JPH03250866A - Picture reader and facsimile equipment using picture reader - Google Patents

Abstract

PURPOSE:To improve data compression efficiency due to coding regardless of the insertion direction of an original by correcting the arrangement of a read picture data by a tilt of a character string and making the arrangement the same when the character is read under the state that the arrangement direction of the character string of the picture is matched for the scanning direction. CONSTITUTION:A tilt correction section 5 corrects the arrangement of a read picture data before coding by a tilt of a character string detected by a tilt detection section 4 so that the same arrangement as that when the picture data is read under the state that the arrangement direction of the character string of the picture is made matched for the scanning direction. The corrected picture data is coded. Thus, the scanning line for reading is overlapped to the character string included to the picture of read object. That is, even when the reading state where one scanning line is in crossing with plural character strings, the data compression efficiency of the coded picture data is enhanced.

Description

[Detailed description of the invention]

[0001]

[Industrial application field]

The present invention relates to an image reading device that has a function of sequentially scanning and reading a printed image or the like in a fixed direction, encoding the read image data, and compressing the data. The invention further relates to a facsimile machine having such an image reading device. [0002]

[Conventional technology]

In recent years, facsimile machines have played a very important role in the field of data communication. A facsimile machine uses a charge-coupled device (CCD) to capture images printed on paper or handwritten images.
(Device) A scanner consisting of an image sensor etc. reads the data and converts it into a digital signal. This digital signal is
MH (Modi f ie, d Huf fman)
, MR (Modified READ) , M
It is converted into a compressed digital signal by an encoding system called MR (Modified MR). The compressed digital signal is converted to an analog signal and sent out over the telephone line. [0003] FIG. 25 shows the principle of reading a document in a facsimile machine according to the CCITT recommendation. Referring to FIG. 25, a document 10 of, for example, 15OA4 size is read by a CCD sensor or the like along the main scanning direction. 1
The main scanning row is divided into 1728 pixels and converted into digital signals according to their luminance. Main scanning is repeated in a sub-scanning direction orthogonal to the main scanning direction. By completing the scanning in the sub-scanning direction, the entire original 10 has been converted into digital signals. [0004] The image converted into a digital signal is processed by the above-mentioned MHSMRS.
It is compressed according to a method such as MMR. The compressed signal is sent out over the line via a modem and sent to the other party's facsimile machine. [0005] On the receiving side, processing opposite to that on the transmitting side is performed. That is, the facsimile machine on the receiving side converts the received analog signal into a digital signal. The converted digital signal is
It is converted into an uncompressed digital signal by decompression processing, which is the opposite of compression. The original image is obtained by printing the obtained digital signal using the same procedure as when it was read. [0006] Compression/expansion of a document is performed for the purpose of increasing line usage efficiency and reducing communication time by reducing the amount of data to be transmitted. [0007] FIG. 27 and FIG. 28 show an MH code system as an example. The MH code is a -dimensional run-length encoding method. In this method, one line of data consists of a series of variable length codes. Each symbol represents a white or black run length. A white run length is called a white run, and a black run length is called a black run. White runs and black runs occur alternately. All lines begin with a white run symbol to ensure that the receiver is motivated by the signal. If the actual scan line begins with a black run, a code indicating a white run of length O is sent first. The MH code includes two types of codes: a terminating code and a makeup code. [0008] FIG. 27 shows a terminating code. The terminating code represents a run length from 0 to 63 pixels. To increase compression efficiency, run lengths that frequently appear in image information are assigned short codes. [0009] FIG. 28 shows a makeup code. The makeup code is combined with the terminating code to represent run lengths of 64 to 1728 pixels. Run lengths from 64 to 1728 pixels are first encoded with a make-up code representing a run length equal to or shorter than the run length. This is followed by a terminating code representing the difference between the actual run length and the run length represented by the makeup code. [00103] The receiver can obtain uncompressed data by comparing the received data with a decoding table prepared in advance. [0011] In this way, by compressing transmission data using codes, images can be transmitted efficiently by facsimile machines.
It can be done in a short time. [0012] However, facsimile machines have the following problems. As shown in FIG. 25, if the document 10 is correctly supplied to the apparatus and the main scanning direction and the direction in which the character string 12 extends match, compression by encoding is efficiently performed. [0013] However, as shown in FIG. 26, when the original 10 is fed to the apparatus at an angle and the main scanning direction does not match the direction of the character string 12, the compression efficiency decreases. This is due to the following reasons. [0014] As mentioned above, coding systems assign short codes to run lengths that occur frequently in image information and relatively long codes to run lengths that occur only infrequently. Therefore, as shown in FIG. 26, when the direction of the main scanning line 14 read in one main scan and the direction of the character string 12 do not match, the frequency of appearance of run lengths becomes irregular, which reduces the efficiency of data compression. becomes worse. [0015] Furthermore, as shown in FIG. 25, consider a case where character strings are arranged at intervals of dO. If the document 10 is fed to the facsimile machine in the correct direction, this distance do
The main scanning lines existing between O include only white runs. Therefore, the above-mentioned makeup code can be used to code this part of the image with very high efficiency. On the other hand, as shown in FIG. 26, when the document 10 is supplied to the facsimile machine in an inclined manner, the main scanning line consisting of only white runs is limited to the interval d1. As is clear from the figure, the interval d1 is much shorter than the interval do. Therefore, compared to the case shown in FIG.
The compression efficiency in the case shown is quite low. [0016] In order to solve this problem, for example, Japanese Patent Application Laid-Open Nos. 55-154871, 62-206962, and 63-88963 detect this slope before encoding. Facsimile machines and the like that are equipped with a tilt correction function have been proposed. [0017] JP-A-55-154871 proposes a method for reading a document in a facsimile machine using a sheet of paper with special marks as a document. By detecting this mark when reading a document, it is possible to know the skew of the document. Coordinate transformation is applied to the data so as to offset this inclination. This allows for improved encoding efficiency. [0018] JP-A No. 62-206962 proposes the use of a sensor that is provided in a document reading section of a facsimile machine and detects the passing time of the leading edge of a supplied document. By detecting the time difference between two places on the leading edge of a document fed at a predetermined speed and passing in front of a predetermined sensor, it is possible to know the spatial relationship between these two places. The size of the supplied document can also be known at the same time. From this information, it is possible to know the skew of the document. By performing a coordinate transformation on the data to cancel the known slope,
Data compression efficiency can be improved. [0019] Japanese Patent Application Laid-Open No. 63-88963 uses a method of detecting the reading time of one end of a document along the sub-scanning direction based on fluctuations in signal level when reading the document, for reading the document by a facsimile machine. I am proposing that. During main scanning, writing of main scanning line data to the image memory is started based on the time when the edge of the paper is detected. By using this method, one edge of the document is stored in the image memory.
It is always stored parallel to the edge of the image memory. The image itself also has the original skew removed, improving compression efficiency. [0020]

[Problem to be solved by the invention]

However, even with the above-mentioned conventional technology, the following problems remain. For example, consider a language like Japanese, where characters can be written both vertically and horizontally. In such a language, as shown in FIG. 29, there is a case where a document written vertically on an A4 size sheet is fed to a document reading device of a facsimile machine along the vertical direction of the sheet. In such a case, even if the inclination of the document itself is detected and corrected, the compression efficiency cannot be improved in principle. [0021]Also, even if the document is written horizontally, a similar problem occurs when the document is fed to the reading device along the direction in which the character strings extend. [0022] Therefore, an object of the present invention is to provide an image reading device and a facsimile device that can perform data compression through encoding with higher efficiency regardless of the direction in which a document is inserted. [0023]

[Means to solve the problem]

The image reading device according to the invention according to claim 1 is a device that scans and reads an image in a fixed direction and encodes the read image data, and the image reading device scans and reads an image in a fixed direction and encodes the read image data. a tilt detection means for extracting the part of the character row that is located in the area and detecting the inclination of the character row with respect to the scanning direction; and a state in which the read image data array for one image is made such that the arrangement direction of the character row matches the scanning direction. a correction means for correcting the inclination of the character string detected by the inclination detection means so that the arrangement is the same as the arrangement of the image data read under means. [0024] The image reading device according to the invention described in claim 2 is a device that scans and reads an image in a certain direction and encodes the read image data, and the image reading device scans and reads an image in a certain direction and encodes the read image data. A vertical/horizontal determining means extracts a part of a character string arranged in a row, and determines whether the character string is arranged in a horizontal direction matching the scanning direction or a vertical direction perpendicular to the scanning direction; When it is determined that the orientation is vertical, the array of image data for one read image is made the same as the array of image data that is read with the character string array direction aligned with the scanning direction. A correction means corrects the rotation by 90 degrees so that the image data is rotated by 90 degrees, and when the vertical/horizontal determining means determines that the orientation is vertical, the correcting means encodes the corrected image data, and the vertical/horizontal determining means determines that the orientation is horizontal. and encoding means for encoding the image data as read without being corrected by the correction means. [0025] The facsimile device according to the invention recited in claim 3 divides and reads an image on a document into a plurality of pixels arranged two-dimensionally, thereby determining the brightness of each pixel for each pixel. and pixel signal storage means each including a plurality of storage cells for storing a pixel signal in logical association with one of the pixels. The plurality of memory cells are logically two-dimensionally arranged in a predetermined first direction and a second direction intersecting the first direction, corresponding to the arrangement of the plurality of pixels. . The facsimile apparatus according to the present invention further includes an arrangement of a plurality of storage cells, based on the pixel signals stored in the pixel signal storage means, for storing character strings included in the image on the document expressed by the stored pixel signals. a character arrangement direction detecting means for detecting a character arrangement direction indicating an arrangement direction in the character arrangement direction; and a character arrangement direction for determining which of a first direction and a second direction the detected character arrangement direction is closer to. direction determining means; and in response to the output of the character arrangement direction determining means, when the character arrangement direction is closer to the second direction than the first direction, the image on the document is expressed by the stored pixel signals. image rotation means for rotating the image on the document represented by the stored pixel signals by a predetermined angle so that the character arrangement direction is closer to the first direction; Further, a signal compression means reads the stored pixel signal from the pixel signal storage means along the first direction and compresses it using a compression code, and the pixel signal compressed by the signal compression means is transmitted via a communication line. and transmitting means for transmitting the information to another station. [0026]

[Effect]

According to the image reading device according to the first aspect, the arrangement of the read image data is corrected by the correction means by the inclination of the character row detected by the inclination detection means, so that the arrangement direction of the character row of the image is corrected. The array is the same as when read under conditions that match the scanning direction. [0027] According to the invention described in claim 2, when the arrangement direction of the character strings of the image to be read is the vertical arrangement direction perpendicular to the scanning direction, the arrangement of the read image data is corrected by 90 degree rotation. The characters are arranged in the same arrangement as when read in the horizontal arrangement direction where the arrangement direction of the character strings coincides with the scanning direction. [0028] In the facsimile apparatus according to the third aspect of the invention, the read image on the document is converted into pixel signals and stored in the array of memory cells. The direction in which the character string of the document extends in the array of memory cells is detected;
Furthermore, the character arrangement direction is the first and second direction of the memory cell arrangement.
It is determined which of the directions is closer. If it is determined that the character arrangement direction is closer to the second direction, the image rotation means rotates the image by a predetermined angle so that the character arrangement direction becomes closer to the first direction. The signal compression means reads and encodes pixel signals from the array of storage cells along the first direction. Since the blank lines between character strings also extend along the same direction as the character arrangement direction, by rotating the image as described above, the average length of the white run and black run along the first direction is , the number of white runs and black runs is longer and fewer than when the image is not rotated. [0029]

【Example】

FIG. 1 is a block diagram showing a schematic configuration of an image reading device that is an embodiment of the invention. The scanner 1 is a device that sequentially scans and reads an image to be read, such as a printed image, in a fixed direction, and the binarization unit 2 is a circuit unit that performs binarization processing on the image data read by the scanner 1. It is. The binarized image data is -
The image is then stored in the image memory 3. [00303 The tilt detection unit 4 detects a portion of the image that corresponds to a character string from the image data for one image stored in the image memory 3, and further detects the tilt of the character string (the arrangement of the character string with respect to the scanning direction). This is a circuit unit for detecting directional inclination (hereinafter referred to as skew). [0031] The skew correction 5 corrects (rotates) the array of binarized image data in the image memory 3 by the amount of skew detected by the skew detector so that the array direction of the character strings matches the scanning direction. This is a circuit section for making the arrangement of image data equal to the arrangement of image data when reading is performed under the same condition. The corrected image data is then stored in the image memory 3. [00321] The encoding unit 6 is a circuit unit for performing encoding processing such as the above-mentioned MH, MR, MMR, etc. on the corrected image data. The encoded image data is either stored in the -H image memory 3 or directly transmitted to another facsimile machine via a transmission mechanism (not shown). [0033] The control unit 7 is a circuit unit that controls the entire image reading device, including the scanner 1, the binarization unit 2, the image memory 3, the tilt detection unit 4, the tilt correction unit 5, and the encoding unit 6. 0034
] FIG. 2 is a flowchart showing the general operation of this image reading device. The operation when the image reading device reads an image including a character string is as follows. First, the scanner 1 sequentially scans a print image to be read in a fixed direction and reads image data for one image. The binarization unit 2 performs binarization processing on the read image data. The rM image memory 3 stores binarized image data. [0035] The tilt detection section 4 extracts a data portion corresponding to a character display area from the image data for one image stored in the image memory 3. Then, from among the data portions, extracts a data portion corresponding to a character string. Extract step n2)
Furthermore, the inclination angle of the character string with respect to the image scanning direction by the scanner 1, that is, the skew of the character string, is detected from the data portion of the character string (step n3). [0036] The tilt correction section 5 rotationally corrects the array of accumulated image data for one image in the image memory 3 by the amount of skew detected by the tilt detection section 4 (step n4). As a result, the stored image data in the image memory 3 is stored in the same arrangement as the image data when reading is performed with the character string arrangement direction aligned with the scanning direction. The encoding unit 6 performs encoding processing such as MHSMRSMMR on the image data that has been subjected to the correction processing. (Step n5). [0037] The encoded image data is transmitted as it is to, for example, another facsimile machine via a transmission mechanism (not shown), or
Alternatively, it is stored in the image memory 3. By correcting the slope in this way, compression efficiency during encoding is improved. [0038] FIG. 4 is a block diagram showing a schematic configuration of an image reading device that is an embodiment of the invention as set forth in claim 2. The scanner 1, binarization unit 2, and image memory 3 shown in FIG. 4 are the same as those shown in FIG. Therefore, detailed explanations about them will not be repeated here. [0039] The vertical/horizontal detection unit 104 detects the 1
The part of the image corresponding to the character area is detected from the image data for the image, and the character arrangement direction of the character area is determined to be horizontal (alignment direction that matches the scanning direction) or vertical alignment direction (alignment direction that matches the scanning direction). This is a circuit unit that determines and detects whether the array direction is orthogonal to the array direction. [0040] When the vertical/horizontal detection unit 104 detects the vertical alignment direction, the rotation processing unit 105 corrects the rotation of the binarized image data array in the image memory 3 by 90 degrees, and rotates the character. This is a circuit unit for making the arrangement of image data equal to the arrangement of image data when reading is performed with the arrangement direction set to the horizontal direction. The corrected image data is stored in the -1 image memory 3. [0041] When the vertical/horizontal detection section 104 detects the vertical arrangement direction, the encoding section 106 also detects the horizontal alignment direction for the image data rotationally corrected by the rotation processing section 105. This is a circuit unit for performing encoding processing on the image data that has been binarized (not subjected to rotation correction processing) when the arrangement direction is detected. The encoded image data is -1
The data is either stored in the image memory 3 or transmitted directly to another facsimile machine via a transmission mechanism (not shown). The control unit 107 is a circuit unit for controlling the entire image reading apparatus, including the scanner 1, the binarization unit 2, the image memory 3, the vertical/horizontal detection unit 104, the rotation processing unit 105, and the encoding unit 106. be. [0042] The operation of the image reading device when reading a document in which ξ characters are written horizontally or a document in which characters are written vertically as an image will be described below. In the following, a row or column of characters will be referred to as a "character string" in both horizontal and vertical writing. [0043] Here, to simplify the explanation, a horizontally written document is read with the direction of its character strings set to match the scanning direction, and a vertically written document is read with the direction of its character strings set to match the scanning direction. It is assumed that reading is performed with the setting perpendicular to the . [0044] First, the scanner 1 sequentially scans a document to be read, such as a horizontally written document or a vertically written document, in a fixed direction to read image data for one image. The binarization unit 2 performs binarization processing on the read image data. Image memory 3
stores the binarized image data. [0045] The vertical/horizontal detection unit 104 detects 1 stored in the image memory 3.
Extract the data portion corresponding to the character area from the image data of the image (step n11) Extract the data portion corresponding to the character string from the image data portion (step n12) It is determined whether the image scanning direction coincides with or perpendicular to the image scanning direction according to No. 1, that is, whether the document is horizontally written or vertically written, and the result is detected (step n13). [0046] When the vertical/horizontal detection unit 104 detects a horizontally written document, that is, when it is determined that the character arrangement direction is the horizontal arrangement direction that matches the scanning direction, the encoding unit 106 reads and corrects the text. Encoding processing is performed on image data that has not been encoded (step n14). [0047] When the vertical/horizontal detection unit 104 detects a vertically written document, that is, when it is determined that the character arrangement direction is the vertical arrangement direction perpendicular to the scanning direction, the rotation processing unit 105 uses the image memory 3
The array of accumulated image data for one image in is rotated and corrected by 90 degrees (step n15). As a result, the image data stored in the image memory 3 is stored in the same arrangement as the image data when it is read with the horizontal arrangement direction, that is, the character arrangement direction aligned with the scanning direction. become. [0048] The encoding unit 106 performs encoding processing such as MH, MR, and MMR on the rotationally corrected image data (step n14). [0049] The encoded image data may be directly transmitted to another facsimile machine via a transmission mechanism (not shown), or may be stored in the image memory 3. [0050] By doing this, when a document is read in a vertical arrangement direction in which the character arrangement direction is orthogonal to the scanning direction, data compression efficiency is improved in encoding the read image data. It will increase significantly. [0051] When the above-described image reading device is used for facsimile, sub-data for recognizing whether the encoded image data is a horizontally written document or a vertically written document is added to the encoded image. It can be added to data and sent. A facsimile machine that receives this image data may refer to this sub-data when decoding the image data, perform 90 degree reverse rotation correction as necessary, and reproduce the image. [0052]Furthermore, even if the image reading device is an image file device that does not transmit image data, the determination result of the vertical/horizontal detection unit 104 can be stored together with the encoded image data. When decoding the image data, if it is identified as image data of a vertical document,
If necessary, the image can be reproduced by correcting the 90-degree reverse rotation. [0053] FIG. 6 is a block diagram showing a schematic configuration of an image reading device according to another embodiment of the invention claimed in claim 2. [0054] The image reading device of this embodiment not only determines whether the character arrangement direction is horizontal or vertical, but also determines whether the character arrangement direction is horizontal or vertical with respect to the scanning direction. This embodiment differs from the previous embodiment in that the present invention detects whether the image data has a tilt angle of 100 degrees, and also performs correction for the tilt angle before encoding the read image data. In FIG. 6, the tilt angle detection unit 108 is a circuit unit for detecting the tilt angle of the character arrangement direction with respect to the scanning direction from the read image data. [0055] The rotation processing unit 115 is similar to the rotation processing unit 105 (
Unlike Fig. 4), the function rotates and corrects the arrangement of image data for one read image by the above-mentioned tilt angle, not only when characters are arranged in the vertical direction but also in the horizontal direction. It is a circuit section with [0056] Regarding the other circuit units, that is, the scanner 1, the binarization unit 2, the image memory 3, the vertical/horizontal checking unit 104, the encoding unit 106, and the control unit 107, the previous implementation shown in FIG. Same as example. Therefore, detailed explanations about them will not be repeated here. [0057] FIG. 7 is a flowchart showing a document reading operation by the image reading device of this embodiment. FIG. 8 is a diagram for explaining the relationship between the arrangement direction of characters of a document and the scanning direction in the reading operation. [0058] The image reading device of this embodiment operates as follows during reading. A scanner 1 sequentially scans a document to be read, such as a horizontally written document or a vertically written document, in a fixed direction and reads image data for one image. The binarization unit 2 performs binarization processing on the read image data. #J
The image memory 3 stores binarized image data. [0059] The vertical/horizontal detection unit 104 detects 1 stored in the image memory 3.
A data portion corresponding to a character area is extracted from the image data for the image (step n21). Next, the vertical/horizontal detection section 1
04 extracts a data part corresponding to a character string from the data part, and determines from the data part of the character string whether the character arrangement direction is almost the same as the image scanning direction by the scanner 1 or almost orthogonal, that is, the document is written horizontally. It is determined whether the document is vertically written or not, and the result is detected (step n22). [00601 Regardless of whether the vertical/horizontal detecting unit 104 detects a horizontally written document or a vertically written document, the tilt detecting unit 108 detects the tilt angle of the arrangement direction with respect to the scanning direction (step n
23). [0061] As shown in FIG. 8(1), when the document to be read is a horizontally written document A1 containing character lines r, and the arrangement direction of the character lines r has a slight inclination (angle α) to the scanning direction S, That is, when the character arrangement direction is the horizontal arrangement direction, the angle α detected by the inclination angle detection unit 108 has a value of -45 degrees or more and less than 45 degrees. As shown in FIG. 8(2), the document to be read is a vertically written document A2 containing a character string C, and the arrangement direction of the character string C is slightly tilted (angle β) from the direction perpendicular to the scanning direction S. In other words, when the character arrangement direction is the vertical arrangement direction, the angle β detected by the inclination angle detection unit 108 has a value of 45 degrees or more and less than 135 degrees. [0062] The rotation processing unit 115 rotates and corrects the array of accumulated image data for one image in the image memory 3 by the tilt angles α and β detected by the tilt angle detection unit 108 (step n24).
. As a result, the image data stored in the image memory is stored in the same arrangement as the image data when read with the character arrangement direction aligned with the scanning direction. [0063] The encoding unit 106 performs encoding processing on the image data that has been subjected to the correction processing. The encoded image data is either directly transmitted to another facsimile machine via a transmission mechanism (not shown), or is stored in the image memory 3. [0064] In this way, not only is the image data when the characters are arranged in the vertical direction corrected by rotation so that the image data is read in the same direction as in the horizontal direction, but also when the characters are arranged in the horizontal direction, the image data is read in the vertical direction. In the case of direction as well, encoding is performed after rotationally correcting the array of accumulated image data by the inclination angle with respect to the scanning direction. Therefore, data compression efficiency does not deteriorate due to overlap where scanning lines intersect character strings. [0065] In this embodiment, as in the apparatus shown in FIG.
Sub-data for identifying whether the encoded image data is a horizontally written document or a vertically written document can be added to the image data. Thereby, the image data can be corrected by 90 degrees of reverse rotation as necessary during playback. [0066] FIG. 9 is a circuit block diagram of a facsimile apparatus according to an embodiment of the invention set forth in claim 3. Referring to FIG. 9,
This facsimile device includes a control circuit 16 for controlling the entire device, an operation panel 18 connected to the control circuit 16, and a control panel 18 for supplying a document to be transmitted for reading, which is controlled by the control circuit 16. The original feed mechanism 24 and the original feed mechanism 2
a fluorescent lamp 22 for illuminating the document conveyed by 4 for reading; a lighting circuit 20 controlled by the control circuit 16 to control lighting of the fluorescent lamp 22;
A reading mechanism 28 receives light rays emitted by the fluorescent lamp 22 and reflected by the original and reads the original by photoelectric conversion, and a printing mechanism connected to the control circuit 16 and prints the received original onto recording paper 44. A line interface (LIU) for interfacing the facsimile machine 26 with a telephone line.
e Unit) 32, and a modem 30 for mutually converting analog signals transmitted over the line and digital signals handled within the control circuit 16. [0067] The control circuit 16 is a sensor circuit 6 for controlling the sensors provided in each part and processing signals from the sensors.
4, key scanning of the operation panel 18 allows you to monitor the human power of various keys, and the LEDs on the operation panel 18 (
a panel controller 66 for controlling lighting of a light emitting diode;
A mechanics/recording control circuit 68 for controlling the motor that feeds the original or the recording paper 44, the print head and printing pressure of the printing mechanism 26, and analog pixels obtained by the reading mechanism 28. The signal is converted to AGC (Au
t. A reading processing circuit 70 for controlling the signal to a constant level through processing (Galn Control) and correcting optical distortion caused by the optical system included in the reading mechanism 28 and signal distortion due to variations in sensitivity of the CCD image sensor. , and a transmission control circuit 72 for controlling data transmission processing by the modem 30. [0068] This facsimile device is further connected to a sensor circuit 64, a panel controller 66, a mechanics/recording control circuit 68, a transmission control circuit 72, and a modem 30, and transmits control procedures and codes for each mode performed in each of these circuits. a microcomputer (abbreviated as "microcomputer") 62 for controlling encoding/decoding processing, operation in self-diagnosis mode, etc.; and a page memory for storing digital signals of read images output by the reading processing circuit 70. 7
6 and is stored in the page memory 76 and is stored in the microcomputer 62.
page memories 78 and 80 for storing digital image signals that have been corrected as a feature of the present invention;
The page memory 76 is connected to the microcomputer 62, the mechanics/recording control circuit 68, the reading processing circuit 70, and the page memory 76.78.8o, and is controlled by the microcomputer 62.
．． 78 and 80, a page memory control circuit 74 for controlling image signal reading from there, and encoding/decoding processing. [0069] The document feed mechanism 24 includes a feeding roller 34 for feeding documents one by one for reading, a rubber plate 38 for separating documents, and a transport roller 36 driven by a pulse motor (not shown) to feed the documents. A document sensor 40 is provided between the feeding roller 34 and the rubber plate 38 to sandwich the plane on which the document is conveyed, and detects that the document has been set and provides a signal to that effect to the sensor circuit 64. and the transport roller 36, and are provided immediately before the position where the document is illuminated by the fluorescent lamp 22, and are for detecting both ends of the leading edge of the document being fed and providing a signal to that effect to the sensor circuit 64. document edge sensor 42. [0070] The printing mechanism 26 includes a paper-out sensor 48 for detecting paper-out of the recording paper 44, and a mechanics/recording control circuit 6.
8 and scans the recording paper 44 according to a signal given from the mechanics/recording control circuit 68.
Thermal head 5 for forming an image on recording paper 44
0 and a thermal head 50 to sandwich and hold the recording paper 44, and a pulse motor 46 for rotating the roller 52 and pulling out the recording paper 44. [0071] The reading mechanism 28 includes a mirror 56 for reflecting light emitted from the fluorescent lamp 22 and reflected on the surface of the document and guiding it to a predetermined optical path, and a mirror 56 provided on the optical path of the light reflected by the mirror 56. a lens 58 for condensing light rays from the original to form an image of the original, and a lens 58 for converting the image of the original formed by the lens 58 into an electrical signal by photoelectric conversion and providing it to the reading processing circuit 70. 2048-bit CCD60 and
It includes a motor 54 for moving the lens 58 and CCD 60 to predetermined positions according to the magnification. [0072] Referring to FIG. 10, the document edge sensor 42 shown in FIG. 9 includes a sensor 42a for detecting the left end of the leading edge of the document 10 supplied to the sensor 42, and a
They are placed the same distance apart as the width of the S size paper.
A sensor 42b for detecting the right edge of the front edge of B5 size paper, and a sensor 42c provided at a distance equal to the width of the 15OA4 size paper from sensor 42a and for detecting the right edge of A4 size paper. including. [0Q73] The facsimile device according to the present invention is characterized by a microcomputer 6
2, the program executed in page memory 76 not only corrects the skew of the document, but also detects the direction of the character strings in the document so as to achieve the highest compression efficiency.
It has a function of converting an image signal read using 80 to 80. [0074] Referring to FIGS. 9 to 15, the facsimile apparatus according to the present invention operates as follows. First, the principle of operation of a facsimile apparatus when reading a document, to which the present invention is primarily concerned, will be explained. [0075] The operator sets the document to be transmitted on a document table (not shown) provided in contact with the feeding roller 34. Sensor 40 detects that a document has been set, and provides a signal to that effect to sensor circuit 64. Furthermore, the operator presses a start button (not shown) on the operation panel 18. Since the panel controller 66 scans each key on the operation panel 18 as described above, it detects that the start button has been pressed by the operator. The panel controller 66 gives a signal to that effect to the microcomputer 62. [0076] The microcomputer 62 receives a signal from the sensor circuit 64 indicating that a transmission document has been set, and a signal indicating that the start button has been pressed from the panel controller 66, and operates as follows. The microcomputer 62 first controls the page memory control circuit 74 so that the page memories 76 to 80 can write images to the page memories 76 to 80.
Enable 0. The microcomputer 62 then controls the mechanics/recording control circuit 68 to operate a pulse motor (not shown) for conveying the document. The conveyance roller 36 is rotated by this pulse motor, and the document is conveyed along the sub-scanning direction. At this time, the separated documents are conveyed by the rubber plate 38 for reading. [0077] The transported document passes in front of the sensor 42 before being read by the CCD 60. Referring to FIG. 10, it is assumed that the leading edge of document 10 is inclined by an angle θ with respect to the main scanning direction (perpendicular to the document conveyance direction). Further, it is assumed that this document 10 is A4 size. Due to the inclination θ, the left end and right end of the leading edge of the document 10 are deviated by a distance D1 in the sub-scanning direction. [0078] The conveyance speed of the original 10 is detected by the circuit 68,
The microcomputer 62 can know the speed in advance. The width of the document 10 is also known from the sensor output. therefore,
By knowing the time from when the sensor 42a detects the passage of the left end of the leading edge of the original 10 to when the sensor 42C detects the passage of the right end of the leading edge of the original 10, the inclination θ can be easily determined. . [0079] If the paper 10 is B5 size, the sensor 42
a, the inclination θ can be known from the signal from the sensor 42b. [0080] The signals of sensors 42a-c are provided to sensor circuit 64. The sensor circuit 64 provides the output results of each sensor 42a to 42c to the microcomputer 62. The microcomputer 62 calculates the slope θ according to a predetermined calculation. [0081] The microcomputer 62 controls the lighting circuit 20 to turn on the fluorescent lamp 22.
lights up. The light beam emitted from the fluorescent lamp 22 is reflected on the surface of the document and reaches the mirror 56. The light beam reflected by the mirror 56 forms an image of the original on the light receiving surface of the CCD 60 by the lens 58. At this time, the positions of the lens 58 and CCD 60 are moved by the motor 54 according to the size of the document, and the magnification of the image is adjusted. The CCD 60 electrically scans the image of the original in the main scanning direction, outputs an analog signal corresponding to the brightness on the surface of the original, and supplies it to the reading processing circuit 70 . [0082] The reading processing circuit 70 makes the level of the analog signal given from the CCD 60 constant through AGC processing,
Further, optical system distortion contained in the signal output from the CCD and distortion due to the influence of sensitivity variations of the CCD are corrected, and the corrected signal is applied to the page memory control circuit 74. [0083] At the same time that the sensor 42a shown in FIG. 10 detects passage of the left edge of the leading edge of the document, the page memory control circuit 74
Writing of the output of the read processing circuit 70 to the page memory 76 is started. [0084] Writing to the page memory 76 is performed as follows. Referring to FIG. 11, page memory 76 has an image storage area 82 consisting of the X direction and memory cells arranged two-dimensionally in the X direction. The page memory 76 sequentially stores signals obtained by one main scan in one row in the X direction of memory cells at fixed addresses in the X direction. The address in the X direction is sequentially incremented along with the sub-scanning. Therefore, in the image storage area 82,
Original image 84 as a collection of main scanning lines in the sub-scanning direction
is formed. [0085] With reference to FIGS. 9 and 10, sensor 42a is connected to document 10.
After the left end of the leading edge of the document is detected until the same part is detected by the CCD 60, there is a difference between the position of the sensor 42 and the position of the document when it is actually read by the CCD 60. , - a constant time difference occurs. Therefore, Figure 1
1, the leftmost edge of the leading edge of the document image 84 is stored at an address X1 in the X direction and an address Y1 in the X direction in the image storage area 82 of the page memory 76. Thereafter, the image of this original document is stored in the page memory 76 in the form shown in FIG. [0086] Referring to FIG. 11, it is assumed that the image storage area 82 of the page memory 76 has an address Xo-Xn in the X direction and an address yo-yn in the Y direction. It is assumed that the left end of the leading edge of the original image 84 stored in the page memory 76 has an address (Xl, yl), and the right end of the leading edge of the original image 84 has an address (X 2, y2 ). The difference between address (Xl, Yl) and address (X2, Y2) is due to the inclination θ of the original as described above. [0087] The inclination θ of the document can be easily determined from the outputs of the sensors 42a and 42c. The microcomputer 62 converts the coordinates of the original image 84 using the obtained inclination θ, and
The converted data is written to the page memory 78. This coordinate transformation uses the method disclosed in US Pat. It is done. [0088] As a result of the coordinate transformation, the data stored in the page memory 78 is parallel to the corresponding side of each of the four printed image storage areas 82 of the original, as shown in FIG. As a result, when a document is supplied as shown in FIG. 11, the character arrangement direction and the y direction of the memory cell arrangement of the page memory match. [0089] By converting the image of the original in this way, if this image signal is read along the y direction of the page memory address, the signal will be read under the same conditions as when the original is inserted at an angle of O. It becomes possible to perform compression of . Therefore, the compression efficiency of the image signal is significantly improved compared to the case where the image signal as shown in FIG. 11 is compressed. [0090] As shown in FIG. 10, when the original 10 is fed tilted so that the direction of the character string 12 on the original 10 deviates from the main scanning direction of the image by a small angle θ, the above-described By such coordinate transformation, it is possible to immediately obtain a corrected image signal with the highest compression efficiency. However, documents are not always fed into a facsimile machine in this orientation. [0091] For example, with reference to FIG. 13, as in Japanese,
When feeding a vertically written original to a facsimile machine as is, or when feeding a horizontally written original (Japanese or English) to a facsimile machine after rotating it approximately 90 degrees, the above correction alone will not be enough to correct the image signal. Compression efficiency is not improved. That is, the sensor 42a shown in FIG.
, 42c, the corrected image signal will only be as shown in FIG. 14. [0092] Referring to FIG. 14, the manuscript in which characters are arranged vertically is
If the above-described correction is only performed, the direction of the character string and the y direction of the image storage area 82 will not match. Even if this image is read out along the y direction and compressed,
Compression efficiency does not improve. [0093] Therefore, the program executed in the microcomputer 62 of the facsimile apparatus according to the present invention performs coordinate conversion of the image based on the inclination θ of the original obtained by the sensors 42a and 42c as described above. It is characterized by detecting the direction of the character string. Detection of the direction of a character string is performed according to the following procedure. [0094] First, the microcomputer 62 examines the document image stored in the page memory 76 and calculates the ratio of white dots to black dots in each part of the document image. Area where only character strings are arranged (hereinafter referred to as "character array area")
It is known that there is a characteristic difference between the ratio of white points to black points in a region where an image other than characters is formed. Therefore, by examining the distribution of white dots and black dots in the original image 84, it is possible to distinguish between a character array area where a line image 86 containing only characters is formed and an image area 88 where an image other than characters is formed. can do. [0095] It is generally considered that the compression efficiency of the image does not change when the image region 88 is read along any direction. On the other hand, as described above, when encoding an area where only the row images 86 are formed, if the document is read and encoded along the direction in which the row images 86 extend, the signal will be lower than in other cases. It is known that compression efficiency is significantly improved. Therefore, after detecting the character array area as described above, if the direction of the line image 86 included in the character array area is detected and the image signal is encoded along that direction, the code can be compressed. Efficiency is expected to improve. [0096] Once the character arrangement area is specified, the direction in which the line image 86 as shown in FIG. 11 extends can be easily determined. For example, Japanese Patent Laid-Open No. 2-215269 discloses a scanner that determines the effective image width in the main scanning direction based on the continuity of black bits in the main scanning direction and the number of black bits in the sub-scanning direction. The device is disclosed. Using the technique disclosed in this publication, it is possible to know the general direction in which the line image 86 of the original image 84 extends. Let θ1 be the angle formed between the arrangement direction of this character string and the y-axis. [0097] Referring to FIG. 15, if the angle θ1 is between 145 degrees and 45 degrees, it is determined that the original is supplied to the facsimile machine in the state shown in FIG. . In this case, it is sufficient to correct the image by correcting the angle θ detected by the sensors 42a and 42c.

On the other hand, if the angle θ1 is between 45 degrees and 135 degrees, it is considered that the document was supplied to the facsimile machine in the state shown in FIG. In this case, as described above, before the angle θ is corrected by the sensors 42a and 42c, the document image 84 is rotated 90 degrees around a predetermined point in the image storage area 82. The rotated image signal is stored in page memory 80. The document image after rotation by 90 degrees is similar to that shown in FIG. [0099] The microcomputer 62 sends the sensor 42a to the original image 84 converted into the form shown in FIG.
, 42c is performed to correct the angle θ detected by the angle θ. The resulting image is as shown in FIG. 12. By sequentially reading and encoding these image images, it is possible to compress the image signal with high efficiency. [0100] In the above description, the page memory 76 for initially storing the read image signal, and the case where the angle θ1 between the direction of the character string and the main scanning direction is between 45° and 135°. , a page memory 80 for storing image signals rotated by 90 degrees, and a page memory 78 for storing finally corrected image signals. However, the present invention is not limited to this, and it is also possible to perform the above-described correction processing by dividing one page memory area, for example. It is also conceivable to use the same page memory for different purposes. [0101] FIGS. 16 to 23 are schematic flowcharts of programs executed in the microcomputer 62 of the facsimile machine of the above-described embodiment. The following flowchart shows a program having a structure in which processing is divided into a plurality of subroutines and processing is performed by calling the subroutines from the main routine. However, the structure of the program is not limited to this, and the same effect can be obtained by using a program in which all subroutines are incorporated into the main routine as they are, for example. [0102] Referring to FIG. 16, the main routine of this program has the following structure. In step 5OO1, the sensor 40 (FIG. 9) detects that the original to be sent has been set.
Detected by. [0103] Subsequently, in step 5OO2, necessary initial processing is performed. Control proceeds to step 5OO3. [0104] In step 5003, the sensor 42 (FIG. 9), that is, the sensors 42a and 42c (FIG. 10), detects the angle θ that the leading edge of the document makes with the main scanning direction, that is, the inclination θ of the document. Control proceeds to step 5OO4. [0105] In step 5OO4, image information of all originals set on an unillustrated original table is read and stored in page memory 76 (FIG. 9). Control is step 5O
Proceed to O5. [0106] In step 5005, the location of the character arrangement area in the document image stored in the page memory 76 is detected. Control proceeds to step 5OO6. [0107] In step 5OO6, it is detected in which direction the characters are arranged in the detected character arrangement area. As a result, the angle θ1 between the character arrangement direction and the main scanning direction is determined. Control then proceeds to step 5OO
Proceed to step 7. [0108] In step 5OO7, coordinate transformation processing is performed to correct the image tilt based on the document tilt θ determined by the sensors 42a and 42c. When the angle θ1 is between 145 degrees and 45 degrees, the correction result is written to the page memory 78. Angle θ1 is 4
If the angle is between 5 degrees and 135 degrees, the corrected image signal is written to the page memory 80. Control proceeds to step 5OO8. [0109] In step 5OO8, the following processing is performed based on the angle θ1 detected in step 5OO6. If the angle θ1 is between 145 degrees and 45 degrees, the corrected image signal has already been stored in the page memory 78. Therefore, no further correction processing is performed. On the other hand, if the angle θ1 is between 45 degrees and 135 degrees,
It is also necessary to rotate the image by 90 degrees. The microcomputer 62 performs a 90 degree coordinate transformation on the image signal stored in the page memory 80 and writes the transformed image signal into the page memory 78 . Therefore, regardless of the angle θ1, the page memory 78 is
An image signal representing an image such as that shown in FIG. 2 is stored. Control proceeds to step 5OO9 [01101 In step 5OO9, the data stored in the page memory 78 is sequentially read out, encoded, and transmitted. This program ends when the sending process ends. [0111] Referring to FIG. 17, in the initial processing at step 5OO2, the following processing is performed. In step 5201, the panel controller 66 (
From FIG. 9), it is determined whether a signal indicating that the start button has been pressed is given. If the result of the determination is NO, control returns to step 5201; otherwise, control proceeds to step 5202. [0112] In step 5202, image memories 76-80 are enabled for writing. Control continues to step 3203. [0113] In step 5203, a signal is given to the mechanics/recording control circuit 68 to indicate that conveyance of the document should be started. In response to this, the feed roller 34 and the conveyance roller 36 start rotating, and after the conveyance of the document to the reading position starts, the control returns to the main routine. [0114] With reference to FIG. 18, the process in step 5OO3 is as follows:
It is done as follows. In step 5301,
Both ends of the leading edge of the document supplied to the reading position are connected to the sensor 4.
2a, 42C (FIG. 10), the information about the time of the detection is notified to the program. Control proceeds to step 83O2. [0115] In step 5302, the time difference between the detection of the document by sensor 42a and the detection of the document by sensor 42C is calculated. Control continues to step 5303. [0116] In step 5303, a lookup of the tilt table stored in advance in the ROM in the microcomputer 62 is performed based on the time difference obtained in step 5302 and information on the paper size of the supplied original. As described above, if the time difference between the detection of the document by the sensors 42a and 42c is known, the inclination θ of the document can be determined according to a certain calculation formula. However, generally, the relationship between the time difference and the slope θ is tabulated in advance and stored in the ROM. By doing this, if the time difference and document size are obtained, the slope θ can be looked up at high speed by searching the table. [0117] In step 5304, the desired tilt angle θ is determined from the data stored in the tilt table. Control continues to step 5305. [0118] In step 5305, the obtained tilt angle θ is
It is stored in the memory within the microcomputer 62. Control returns to the main routine. [0119] Referring to FIG. 19, reading of image information in step 5OO4 is performed as follows. Step 54
At 01, a one-page document separated by the rubber plate 38 and supplied to the reading position is converted into image information by the CCD 60 and provided to the reading processing circuit 70. Control continues to step 8402. [01201 In step 84O2, the image signal for one page inputted from the reading processing circuit 70 is written into the page memory 76. Control continues to step 5403. [0121] In step 5403, it is determined whether all pages have been read. This determination is made, for example, by using the document sensor 40 to detect whether there is a document to be read next. If the answer is NO, control returns to step 5401, otherwise control returns to the main routine. [0122] Referring to FIG. 20, the detection of the character arrangement direction in step 5OO6 is performed as follows. Step 56
At step 01, an angle θ1 formed between the arrangement direction of characters included in the original image and the main scanning direction is detected. Control continues to step 86O2. [0123] In step 5602, it is determined whether θ1 is between 145 degrees and 45 degrees. If the answer to the determination is NO, control proceeds to step 5603; otherwise, control proceeds to step 5604. [0124] In step 5603, 1 is set to a flag prepared in advance in the program and indicating the character string direction of the document. Control then returns to the main routine. [0125] When control proceeds to step 5604, O is set in the aforementioned flag. Control then returns to the main routine. [0126] As can be easily seen, when this flag is 0, the direction of the character string and the main scanning direction generally match, and when the flag is 1, the direction of the character string and the main scanning direction They intersect at approximately right angles. [0127] The tilt correction process in step 5OO7 is performed as follows. In step 5701, the sensor 42
It is determined whether or not there is an inclination in the insertion direction of the document detected by a and 42c. If the answer is YES, control proceeds to step 57O2, otherwise control returns to the main routine. [0128] When the control proceeds to step 57O2, the inclination angle θ stored in the memory in step 5305 of FIG.
is read from memory. Control is at step 5703
Proceed to. [0129] In step 5703, the page memory 76 is
Images for each page are read out. Control is step 5
Proceed to 704. [01301 In step 5704, a correction calculation is performed on the read image signal to offset the tilt angle θ. This correction calculation is performed according to a well-known formula for coordinate transformation by rotation. The operation is also USP4,
829,452. Control continues to step 5705. [0131] In step 5705, the corrected image signal is written to the page memory 78.80. Control is step 5
Proceed to 706. [0132] In step 5706, it is determined whether the correction calculation has been performed on the image signals of all pages of the document stored in the page memory 76. The answer to the judgment is N. If so, control returns to step 5703; otherwise control returns to the main routine. [0133] Referring to FIG. 22, the image signal correction based on the character arrangement direction in step 5OO8 is performed as follows. In step 5801, it is determined whether the value of the flag set in steps 3603 and 5604 of step 8006 is O. step 580
If the answer to the question 1 is YES, it means that the direction in which the characters on the document are arranged is substantially the same as the main scanning direction, as described above. Therefore, in this case, no correction processing is performed and control is returned to the main routine. If the answer to the determination in step 5801 is NO, control proceeds to step 5802. [0134] If the control proceeds to step 5802, this means that the arrangement direction of the characters on the supplied document is approximately 90 degrees different from the main scanning direction. In step 5802, image signals for one page are read from page memory 80. Control continues to step 5803. [0135] In step 5803, a coordinate transformation calculation is performed on the read image signal to rotate the image at an angle of 90 degrees. Control continues to step 5804 [013
6] In step 5804, the image signal corrected in step 5803 is written to the page memory 78. Control continues to step 5805. [0137] In step 5805, it is determined whether or not correction calculations have been performed on all the image signals of the originals stored in the page memory 80. If the answer is NO, control returns to step 58O2, otherwise control returns to the main routine. [0138] A more specific explanation of the image signal conversion performed in steps 5OO7 and 5OO8 will be as follows. When the inclination θ1 is between 145 degrees and 45 degrees, the image signal formed in the page memory 76 represents the image shown in FIG. In this case, step 5OO7
By the correction calculation performed in step 5704, the image signal is converted into one representing the original image as shown in FIG. In this case, the correction of the character arrangement direction in step 5OO8 is not performed. therefore,
The page memory 78 stores an image signal representing a document image as shown in FIG. [0139] On the other hand, when the inclination θ1 is between 45 degrees and 135 degrees, the document image represented by the image signal stored in the page memory 76 is shown in FIG. As a result of the correction calculation performed in step 5704 (FIG. 21) of step 5OO7, an image signal representing a document image as shown in FIG. 14 is stored in page memory 80. As mentioned above, this image signal cannot be compressed with high efficiency as it is. [0140] In step 5008, the original image 84 shown in FIG. 14 is rotated 90 degrees and written to the page memory 78. Therefore, the document image stored in page memory 78 is as shown in FIG. [0141] The transmission process in step 5OO9 is performed as follows. Referring to FIG. 23, in step 5901, microcomputer 62 causes page memory control circuit 74 to
Gives a signal to start encoding. Control proceeds to step 5902 [0142] In step 5902, page memory control circuit 7
4, image signals for one page are read out from the page memory 78 and encoded. The encoded signal is stored in a dedicated memory for transfer, not shown. Control continues to step 5903. [0143] In step 5903, it is determined whether all pages of the document stored in page memory 78 have been encoded. If the answer to the determination is No, control returns to step 5902; otherwise, control proceeds to step 5904. [0144] When the control proceeds to step 5904, the microcomputer 6
2 gives a signal to the transmission control circuit 72 and modem 30 to send out the compressed signal stored in the dedicated memory for transfer onto the telephone line. The transmission control circuit 72 and modem 30 convert the compressed signal stored in the dedicated memory into an analog signal and send it to the LIU according to a predetermined procedure.
32 onto the telephone line. Step 5904
After the processing in is completed, control of the program is returned to the main routine. [0145] The image signal encoding process performed in step 89O2 is performed on an image signal representing a document image as shown in FIG. In this original image, the direction in which the character string extends matches the reading direction (main scanning direction) of the signal for encoding. The width of the blank line between strings is the widest compared to other cases. Therefore, blank lines are efficiently encoded using the makeup code described above, and the compression rate of the entire document is maximized. [0146] As described above, according to the facsimile apparatus of the embodiment of the present invention, the difference in time between the leading edge of the supplied document and the main scanning direction is determined based on the time difference in which both ends of the leading edge of the supplied document pass in front of a predetermined sensor. The angle formed is found. Furthermore, the direction in which the character strings on the original image extend is detected from the original image that has been digitized by the CCD and the reading processing circuit and stored in the page memory. The obtained original image signal is first subjected to coordinate transformation so as to correct the inclination of the transmitted original. The coordinate-converted image signal is further processed so that the extending direction of the character string on the original image represented by the image signal matches the direction (main scanning direction) in which the image signal is read from the page memory during encoding. If necessary, it is rotated an additional 90 degrees. As a result, when the corrected image signal is encoded, blank lines between character strings are encoded most efficiently, and the compression efficiency of the entire document image is also maximized. [0147] Although this invention has been described above based on one embodiment, this invention is not limited to the above-described embodiment. For example, the following embodiments may also be considered. [0148] The hardware of the facsimile machine for realizing the second embodiment of the invention recited in claim 3 is the same as that shown in FIG. The second embodiment differs from the first embodiment in the structure of the program executed in the microcomputer 62. [0149] Referring to FIG. 24, the program executed in the microcomputer 62 of the facsimile machine of the second embodiment has the following control structure. In step 5oil, it is detected whether the original to be transmitted has been set on a document table (not shown). This process is performed in step 5OO of FIG.
The processing is the same as that performed in 1. Control proceeds to step 5O12. [01501 In step 5O12, necessary initial processing is performed. This process is the same as the process performed in step 5OO2 of FIG. Control is step 5
Proceed to O13. [0151] In step 5O13, image information is read for all documents set on the document table. This process is the same as the process performed in step 5OO4 of FIG. Control proceeds to step 5O14. [0152] In step 5O14, it is detected which portion of the read document image is the character arrangement area. This process is the same as the process performed in step SOO5 of FIG. Control is step 5O1
Proceed to step 5. [0153] In step 5O15, the direction in which the character string extends in the detected character array area is detected. This process is almost the same as the process performed in step 5OO6 of FIG. However, in this step 5O15, the direction in which the character string extends is detected more precisely than was done in the first embodiment. Control continues to step 5O16. [0154] In step 5O16, based on the detected character arrangement direction, the document image is rotated by 90 degrees (if necessary) so that the character strings are arranged along a direction that roughly matches the main scanning direction. ) is carried out. This process is the same as the process performed in step 5oos of the first embodiment. Control proceeds to step 5O17. [0155] In step 5O17, the image signal that has been corrected so that the extending direction of the corrected character string approximately matches the main scanning direction is corrected so that the extending direction of the character string and the main scanning direction more accurately match. Corrected. This process is almost the same as the process performed in step 5OO7 of the first embodiment. However, it should be noted that in this case, unlike the first embodiment, the angle for correction is derived from the character arrangement direction obtained in step 5O15. After the process of step 5O17 is performed on the image signals of all pages, this program ends. [0156] Even with a program having the above structure, the first
The same effects as in the embodiment can be obtained. That is,
This embodiment also makes it possible to compress image signals with high efficiency regardless of the direction of the document supplied to the facsimile machine. [01,57] In this second example, the processes in step 5O16 and step 5O17 were described as separate. However, the invention is not limited thereto. For example, it is also possible to create a program to correct the angle between the character arrangement direction and the main scanning direction obtained in step 5O15 by calculating -degrees. [0158] In each of the above embodiments, the case of reading an image including a character string has been described. However, the present invention is not limited thereto. For example, if the image includes a character string in which a plurality of symbols and other characters are arranged in a fixed direction, not just a character string, the image can be read and the signal compressed efficiently using the same procedure. [0159]

【Effect of the invention】

As described above, according to the image reading device according to claim 1, the arrangement of read image data before encoding is read with the arrangement direction of character strings of the image aligned with the scanning direction. The corrected image data is corrected by the correcting means by the amount of the inclination of the character string detected by the inclination detecting means so that the arrangement is the same as when Even in the case of a reading state in which the scanning line S for reading overlaps the character string C included in the image to be read, that is, one scanning line intersects multiple character strings C, encoding is possible. The data compression efficiency of image data to be processed can be improved. [01601 According to the image reading device according to claim 2, when the arrangement direction of the character strings of the image to be read is in the vertical arrangement direction perpendicular to the scanning direction, the arrangement of the read image is corrected by rotating by 90 degrees. The configuration is such that the character strings are arranged in the same way as when read in the horizontal direction, where the character strings are arranged in the same direction as the scanning direction, and then encoded. In this case, the compression efficiency of image data encoding also improves. [0161] According to the facsimile apparatus according to claim 3, it is determined whether the character arrangement direction on the document is closer to the first direction or the second direction of the arrangement of the memory cells, and if necessary, the image is predetermined. The characters are rotated by an angle such that the character arrangement direction is closer to the first direction. The signal compression means includes a first
Image signals are read out and encoded from the array of storage cells along the direction. By rotating the image as described above,
The average length of the white runs and black runs along the first direction is longer than when the image is not rotated, and the number of white runs and black runs is smaller. Generally, in compression encoding systems, one white run or one black run is converted into one code. Therefore, according to the facsimile apparatus according to the present invention, the number of codes after compression is reduced, and compression efficiency is improved. It also improves transmission efficiency.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing a schematic configuration of an image reading device that is an embodiment of the invention as set forth in claim 1. FIG. [FIG. 2] A flowchart showing the operation of the image reading device shown in FIG. 1.

FIG. 3 is an explanatory diagram illustrating a state in which character rows of an image are read in a tilted state with respect to the scanning direction.

FIG. 4 is a block diagram showing a schematic configuration of an image reading device that is an embodiment of the invention as set forth in claim 2;

5 is a flowchart showing the operation of the image reading device shown in FIG. 4. FIG.

FIG. 6 is a block diagram showing a schematic configuration of an image reading device according to another embodiment of the invention according to claim 2.

7 is a flowchart showing the operation of the image reading device shown in FIG. 6. FIG.

8 is a diagram for explaining the relationship between the character arrangement direction and the scanning direction in the operation of the image reading device shown in FIG. 3. FIG.

FIG. 9 is a block diagram of a facsimile apparatus according to the invention according to claim 3.

FIG. 101 is a diagram schematically showing the arrangement of sensors for detecting the leading edge of a document supplied to a facsimile machine. FIG. 11 is a schematic diagram of a document image stored in a page memory.

FIG. 12 is a schematic diagram of a document image stored in a page memory.

FIG. 13 is a schematic diagram of a document image stored in a page memory.

FIG. 14 is a schematic diagram of a document image stored in a page memory.

FIG. 15 is a diagram schematically showing the angle formed between the orientation of character strings on a document and the main scanning direction.

FIG. 16 is a flowchart of a main routine of a program executed by a microcomputer of a facsimile machine according to a first preferred embodiment of the invention;

[Figure 17] It is a flowchart of initial processing.

[Figure 18] 3 is a flowchart of document tilt detection processing.

[Figure 19] 3 is a flowchart of reading image information.

FIG. 20 is a flowchart of a routine for detecting character arrangement direction.

[Figure 21] It is a flowchart of tilt correction.

[Figure 22] 3 is a flowchart of a routine for character arrangement direction correction.

[Figure 23] 3 is a flowchart of a transmission processing routine.

FIG. 24 is a flowchart of a main routine of a program executed in a microcomputer of a facsimile machine according to a second preferred embodiment of the invention;

[Figure 25] It is a schematic diagram of a manuscript.

[Figure 26] It is a schematic diagram of a manuscript.

FIG. 27 is a diagram showing a part of the terminating code included in the MH code.

FIG. 28 is a diagram showing a part of the makeup code included in the MH code.

[Figure 29] It is a schematic diagram of a manuscript.

[Explanation of symbols]

1 Scanner 2 Binarization section 3 Image memory 4 Tilt detection section 5゛Tilt correction section 6. Encoding section 7 Control section 10 Manuscript 42° control circuit Original sensor 42a-42c Original edge sensor CCD image sensor Microcomputer Vertical/horizontal detection section Rotation processing section Encoding section Tilt angle detection section

[Document core]

drawing

[Figure 1]

[Figure 2]

[Figure 3]

[Figure 4]

[Figure 5]

[Figure 6]

[Figure 7]

[Figure 8]

[Figure 9]

[Figure 101 [Figure 11] y

[Figure 12]

[Figure 13]

[Figure 14] Vice-retirement

[Figure 16]

[Figure 17]

[Figure 18]

[Figure 19]

[Figure 20]

[Figure 21]

[Figure 22]

[Figure 23]

[Figure 24]

[Figure 25] 225 Jiku scanning way yuzu

[Figure 26]

[Figure 28]

[Figure 29] (b) Main ii square

Claims (3)

[Claims] 1. An image reading device that scans and reads an image in a fixed direction and encodes the read image data,
Tilt detection means extracts a character string part in which a plurality of characters are arranged in a predetermined direction from the image and detects the inclination of the character string with respect to the scanning direction; a correction means for correcting the inclination of the character row detected by the inclination detection means so that the arrangement direction of the character row is the same as the arrangement of image data read under a state in which the arrangement direction of the character row is made to match the scanning direction; An image reading device comprising: encoding means for encoding image data corrected by the means. [Claim 2] An image reading device that scans and reads an image in a fixed direction and encodes the read image data, wherein a character string portion in which a plurality of characters are arranged in a predetermined direction is extracted from the image. , a length/width determination means for determining whether the character string is arranged in a horizontal direction that matches the scanning direction or a vertical direction that is orthogonal to the scanning direction; a correction means for rotating and correcting the arrangement of the image data for one image by 90 degrees so that it becomes the same as the arrangement of the image data read under the condition that the arrangement direction of the character rows is aligned with the scanning direction; If the vertical/width determining means determines that the image data is in the vertical alignment direction, the corrected image data is encoded by the correcting device, and if the vertical/horizontal determining device determines that the image data is in the horizontal alignment direction, the corrected image data is encoded without being corrected by the correcting device. An image reading device comprising: encoding means for encoding raw image data. 3. A pixel for generating a pixel signal indicating the brightness of each pixel by dividing and reading an image on a document into a plurality of pixels arranged two-dimensionally. signal generating means; pixel signal storage means each including a plurality of storage cells for storing the pixel signal in logical association with one of the pixels; The pixels are logically arranged two-dimensionally in a predetermined first direction and in a second direction intersecting the first direction, corresponding to the arrangement of the pixels, and the pixel signals are stored in the pixel signal storage means. Based on the stored pixel signal, a character arrangement direction indicating the arrangement direction of the character string included in the image on the document expressed by the stored pixel signal in the arrangement of the plurality of storage cells is determined. character arrangement direction detection means for detecting the character arrangement direction; character arrangement direction determination means for determining which of the first and second directions the detected character arrangement direction is closer to; In response to the output of the arrangement direction determining means, if the character arrangement direction is closer to the second direction than the first direction, the image on the original document is expressed by the stored pixel signal. , so that the character arrangement direction is closer to the first direction,
an image rotation means for rotating the image on the document represented by the stored pixel signal by a predetermined angle; a signal compression means for reading out the signal along the first direction from the signal storage means and compressing it using a compression code; and for transmitting the pixel signal compressed by the signal compression means to another station via a communication line. a facsimile machine comprising: a transmission means;