JP3333889B2 - Image processing device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus capable of reproducing an image of a book original or the like placed on an original platen in an open state without distortion.

[0002]

2. Description of the Related Art In recent years, in order to increase the efficiency of document input, a device for reading the contents of a document placed on a document table has been used. In the case of a bound document such as a book document, the read image is distorted due to the lifting of the binding portion, and thus a device for correcting the distortion has been devised.

For example, in the invention disclosed in Japanese Patent Application Laid-Open No. Sho 62-143557, a rising state of a binding portion is detected by a distance sensor, and reading in a sub-scanning direction is performed in accordance with a distance between a document surface and a reading portion. The pitch, that is, the scanning speed is changed to correct image distortion.

In the invention disclosed in Japanese Patent Application Laid-Open No. 3-117965, the degree of curving of a document surface is detected, and a document image stored in a page memory is expanded to perform processing.
The image distortion is corrected.

[0005]

However, in such a conventional image processing apparatus in which the scanning speed is changed, it is necessary to slightly change the scanning speed in accordance with the floating state of the document. Therefore, the occurrence of speed unevenness becomes a problem, and very precise control accuracy is required. In the case of performing the decompression process, it is necessary to secure a sufficient memory capacity for storing the entire image.

The present invention has been made in view of such a conventional problem, and does not change the scanning speed of the line sensor and only secures a minimum memory capacity, and is capable of performing the scanning in the sub-scanning direction. It is an object of the present invention to provide an image processing device capable of correcting an image distortion.

[0007]

According to the present invention, there is provided an image processing apparatus which scans a line sensor to read an original image and records the image as two-dimensional image information. Line sensor driving means for driving the image sensor at a constant speed in the sub-scanning direction, image storage means for storing image information of a plurality of lines in the main scanning direction read by the line sensor, and a line sensor driven by the line sensor driving means sub-scanning direction and the image distortion detecting means for detecting a distortion of said original image, according to a distortion of the original image detected by the image distortion detection means, the La
To the image storage means of the image information read by the in-sensor
Control means for repeatedly reading out the image information stored in the image storage means in units of the main scanning direction of the line sensor at the same time as the writing of the image data.

[0008]

According to the present invention constructed as described above, the image storage means temporarily stores the image read by the line sensor for at least the number of pixels of the line sensor, and detects the state of the image in the sub-scanning direction. The means detects the distortion of the image, and in response thereto, the control means corrects the photographed image by repeatedly reading one line of the image information of the distorted portion from the image storage means a plurality of times. Things.

[0009]

Embodiments of the present invention will be described below in detail with reference to the drawings.

FIG. 1 is a drawing showing the overall configuration of an image reading apparatus. A document such as a book or a file is placed upward on a document table 1 of the image reading apparatus, and an imaging camera unit 2 for reading the document by scanning of a line sensor is provided above the document. An illuminating unit 3 disposed above the document table 1 for illuminating the document, an operation unit 4 for setting image reading conditions and the like, and an image of an end portion of the document arranged on the document table 1 at the back of the document table 1. A distance measuring mirror 5 is provided.

FIG. 2 is a front view showing a schematic configuration of the apparatus. The imaging camera unit 2 includes an imaging lens 6 for forming an image of the document on the image sensor surface, and a motor and a driving belt and gears (not shown) on a focal plane on which the image of the document is formed. And a line sensor 7 that is an image sensor that is driven by the driving means and scans in the sub-scanning direction (the direction indicated by the arrow in FIG. 2). The document 10 placed on the platen 1 is, for example, a book or a file that is bound at one end and has a columnar shape in which each page is spatially curved when opened left and right.

FIG. 3 is a block diagram for explaining the image processing circuit.

The number of pixels of the line sensor 7 used in the imaging camera unit 2 is determined by the size of the document to be photographed and the resolution for reading the document, and in this embodiment, the short side of A3 (29
7 mm) at a resolution of 400 dip, so that 50
A lens reduction type CCD line sensor of 00 pixels is used.

The capacity of a memory serving as image storage means for storing image information is determined by how many bits are allocated to one pixel of image information, which is the number of read gradations for reproducing an original. In this embodiment, 8 bits are assigned to one pixel so that 256 gradations can be expressed.

[0015] Generally, commercially available semiconductor memories include:
The bit configuration is determined in base units. For example, 4MD
In a RAM, the size is 1024 (2 ¹⁰ ) × 1024 × 4 bits, 2048 (2 ¹¹ ) × 2048 × 1 bit, and the like. Therefore, in order to efficiently record the image information of the line sensor by using these semiconductor memories, one column has 1024 lines.
Five memories each having a bit configuration are connected in series (1024 × 5
= 5120 ≧ & $ 5000).

A memory having such a configuration includes, for example, a 1M DRAM (1024 × 1024 × 1 bit) and a 4M DRAM.
Although there is a DRAM (1024 × 1024 × 4 bits) and the like, a 4MDRAM is selected in this embodiment in order to reduce the number of chips on the substrate and simplify the circuit configuration. Since this memory has a depth of 4 bits per address, two memories are arranged in parallel and correspond to the number of gradations of 8 bits. FIG. 4 conceptually shows this state, in which a total of ten 4MDRAMs are used.

Next, storage of image information (writing to the memory) will be described. The line sensor 7 has a light receiving element (phototransistor) for 5000 pixels, and an output analog signal is converted into a digital signal by the A / D converter 11 and then divided into upper and lower 4 bits.
The vertically divided image information is input to an S / P (serial / parallel) converter 12, and in this embodiment, one pixel can be simultaneously written into each of five DRAMs arranged in series. The five pixels are grouped as one unit, and written in the image memory 15 in units of one unit in accordance with an instruction from the CPU 13 and the address controller 14.

Next, execution of the recorded image information (call from the memory) will be described. In the present embodiment, the image information written in the image memory 15 is read from the image memory 15 at the same time for five pixels in accordance with an instruction from the CPU 13 and the address controller 14. The read image information is arranged in the P / S (parallel / serial) converter 16 in the order of one pixel at a time in the same manner as at the time of writing, and then the upper and lower bits are put together. Printers and CRTs,
It is sent to the subsequent device.

FIG. 4 is a diagram showing an arrangement of the image memory 15. As described above, the image memory 15 is composed of ten memories having a storage capacity of 4 Mbits per chip. Each chip has a 1M × 4 bit configuration and has 1024 storage elements in each of a row direction and a column direction. The 10 memory chips have upper bits, lower bits
And the five memory chips in each group are S / P
This corresponds to 5000 pixel information of the converted image information. Since there are 1024 storage elements in the column direction, 1024 lines of image information can be stored in all of the ten memories.

FIG. 5 is a diagram showing the address arrangement of the upper bits 1 to 5 of the image information stored in the image memory 15.

First, a method of controlling an address at the time of writing to the image memory will be described. In the image information in the line sensor main scanning direction, the first to fifth data are stored in the first row of the memory chips for higher order 1 to 5, and
The 0th data sequentially proceeds to the second row, the 11th to 15th third rows, and when the 0th data comes to the 1000th row, writing is completed up to the 4996th to 5000th data.

The image information in the line sensor sub-scanning direction is 1
The data of the line is stored in the first column of the memory chips for the higher order 1 to 5, the second line sequentially proceeds to the second column, the third line proceeds to the third column, and the 1024th line is stored in the 1024 column. Then, the process returns to the first column and the 1025th line is stored. That is, in the present embodiment, the entire original surface is not stored in the image memory as the original image, but only an image of 1024 lines is temporarily stored.

Next, a method of controlling an address at the time of reading from the image memory will be described. The image information is read in an early write mode performed after writing, and the image information in the sensor main scanning direction is read in the same order as when writing.

On the other hand, the image information in the line sensor sub-scanning direction is read out in the steady state (when the original is not distorted) as in the writing, but the original is distorted and expanded to correct the distortion. When a signal is input from the CPU 13,
The reading column is stopped, and the same column is read a plurality of times. The same applies to the lower-order 1 to 5 memory chips. That is, an image read and reproduced is an image enlarged in the sub-scanning direction only when an expanded signal is included. Thus, when a document (a book or the like) in which both sides of the binding portion are raised is read as shown in the side view of the document shown in FIG. 6A, for example, as shown in FIG. When the image is read out from the image memory and is distorted by being reduced in the sub-scanning direction (the arrow in the figure or the direction opposite to the arrow), FIG.
As shown in (C), a normal image enlarged in the sub-scanning direction is obtained.

FIG. 10 shows the address arrangement of the memory in the conventional image reading apparatus. As shown in the figure, in the related art, the output which is read by the line sensor and converted into a digital form is sequentially written in line order and written. As in the present embodiment, control for changing the row or column of the memory at a line break is not performed.

Next, the operation of the present apparatus will be described. FIG. 7 is a flowchart for explaining the operation. First, a copy start determination is made (S1). When it is detected that a read operation start instruction (copy start) has been input, the lamp of the illumination unit 3 is turned on (S2), and the original 1
Illuminates 0. Next, a preliminary scan for the distance measurement operation is started (S3). Here, the distance measuring mirror 5 is arranged at an angle of 45 °, and the CCD line sensor 7 photographs the upper end of the document from a substantially horizontal direction. In this ranging operation,
By moving the CCD line sensor 7 from one end and taking an image of the upper end of the document reflected on the distance measuring mirror 5,
The height of the document is calculated and stored. That is, the image distortion is detected. At this time, the line sensor 7 and the distance measuring mirror 5 function as image distortion detecting means.

Next, an operation for correcting image distortion is performed (S4). In this correction operation, as shown in FIG. 8, based on the obtained height information of the original, a height distribution at regular intervals ΔX is obtained in the scanning direction, and the enlargement correction rate of the original in this section is calculated as follows. It is determined according to an approximate expression. ΔL ² = ΔX ² + ΔH ² Enlargement correction rate = ΔL / ΔX where ΔL is the length of the document surface, and ΔH is the height of the document at ΔX. The number of repetition lines in the section ΔX is calculated from the enlargement correction rate, and the number of lines is evenly distributed in the section. At this time, the CPU 13 and the address controller 14
Act as control means.

In the present embodiment, as will be described later, at the time of the main scan, reading is performed while performing correction while simultaneously writing to the image memory. Therefore, the number of lines required for repeated reading for correction is required. If the value exceeds the number of columns in the memory, the calculation result needs to be corrected. In this embodiment, the maximum number of repeatable lines is 1024, which is the number of memory columns.
When it is necessary to repeatedly read data exceeding the number of repetitions (1024), it is necessary not to exceed the number. This is because, for example, image correction is necessary for the n-th line, and when the number of repetitive readings for that line exceeds 1024, the n + 1024-th memory in the reading operation is stored in the n-th memory column as described above. This is because image information is stored and cannot be repeatedly read beyond 1024.

In this embodiment, this correction is made by changing the scanning speed in the sub-scanning direction. Specifically, the correction is performed based on the number of lines of one page when a document without distortion is read. The difference from the number of lines when a distorted image is read, that is, the number of times of repeated reading per line required for correction does not exceed 1024.

Therefore, in this embodiment, in order to make the state of the original for reading an image (the state of the height of the original) correspond to as many states as possible, it is determined that one memory row per line does not exceed 1024. Rather, control is performed so that the total number of lines including the number of repetitive readings per page does not exceed the total number of lines when reading a document without distortion requiring no correction.

This is, for example, at a resolution of 400 dpi.
When the short side of A4 is photographed in the sub-scanning direction, the number of photographing lines is about 3 in a state without distortion (a state in which no correction is necessary).
It will be 300 lines. Assuming that the number of lines for A4 and one page in the preliminary scan when a document (a book or the like) distorted due to a curvature is photographed is 2000 lines,
It is necessary to decompress the image by 3300-2000 = 1300 lines. That is, the number of repeating lines required for correction exceeds 1024. Therefore, by changing the scanning speed of the line sensor at the time of the main scan according to the following equation, the number of photographing lines of the original (2,000 above) is increased, and as a result, the interval of ΔX is shortened. Is controlled to reduce the number of repetitive read lines (1300 described above) necessary for the above and to control the required number of expanded lines within the memory capacity. Thus, even when the scanning speed is reduced, the data output timing remains constant and does not change. V ′ = V × (2000 + 1024) / 3000 where V is the speed before change (for example, preliminary scan speed), and V ′ is the speed after change (for example, main scan speed).
Represents

When the above calculation is completed, the line sensor 7
Is scanned in the direction opposite to the preliminary scan, and a main scan for photographing the original is performed (S5). At this time, the CPU 13 controls the address based on the number of repetition lines obtained by the calculation, thereby performing partial enlargement in the sub-scanning direction.

In the present embodiment, the operation for correcting the number of repetition lines is performed by changing the scanning speed. However, the present invention is not limited to this, and the memory capacity may be increased. Since it is expected that it is difficult to read an image, it is possible to display an inability to read by providing an alarm or the like, and to prevent reading.

FIG. 9 is a diagram for explaining the timing signals of the image memory. In this apparatus, since it is necessary to simultaneously write and read image information to and from the image memory during the main scan, a very high-speed operation is required, and the number of repetition lines increases. Therefore, in this embodiment, a DRAM capable of high-speed operation and low in cost is used as a memory element.

In the figure, RAS (row address strobe) indicates a row address activation signal, and CAS (alum address strobe) indicates a column address activation signal. In this embodiment, since writing is performed to the same row address, the row address may be activated only once, and then the write row address and the read column address may be sequentially applied. Writing and reading of image information to and from the memory are controlled by a WE (write enable: writable) signal. When this signal is at a low (L) level, write data is at a high (H) level and read data is at a data line. Appears in

In general, a DRAM requires a refresh operation for rewriting data in order to retain the memory contents. This refresh operation rewrites the data to the same address every predetermined period (refresh cycle). In one form, an address is taken in by RAS input, a corresponding word line is selected,
Activate (read / write) the sense amplifier and
Although the refresh of a plurality of bits selected by the input address is performed, as in the present embodiment, the row address is sequentially updated at high speed in accordance with the flow of image information, that is, in a period earlier than the refresh cycle. By performing the access for writing or reading the image information, this access operation becomes a substitute for the refresh operation, and the refresh operation becomes unnecessary.

As described above, according to the present embodiment, in an image reading apparatus for reading a spatially curved original such as a book, an image memory for temporarily storing image information is optimally arranged for a line sensor. With a simple and inexpensive configuration, high-speed and high-performance correction processing can be realized. In addition, in writing and reading operations at one time, it is only necessary to give a column address once, and the operation time can be reduced. Further, in the DRAM, a rewriting operation (refresh operation) which is usually required can be omitted.

In the above description of the present embodiment, the memory circuit is used for partial enlargement in the sub-scanning direction for correcting the distortion of the original, but it is also used as a buffer memory for temporarily storing the entire page. Can be. However, at this time, the number of memory elements in the column direction is increased by the required number of lines, and after the write operation is completed, only the row address is activated until the read operation is performed.
Data must be rewritten.

[0039]

As described above, according to the present invention, it is possible to correct the original image in the sub-scanning direction in one screen without changing the scanning speed, so that it is possible to eliminate image unevenness due to scanning unevenness. There is no need to perform complicated scanning drive system mechanical control.

[Brief description of the drawings]

FIG. 1 is a perspective view illustrating an external configuration of an image reading apparatus according to an embodiment of the present invention.

FIG. 2 is a front view illustrating a configuration of an image reading apparatus according to an embodiment of the invention.

FIG. 3 is a block diagram illustrating a configuration of an image processing circuit in the image reading apparatus according to the embodiment to which the present invention is applied.

FIG. 4 is a diagram illustrating a configuration of an image memory of the image reading apparatus according to the embodiment to which the present invention is applied.

FIG. 5 is a diagram for explaining an address arrangement of an image memory in an embodiment to which the present invention is applied.

FIG. 6 is a diagram for explaining a state of image reading.

FIG. 7 is a flowchart for explaining the operation of the image reading apparatus according to the embodiment to which the present invention is applied.

FIG. 8 is a view for explaining image correction in the operation of the image reading apparatus according to the embodiment to which the present invention is applied.

FIG. 9 is a diagram for explaining access timing to an image memory in the operation of the image reading apparatus according to the embodiment to which the present invention is applied.

FIG. 10 is a diagram for explaining an address arrangement of an image memory in a conventional image reading apparatus.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... platen, 2 ... camera part, 3 ... illumination part, 4 ... operation part, 5 ... distance measuring mirror, 6 ... imaging lens, 7 ... line sensor, 10 ... original, 11 ... A / D converter, 12, 16: S / P converter, 13: CPU, 14: Address controller, 15: Memory.

────────────────────────────────────────────────── ─── Continuation of the front page (56) References JP-A-5-161001 (JP, A) JP-A-5-161000 (JP, A) JP-A-5-161002 (JP, A) (58) Field (Int.Cl. ⁷ , DB name) H04N 1/04-1/207 H04N 1/40-1/409

Claims (1)

(57) [Claims] An image processing apparatus for scanning a line sensor to read a document image and recording the image as two-dimensional image information, wherein the line sensor drives the line sensor at a constant speed in a sub-scanning direction. Sensor driving means; and a plurality of lasers in the main scanning direction read by the line sensor.
An image storage means for storing image information in minutes, and the image distortion detecting means for detecting a distortion in the sub-scanning direction of the document image of the line sensor driven by said line sensor driving means, detected by the image distortion detection means depending on the distortion of the original image, before the image information in which the line sensor has read
Control means for repeatedly reading out the image information stored in the image storage means in units of the main scanning direction of the line sensor at the same time as writing to the image storage means.