversion unit 10 performs a process of converting the photographed image into the resolution of the selected recording image. In the DRAM 6,
An image whose resolution has been converted is obtained. The resolution conversion process is
The processing may be performed in the image processing block 7.

In the above-described embodiment, when the photographer presses the shutter button (operation input section 8), a color image signal captured by the CCD 1 is supplied to the camera block 2, the camera signal is processed, and the resolution is converted. Of the original image data under the control of the memory control block 3
Stored in AM6.

When the original image data is stored in the DRAM 6, the original image data is processed by the image processing block 7, and the compressed image data (JPEG file or GIF) is processed.
File) is stored in another area of the DRAM 6. Then, the compressed image data read from the DRAM 6 by the image processing block 7 is written to the recording medium 9.

When recording compressed image data, a file name is given in the image processing block 7. When the recording medium 9 is a memory card, a still image directory (DCIM) is defined, and the still image directory (D
CIM) defines subdirectories such as MSDCF. A subdirectory is equivalent to an album. If it is one image compressed by JPEG,
DS for subdirectory eg 100MSDCF
C00001. jpg file name and extension are added. Next, the image data recorded on the memory card is GI
F file, the directory and subdirectory are the same, and TXT00002. The file name and extension of the gif are added. Numbers from (0001) to (9999) are added after each of DSC0 and TXT0.

When an image stored in the recording medium 9 is reproduced, desired compressed image data is read from the recording medium 9 by designating a file name, and decompressed by the image processing block 7. The decompressed image data is written to the DRAM 6. Then, the image data stored in the DRAM 6 is displayed on the display device 4 via the memory control block 3.

In the above-described embodiment, the image processing in the second mode mainly used for photographing a character document or the like will be described in more detail. In the second mode, the image processing block 7 performs an image binarization process. That is, an optimum threshold value is calculated based on the color image data taken into the DRAM 6, and the color image data is converted into binary (white and black) using this threshold value. The luminance data in the color image data is binarized. Although the binarization process can be performed in the CCD 1, the binarization process in the image processing block 7 allows the image processing block 7 to perform a process such as setting a threshold value.

As will be described later in detail, the distribution of luminance data is examined for each image to be processed, and a threshold value for discriminating a character from a background is calculated based on the distribution. You. When the data of the original image is used for calculating the threshold value, the number of pixels is large. Therefore, it is preferable to use only the image data obtained by thinning out the original image or only the image data near the center, for example, in the original image. .

Next, to create a GIF file, the LZW
Data compression is performed according to the method. In the LZW method, a pattern of an arbitrary length appearing in a data stream is registered in a dictionary (code table), and when the same pattern appears next, a registration number (variable length code) is used as an encoded output. is there. There is no need to edit a dictionary for registering a pattern prior to encoding, and a dictionary is created while reading data.

A GIF file is created from the compressed data. The structure of the GIF file will be specifically described.
FIG. 2 shows an example of a file structure of a general GIF file. GIF files are roughly divided into header block 1
1, a logical screen description block 12, an application extension block 13, a graphic control extension block 14, an image data block 15, and a trailer block 16. By creating these blocks, a GIF file is created.

The header block 11 is composed of, for example, 6 bytes and is arranged at the head. This header block 11
Indicates that the data stream is in GIF format. The header block 11 includes a signature field indicating the start of the data stream, and a version field necessary for performing decoding completely. In addition,
One header block is required for the data stream.

Following the header block 11 is a logical screen description block 12. The logical screen description block 12 defines parameters (size, aspect ratio, color depth) necessary to define an area of an image plane (display device) for rendering an image.
The logical screen description block 12 defines the presence / absence of a global color table and various parameters thereof. This logical screen description block is also mandatory, and there must be exactly one in the data stream.

Following the logical screen description block 12 is a global color table block 12a. The color table is a palette that represents RGB values for all colors used in the image as a set of 3 bytes (24 bits). Since the GIF supports up to 256 colors, the global color table can be up to 2 colors.
Includes 56 x 3 bytes. This is the default palette and is used when subsequent images do not have a dedicated local palette. Although this block is optional, the maximum number of global color tables that can be specified for one data stream is one.

Global color table block 12a
The application extension block 13 is placed next to the application. The application extension block 13 includes unique information for only a specific application to perform special processing on image data.

The graphic control extension block 14 is provided next to the application extension block 13. The graphic control extension block 14 includes parameters for controlling a method of displaying an image. The adaptation range is only the first image immediately following. Note that a GIF file can be configured without disposing the block 13, and one graphic control extension block 14 can be disposed before the image data.

The image data block 15 is arranged next to the graphic extension block 14. Each image of the data stream is composed of an image descriptor block 15a and compressed data 15c.

The image descriptor block 15a contains the parameters needed to process a table-based image. The coordinates specified in this block indicate the coordinates of the logical screen, and are in pixel units. The image descriptor block 15a is a graphic rendering block, which may be preceded by one or more control blocks such as a graphic restriction extension, or may be followed by a local color table. Note that, after the image descriptor block 15a, the compressed data 15
c follows. That is, the image descriptor block 15a
Image descriptors that are mandatory for the image and can be specified for the image present in each data stream are:
There is only one. Note that there is no limit on the number of images present in the data stream.

The table-based compressed data 15c is composed of an array of sub-blocks. Compressed data 15
Each sub-block making up c is 255 bytes at maximum and contains an index to the color table.

Then, the graphic control extension block 14 and the image data block 15 described above are repeated by the number of images to be displayed as continuous images, and a trailer block 16 is arranged at the end of the file. The trailer block 16 is a block composed of a single field indicating the end of the GIF data stream. In the case of a GIF file, be sure to use trailer block 1
6 and the trailer block 1
No. 6 cannot be changed.

To create the image data block 15, the original image is converted to a binary image, and the binary image is converted to a GIF.
Is converted to an index value indicating the color palette number. As described later, the binarization process and the conversion to the index value may be performed at once. In that case, the memory used for the binarization processing can be effectively used.

As described above, in one embodiment, LZW
Data compression is performed according to the method. In the LZW method, the smaller the number of patterns appearing in a data stream,
The possibility of matching with the registered contents of the dictionary is increased, and the compression ratio can be increased. A binary image is a data stream consisting of only two values (0 and 1), and the number of patterns that appear is significantly smaller than that of a color image, so that the compression ratio can be increased. In other words, the data size of the compressed image data is reduced.

As an example, in the case of (640 × 480) pixels,
The data size is about 370 kB (kilobytes) of the captured original color image data. If this color image is compressed by JPEG, the data size becomes about 70 kB. As in one embodiment, data compression by the LZW method is performed, and the data is converted into a GIF file.
Data size. As described above, the compression ratio of about ４ to Ｊ in JPEG compression is compressed to about 1/30 in LZW.

The LZW method is a reversible compression method in which a data stream before compression can be completely restored from a sequence of dictionary registration numbers, that is, the same data as original data can be restored from compressed data. . Meanwhile, JPE
G is lossy compression. A binary image is an image having an extremely small number of colors and many sharp edges.
When compressed and decompressed using PEG, there is a disadvantage that the decompressed image contains much image noise. In terms of image quality,
GIF files are advantageous.

Next, a binary image obtained by binarizing the image is converted to a GIF.
An example of a process of converting to a format will be described. The GIF file is to generate blocks as shown in FIG. An example of the processing will be described with reference to FIG.

FIG. 3A shows a captured color image data stream. One pixel is represented by RGB 3-byte data. Next, the color image is subjected to binarization processing to obtain a stream of binarized image data as shown in FIG. 3B. In the binarization processing, pixel data representing black is converted into (R = G = B = 0), and pixel data representing white is converted into (R = G = B = 255). And FIG. 3C
As shown in (2), the index value is converted to 0 (black) or 1 (white) of the index value indicating the color palette.

The processing shown in FIG.
A conversion process to a value image and a conversion process from a binary image to an index value are required. By two conversion processes,
Processing time becomes longer, and a problem arises in terms of effective use of the memory (DRAM 6). Thus, in one embodiment,
By the method described below, binarization and GIF file creation processing are performed.

Since the color table in the global color table block 12a in FIG. 2 is a binary image, the colors used for the image can be determined in advance to include only two colors, white and black. . That is, in the configuration of the color table, as shown in FIG. 4, the index value 0 corresponds to black (R, G, B = 0), and the index value 1 corresponds to white (R, G, B = 255). Can be predetermined. The present invention is also applicable to a case where each component of a color image is represented by Y (luminance signal), Cb (blue color difference signal), and Cr (red color difference signal). In that case, the information representing black is (Y = 0, Cb = Cr = 12
8), and the information representing white is (Y = 255, Cb = Cr)
= 128).

Next, in the data block 15, the stream actually compressed by the LZW is not the original image data stream itself but a stream of index values indicating the color of each pixel of the original image. It is. In the case of a black and white binary image having the color table shown in FIG. 4, a stream composed of binary values of index values "0" and "1" is compressed as shown in FIG.

As described above, in the case of a binary image, the white and black pixels can be determined to have index values "1" and "0" in advance, so that the original color image data as shown in FIG. By performing a process of binarizing a stream and a process of converting the stream into index values at a time, FIG.
A stream having an index value as shown in FIG. Therefore, as compared with the processing shown in FIG. 3, the processing can be simplified and the processing time can be reduced. Also, as shown in FIG. 3A, the original image information is represented by 3 bytes per pixel, so that the data after the first conversion is 3 bytes per pixel as shown in FIG. 3B.
Requires a byte memory area. In contrast, FIG.
According to the processing shown in (1), since the data after the first conversion is a stream of index values, one data per pixel
It requires only a byte memory area, and has no memory (DRAM
6) can be effectively used.

Here, in order to facilitate understanding of the present invention, an existing binarization process will be described with reference to FIG. In step S11, the captured image is stored in the DRAM 6
I can be attached. The capture image is a recorded image, that is, an image of a selected size or number of pixels. Then, in step S13, an optimum threshold value is derived based on the color image data of the captured image. In step S14, the color image data of the captured image is binarized using the obtained optimum threshold value.

The recordable image size or the number of pixels can be selected by a user operation, and a plurality of types of images can be recorded in the same system. For example, SXGA, XG
Any of A and VGA can be selected as the size of the recorded image. The resolution conversion unit 10 converts the number of pixels. Therefore, the image stored in the DRAM 6 is
The number of pixels differs depending on the size selected at the time of shooting. Not the resolution conversion unit 10 but the image processing block 7
May be used to convert the number of pixels.

In the existing method, in order to use an image stored in the DRAM 6, a process for deriving a threshold value includes the following steps.
The number of operations differs depending on the number of pixels, and the processing time also differs. Further, when the image sizes are different, the algorithm for deriving the threshold cannot be common. For example, when a histogram is used to derive a threshold, the total number of pixels differs due to a difference in size, and processing such as normalization of the histogram is required.

FIG. 8 shows a process according to an embodiment of the present invention which can solve these problems. In step S21, after the captured image is attached to the DRAM 6, before the process proceeds to step S23 for deriving an optimal threshold value, the captured image is resized to a predetermined fixed size.
Above, it is stored in an area different from the area of the captured image (step S22). The term “resizing” means an image processing for reducing the original image with high accuracy and relatively reducing the total number of data without changing the color information or the like of the image.

FIG. 9 shows a state where both a captured image and a resized image are stored in the DRAM 6. The resizing is performed by, for example, V
Convert the size of the captured image to the size of GA. In this example, if the size of the recorded image is VGA,
In particular, there is no need to perform resizing processing.

In step S23, an optimum threshold is derived based on the color image data of the resized image. Since the color information distributions of the original image and the resized image are relatively the same, the same threshold value can be obtained regardless of which image is used. Since the threshold is derived after resizing to an image of a predetermined size, the same algorithm can always be applied regardless of the size of the selected recording image, and the processing time required for deriving the threshold is also reduced. It can be constant. In addition, since the resized image is stored in an area different from the image captured by the DRAM 6, the captured image is not destroyed. Therefore, the processing after deriving the threshold value needs to be changed from the case where the existing algorithm is used. Absent.

An example of a process of deriving a threshold value used for binarization performed in the image processing block 7 will be described. FIG. 10 is a flowchart of a process for deriving a threshold. In step S1 for starting the derivation of the threshold value, the resized image is attached to the DRAM 6. A histogram of the luminance data of the image taken into the DRAM 6 is created (step S2). The created histogram is stored in a memory of the image processing block 7, for example.

FIG. 11 shows an example of the created histogram. The horizontal axis indicates the level of the luminance data (values from 0 to 255 in the case of 8-bit data), and the vertical axis indicates the number of pixels. When a black character is photographed on a white background, as shown in FIG. 11, a high peak corresponding to the background appears on the white side, and a low peak corresponding to the character appears on the black side. In the image of the characters written in black ink on the blackboard, the height of the peak is opposite to that of FIG.
There is no need to change the following threshold derivation process.

In step S3, a minimum value Ymin and a maximum value Ymax of luminance are obtained from the histogram. In this case, in order to remove noise and invalid pixel data contained in the image, a minimum value Ymin and a maximum value Ymax are obtained after setting a black-side offset value offset1 and a white-side offset value offset2. More specifically, the histogram is traced upward (in FIG. 11, rightward) from (Y = 0), and when the number of pixels exceeds the offset offset1, Ymin
Is calculated, and the histogram is traced downward (leftward in FIG. 11) from (Y = 255), and the number of pixels is offset.
Ymax is obtained when offset2 is exceeded.

Then, in step S4, the difference Ydiff (=
Ymax-Ymin) is obtained, and the difference Ydiff is compared with a preset limit range Yrange. Ydiff> Y
In the case of range, in step S5, the threshold T
hr is calculated by the following equation (1).

Thr = (Ymin + Ymax) / 2 (1) The threshold value can be obtained almost accurately by the equation (1). By adding, a clearer binary image may be obtained. In that case, a term of the adjustment value is added to the equation (1), and is set as the threshold value Thr. Assuming that the range of the adjustment value is ± Adj, the threshold value is calculated as follows: Thr = (Ymin + Ymax) / 2 ± Adj (2)

A histogram is created for each fetched image, the maximum value and the minimum value are obtained again for each image, and the threshold value Thr is determined by equation (2). Therefore, by changing the setting of the camera unit, for example, by adjusting the overall brightness,
Even if the position of the peak of the histogram of the captured image fluctuates, it is possible to derive an optimum threshold value according to the image.

Further, if the obtained difference Ydiff is equal to or smaller than the set range Yrange in step S4, the threshold value is set to a fixed value (step S6). Difference Y diff is Y
As shown in FIG. 12, an image having a size equal to or smaller than the range is estimated to be an image that generates a histogram including only one peak, that is, an image having no characters and only a background. For such an image, it is inappropriate to derive the threshold value by the equation (1) or (2), so a preset fixed value is used as the threshold value. For example, an intermediate value (128) of an 8-bit luminance signal is used as a fixed value.

With the above processing, the threshold derivation processing is completed.
The captured image is binarized by the obtained threshold value Thr (step S7). In the above-described threshold deriving process, since the adjustment value from the offset value and the intermediate value depends on the characteristics of the captured image, it is necessary to set different values depending on the digital image recording device. Therefore, these values can be set to arbitrary values in order to reduce the dependence of the digital image recording apparatus and to provide versatility.

Next, another embodiment of the present invention will be described. In one embodiment described above, the captured image is resized to an image of a fixed size. However, in another embodiment, a thinned image is generated by performing an image thinning process. The thinning-out processing thins out the number of pixels so as not to change the distribution of the color information from the original image. For example, rows (lines) are thinned out at equal intervals in the vertical direction.

When deriving a threshold for binarizing a captured image, it is important to grasp the distribution of color information in the captured image. Therefore,
There is no need to faithfully consider the data of all pixels, and even if the number of rows is appropriately reduced, the threshold can be correctly derived using the histogram, and the processing can be speeded up. Also, by changing the number of rows to be thinned out depending on the subject, a correct threshold value can be derived with a minimum number of data.

A description will be given of still another embodiment of the present invention. In still another embodiment, a frame is set inside a captured image, and a threshold value is derived using only information inside the frame. By this processing, the noise at the peripheral portion of the image and the influence of unnecessary peripheral images other than the characters and the background can be eliminated, and the threshold value can be correctly derived.

That is, when deterioration occurs in a peripheral portion of an image due to a defect in camera signal processing and noise or the like occurs, information including noise or the like is used for deriving a threshold value.
The correct threshold cannot be determined. Further, when a white board or a character document at a remote place is photographed, unnecessary objects around the board other than the white board and the character document are included in the photographed image, and the threshold value cannot be obtained correctly. .

FIG. 13A is an example of an image obtained by photographing a whiteboard. In addition to the whiteboard, unnecessary objects around the whiteboard are photographed. If all the data of the photographed image is used, a correct threshold value cannot be derived, and thus the character of the binarized image becomes unclear as shown in FIG. 13B. In still another embodiment, in FIG. 13A, a frame covering a partial image at the center of the image as shown by a dotted line is set, and a threshold value is derived using only image information in the frame. This makes it possible to correctly derive the threshold value without being affected by the peripheral image, and the binary image becomes clear as shown in FIG. 13C, and the character can be easily read.

The frame indicating the information used to derive the threshold has a predetermined size at the approximate center of the photographed image. However, since there are various kinds of actual captured images, it is more preferable that the position and size of the frame can be set according to the subject. FIGS. 14A and 14B and FIGS. 15A and 15B set a frame (indicated by a dotted line) in accordance with the position and size of a whiteboard on which characters in a photographed image are written.
An example of an image binarized by a threshold value obtained using information in a frame and obtaining the threshold value is shown.

To set the frame, the photographed image is displayed on the display device 4.
, A display indicating the position and size of the frame, for example, a dotted frame is displayed so as to overlap the captured image. The position of the frame can be moved by a direction key of the operation input unit 8, and the size of the frame can be adjusted by a key for image enlargement or reduction. The display of the frame may be something other than the dotted line, and various methods for changing the position and size can be used.
As described above, by making it possible to select a partial image to be used for deriving a threshold value according to a required subject, the threshold value can be obtained with higher accuracy.

The present invention is not limited to the above-described embodiments and the like, and various modifications and applications can be made without departing from the gist of the present invention. For example, by combining the above-described three embodiments, the processing time can be further reduced, and a more accurate derivation of the threshold can be achieved. For example, when resizing to a fixed size, rows may be thinned out. Further, the selection processing of the partial image by the frame and the resizing may be combined. Further, the present invention is not limited to a digital camera and can be applied to other digital image recording apparatuses. For example, the present invention can be applied to a case where a still image recording function is provided as one function of a digital image recording device for recording a moving image, a case where a photographed image is processed by a portable personal computer having a CCD, and the like.

[0077]

According to the present invention, since the threshold value is obtained using the resized image regardless of the size of the image to be recorded, the algorithm for deriving the threshold value can be applied consistently. In addition, it is possible to prevent the processing time from fluctuating.

Further, according to the present invention, since the threshold value is obtained by using the thinned image, the processing can be speeded up without any trouble in deriving the threshold value. By varying the thinning process according to the subject, it is possible to derive a correct threshold value with a minimum number of data.

Further, according to the present invention, instead of using the data of the entire image, the threshold value is derived using the information of the partial image, so that the intended portion is not affected by unnecessary information. It is possible to derive a correct threshold value for optimally binarizing. Also, by allowing the user to arbitrarily set the partial image to be used,
The user can arbitrarily select a part to be binarized.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an overall configuration of an embodiment of the present invention.

FIG. 2 is a schematic diagram illustrating the structure of a GIF file according to an embodiment of the present invention.

FIG. 3 is a schematic diagram used for describing an example of a binarization process of a captured image of a character document or the like and a conversion process to a GIF file.

FIG. 4 is a schematic diagram used for explaining an example of a process of converting a captured image such as a text document into a GIF file.

FIG. 5 is a schematic diagram used to describe a data stream to be compressed in a GIF data block.

FIG. 6 is a schematic diagram used for explaining a binarization process of a photographed image of a character document or the like and a conversion process to a GIF file according to the embodiment of the present invention.

FIG. 7 is a flowchart for explaining existing binarization processing.

FIG. 8 is a flowchart illustrating a binarization process according to an embodiment of the present invention.

FIG. 9 is a schematic diagram schematically showing a state in which a resized image is stored in a DRAM according to an embodiment of the present invention.

FIG. 10 is a flowchart illustrating a threshold deriving process according to an embodiment of the present invention.

FIG. 11 is a schematic diagram illustrating an example of a histogram used for threshold value derivation processing.

FIG. 12 is a schematic diagram of another example of a histogram used for threshold derivation processing.

FIG. 13 is a schematic diagram for explaining still another embodiment of the present invention.

FIG. 14 is a schematic diagram for explaining still another embodiment of the present invention.

FIG. 15 is a schematic diagram for explaining still another embodiment of the present invention.

FIG. 16 is a schematic diagram for explaining a threshold value for binarizing a character document;

[Explanation of symbols]

1 ... CCD, 2 ... Camera block, 3 ... Memory control block, 4 ... Display device, 6 ...
・ DRAM, 7 ・ ・ ・ Image processing block, 8 ・ ・ ・ Operation input unit, 9 ・ ・ ・ Recording media

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) H04N 7/18 G06F 15/68 310A 5C076 9/04 H04N 1/40 103C 5C077 // H04N 101: 00 1 / 46 C 5C079 5/92 HF term (reference) 5B057 AA01 BA02 CA01 CA08 CA16 CB01 CB06 CB16 CD07 CE12 5C022 AA13 AB12 AB22 AC03 AC32 AC42 5C053 FA09 FA27 GA11 GB36 JA16 LA06 5C054 AA05 CA04 CC02 EA01 EA07 EB07 FA05 FC04 FC05 GA04 GB02 GB15 GC01 GD03 5C065 BB02 BB04 BB08 BB11 BB48 CC01 CC09 CC10 DD02 FF02 FF03 GG18 GG30 GG32 GG44 GG49 GG50 5C076 AA22 BA06 BB06 CA02 5C077 LL19 MP08 PP20 PP43 PQ08 RR13P01 RR19 MA17 NA11 NA29 PA00

Claims (14)

[Claims] 1. A digital image recording apparatus for recording an image as a digital signal on a recording medium, comprising: an image capturing means for capturing a color image; an image processing means for converting the captured color image into a binary image; Recording means for recording the output of the image processing means on a recording medium, wherein the image processing means converts the fetched color image into an image of a fixed size or the number of pixels, based on the converted image. The threshold is determined, and 2
A digital image recording device for generating a value image. 2. A digital image recording apparatus for recording an image as a digital signal on a recording medium, comprising: image capturing means for capturing a color image; image processing means for converting the captured color image into a binary image; Recording means for recording the output of the image processing means on a recording medium, wherein the image processing means generates a thinned image by thinning the captured color image, and sets a threshold value based on the thinned image. And a binary image is generated based on the threshold value. 3. A digital image recording apparatus for recording an image as a digital signal on a recording medium, comprising: image capturing means for capturing a color image; image processing means for converting the captured color image into a binary image; Recording means for recording the output of the image processing means on a recording medium, wherein the image processing means stores
A digital image recording apparatus, wherein a threshold value is determined based on a partial image including a subject to be binarized, and a binary image is generated based on the threshold value. 4. A digital image processing apparatus according to claim 1, wherein the captured color image can be converted into a recorded image of a selected size or the number of pixels. 5. The image processing device according to claim 1, wherein the image processing means generates a histogram representing a distribution of the number of pixels of the luminance data in the converted image, and determines a maximum value and a minimum value of the histogram. A digital image recording apparatus, wherein the threshold value is detected and an intermediate value between the maximum value and the minimum value is used as the threshold value. 6. The digital image recording apparatus according to claim 2, wherein the thinning-out processing is processing for thinning out rows. 7. The digital image recording apparatus according to claim 2, wherein the thinning process can be changed depending on a subject. 8. The digital image recording apparatus according to claim 3, wherein a display corresponding to the partial image is performed, and a position and a size of the partial image can be changed. 9. A digital image recording method for recording an image as a digital signal on a recording medium, comprising the steps of: converting a captured color image into a binary image by using a threshold value; and recording the binary image on a recording medium. Converting the captured color image into an image of a fixed size or the number of pixels, determining a threshold value based on the converted image, and generating a binary image based on the threshold value. A digital image recording method. 10. A digital image recording method for recording an image as a digital signal on a recording medium, comprising the steps of: converting a captured color image into a binary image by a threshold value; and recording the binary image on a recording medium. Generating a thinned image by thinning the captured color image, determining a threshold value based on the thinned image, and generating a binary image based on the threshold value. Characteristic digital image recording method. 11. A digital image recording method for recording an image as a digital signal on a recording medium, comprising the steps of: converting a captured color image into a binary image by a threshold value; and recording the binary image on a recording medium. Determining a threshold value in the captured color image based on a partial image including a subject to be binarized, and generating a binary image based on the threshold value. Characteristic digital image recording method. 12. A digital image transmission method for transmitting an image as a digital signal to a communication medium, comprising the steps of: converting a captured color image into a binary image by using a threshold value; and transmitting the binary image to a communication medium. Converting the captured color image into an image of a fixed size or the number of pixels, determining a threshold value based on the converted image, and generating a binary image based on the threshold value. A digital image transmission method. 13. A digital image transmission method for transmitting an image as a digital signal to a communication medium, comprising the steps of: converting a captured color image into a binary image by using a threshold value; and transmitting the binary image to a communication medium. Generating a thinned image by thinning the captured color image, determining a threshold value based on the thinned image, and generating a binary image based on the threshold value. Characteristic digital image transmission method. 14. A digital image transmission method for transmitting an image as a digital signal to a communication medium, comprising the steps of: converting a captured color image into a binary image by using a threshold value; and transmitting the binary image to a communication medium. Determining a threshold value in the captured color image based on a partial image including a subject to be binarized, and generating a binary image based on the threshold value. Characteristic digital image transmission method.