JPS5827145A - Processing method for strengthening contour of color picture signal and scanner for color picture - Google Patents

Abstract

PURPOSE:To perform processing for strengthening contours with each of color signals and to improve the sharpness of color pictures in the three color separation and photometry of the color pictures by fetching and synthesizing sharp picture signals and unsharp picture signals for every color. CONSTITUTION:A color original 1 is illuminated by an illuminating device 2, and the reflected light thereof passes through a filter turret 3 and a contracted optical system 4 and is made incident to a half mirror 5. One of the bisected beams of the light is made incident to a solid state image sensing element 6 for unsharp pictures and the other to a solid state image sensing element 7 for sharp pictures. In the stage of the three color separation and photometry of the original 1, the elements 6, 7 are moved intemittently from a moving start point A up to a reading end point C after, for example, a blue filter 8 is inserted into an optical path. The unsharp picture signal and sharp picture signal of respectively a blue color are outputted therefrom. Thereafter, the elements 6, 7 are returned to the point A, and a green filter 9 is inserted into an optical path. The green components of the respective picture elements are measured through the filter 9 and further the red components are measured.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention extracts and combines sharp image signals and unsharp image signals for each of the three color light components of a color image, thereby performing edge enhancement processing independently for each color signal. The present invention relates to an edge enhancement processing method and a color image scanning device.

In the field of image processing, edge enhancement processing is often performed to sharpen images, and there are two types of edge enhancement processing: electrical and optical. Optical processing involves reading a clear image and a blurred image, and combining the obtained sharp and unsharp image signals to perform edge enhancement processing, and two-dimensional processing is It is widely used because it is simple.

FIG. 1 shows the principle of optical edge enhancement processing. (A) shows a clear image f(x,y),
(B) shows a blurred image g(x,y). These images f (x * y) 9g (x + y
), the contour-enhanced image h(x
, y) is obtained.

h(x,y)=α[f(X,'7)−8g(x
,y))--(1) If α-2 and β=172 in this equation, equation (1) becomes as follows.

h (x, y) = 2f (x, y) - g (x + y) =
(2) ('0) indicates this equation (2), and the image has improved sharpness.

As a method for applying this contour enhancement processing to a color image signal, for example, Japanese Patent Laid-Open No. 52-152301 and No. 53-11
As described in No. 2018, an edge-enhancing signal was obtained by subtracting a gray-level sharp image signal and an unsharp image signal, also a gray-level, in an analog manner, and this was obtained by three-color separation photometry. A common method is to add each color signal in an analog manner.

However, the gradation level of each color signal and the gradation level of the gray level sharp image signal and unsharp image signal for forming the edge emphasis signal are strictly Since they are not the same, the outline enhancement becomes unbalanced for each color, and the sharpness of the reproduced color image cannot be improved.

In view of the above-mentioned drawbacks, the present invention provides a method for contour enhancement processing of color image signals, which is capable of improving the sharpness of a color image by performing contour enhancement processing in a well-balanced manner on each color signal. This is the purpose.

The present invention is characterized in that a sharp image signal and an unsharp image signal are extracted and combined for each color, so that edge enhancement processing is performed independently for each color signal. It has been confirmed through experiments that by performing edge enhancement processing on each color signal in this way, a color image with good balance between colors and significantly improved sharpness can be obtained.

When performing edge enhancement processing using the unsharp image signal and the sharp image signal for each of these three color signals, a three-color separation photometry system for sharp images is used to perform three-color separation photometry for sharp images, and a three-color separation photometry system for three-color separation photometry for sharp images, and a three-color separation photometry system for three-color separation photometry for sharp images, A three-color separation photometry system for unsharp images is required for color separation photometry, making the device complicated.

SUMMARY OF THE INVENTION Therefore, an object of the present invention is to provide a color image scanning device that can perform edge enhancement processing for each color signal using a simple device.

The present invention uses a solid-state image sensor in which a large number of elements are arranged in a line, and arranges a reduction optical system and blue, green, and red filters between a color original and the solid-state image sensor, and sequentially places each color filter in the optical path. By inserting a solid-state image sensor and repeatedly moving it for each color filter, reading a sharp image and a blurred image for each color with this solid-state image sensor, and combining the obtained sharp image signal and unsharp image signal. , contour enhancement processing is applied to each color signal, and then A/D conversion processing and logarithmic conversion processing are performed, and the result is written into the line memory (sofa memory).

Two of the solid-state image sensors are used, one of which is placed on the imaging plane of the reduction optical system to read a clear image, and the other is placed at a position slightly shifted forward from the imaging plane to read a blurred image. Alternatively, two solid-state image sensors may be placed on the image forming surface of the reduction optical system, and a long focal length lens or optical narrow filter may be placed in front of one of the solid-state image sensors to read the blurred image. good. Regarding a device that performs contour enhancement processing using these two solid-state image sensors, Japanese Patent Application Laid-Open No. 56-328
Details are given in issue 70.

In addition, a single solid-state image sensor is used, and a glass plate with a transparent part and a low-pass filter part formed thereon is provided in front of the solid-state image sensor, and these are alternately placed in front of the solid-state image sensor to generate an unsharp image signal and a sharp image signal. It is also possible to take out the image signals alternately, store the image signals outputted earlier in the memory, and combine them when the image signals outputted later. A device that performs edge enhancement processing using a single solid-state image sensor is described in detail in Japanese Patent Laid-Open No. 56-32869.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 2 schematically shows the optical system of the present invention. A color original (color original) 1 is illuminated by an illumination device 2, and reflected light from the color original 1 passes through a filter turret 3 and a variable magnification reduction optical system 4 and enters a half mirror 5. The reflected light is divided into two by this half mirror 5, one of which is incident on the solid-state image sensor 6 for blurred images, and the other is incident on the solid-state image sensor 7 for clear images. This solid-state imaging device 6 for blurred images is disposed at a position slightly shifted forward from the imaging plane of the reduction optical system 4, and outputs an unsharp image signal. The solid-state image sensor 7 for clear images includes a reduction optical system 4
It outputs a sharp image signal for one line of the color original 1.

The solid-state image sensors 6 and 7 are called line sensors, and have a structure in which a large number of elements (pixels) are formed in a line on one semiconductor substrate. Classified as OODXMO type 8, OPD. Each element generates an electric signal according to the amount of light received, and by electrically scanning each element, a signal for one line can be extracted in time series. Moreover, the solid-state image sensors 6 and 7 move intermittently in a direction substantially perpendicular to the element array on a plane parallel to the color original 1. If this mechanical movement of the solid-state image sensors 6 and 7 is referred to as sub-scanning (Y direction), and the electrical scanning is referred to as main scanning (X direction), the color original 1 is two-dimensionally scanned by this main and sub-scanning. scanned.

These solid-state image sensors 6 and 7 move together, and if the position of the solid-state image sensor 7 for clear images is used as a reference, they start moving from a movement start point (scanning origin) A.
During this movement, an image is read between a reading start point (scanning start point) B and a reading end point (scanning end point) C, which are predetermined with reference to the movement start point A. Note that the movement end point restricts the maximum movement range determined by design.

The filter turret 3 includes a blue filter 8, a green filter 9. Red filter 10. An ND filter 11 is provided, and any one of these filters 8 to 11 is selected and inserted on the optical path connecting the color original 1 and the reduction optical system 4. Note that these filters 8 to 11 may be replaced by a sliding type instead of a turret type.

When performing three-color separation photometry on the color original 1, for example, after inserting the blue filter 8 into the optical path, the solid-state image sensors 6 and 7 are moved one line at a time from the reading start point A to the reading end point C.
If the pixel is 13 μm×13 μm, it moves intermittently by 13 μm. Then, within the reading range between the reading start point A and the reading end point C, the color original 1 is passed through the blue filter 8.
The reflected light from the solid-state imaging devices 6 and 7 is measured. The solid-state image sensors 6 and 7 are electrically scanned in the main scanning direction, and the solid-state image sensor 6 for blurred images outputs a blue unsharp image signal, and the solid-state image sensor 7 for sharp images outputs a blue sharp signal. Output.

After reading the blue component of each pixel of the color original 1 by moving the solid-state image sensors 6 and 7 under the blue filter 8, the solid-state image sensors 6 and 7 are returned to the movement starting point A, and then the filter turret 3 to insert the green filter 9 into the optical path, and pass through the entire green filter 9 to measure the green component of each pixel again in the same reading range. After measuring this green component, the red filter 10 is inserted into the optical path to measure the red component.

In this way, the color original 1 is
Color-separated photometry is applied to each part of the image in a field-sequential manner. Note that the ND filter 9 is used when scanning a monochrome original.

By moving the variable magnification reduction optical system 4 and the solid-state image pickup device 10-6.7, the size of the image of the color original 1 changes, so the image of the color original 1 can be converted into a line shape at a desired reduction ratio. Can be read.

FIG. 3 shows the external appearance of the device of the present invention.

The original stage 12 is horizontally adjusted by four adjustment screws 13, and the color original 1 is placed on the original stage 12. In order to set the color original 1 flat without curling, it is held down by a band-shaped original presser or the like.

The color original 1 placed on the original stage 12 is illuminated from one side by illumination devices 2a to 2d provided at the tip of a letter-shaped arm 14a, and from the other side by illumination devices 28 to 2h provided at the tip of the arm 14b. illuminated from. These lighting devices 2a to 2h are shown as lighting device 2 in FIG.

A pair of pillars 15a, 15b are provided that extend in a direction perpendicular to the original drawing stand 12, and these pillars 15a, 1
A camera stand 16 is attached to 5b. This camera stand 16 is moved up and down along the supports 15a and 15b by a motor, chain, etc., as is known in color printers, etc., so that the reduction ratio can be set to an arbitrary value.

A camera head 17 is fixed on the camera stand 16, and the above-described filter turret 3. Reduction optical system 4. Further, reference numeral 18 designates an opening for guiding reflected light from the color original 1 to the camera head 17 .

The control box 20 houses a circuit device for controlling the camera head 17 and the like, and is provided with an operation panel 21 having a plurality of keys on its upper end surface and a CRT (cathode ray tube) 22. This 0RT22 is the operation panel 21
It is used to display data input from the computer or to display the operating means of the scanning device.

FIG. 4 schematically shows the apparatus of the present invention.

The aforementioned filter turret 3 is rotated by the drive motor 24, and a desired color filter is inserted into the optical path 25. The rotation of this drive motor 24 is controlled by a motor drive circuit 26, and this motor drive circuit 26 is controlled by a microcomputer 27. The position detector 28 detects that the color filter is correctly located in the optical path 25 and what color filter it is. This detection involves providing a plurality of holes on the outer periphery of the filter turret 3 to indicate the type of color filter and one hole indicating that the color filter is correctly positioned on the optical path, and detecting these holes with a photodetector. This can be done by Instead of the other holes, electrical contacts may be provided and detected using a brush.

An ND filter turret 29 for sensitivity adjustment is provided below the filter turret 3, and is rotated by a drive motor 30. The sensitivity adjustment ND filter turret 29 is provided with a plurality of ND filters having different transmittances, and by selecting one of them, the amount of light received by the solid-state image sensor 5 is adjusted. The drive motor 30 is driven by operating a key provided on the operation panel 21.

The reduction optical system 4 shown in FIG. 2 is provided within a lens barrel 31, and an infrared cut filter 32 and an aperture 33 are also provided within this lens barrel 31. As is well known, the aperture 33 is made up of a plurality of sector blades arranged concentrically, and by rotating the aperture ring 34, the aperture diameter formed by the inner edges of the sector blades changes. This aperture ring 34 is rotated by a drive motor 35, and the rotation of this drive motor 35 is controlled by operating a key on the operation panel 2.

A rack gear 31a is formed on the outer periphery of the lens barrel 31, and a gear 36 meshes with this rack gear 31a. This gear 36 is interlocked with a sector gear 38 via a ribbon 37. The sector gear 38 meshes with a rack gear 39, and this rack gear 39 moves when a roller 40 provided at its tip contacts a cam surface 42 of a cam plate 41, displacing the sector gear 38 according to the amount of movement. .

By operating the keys on the operation panel 21 and moving the camera head 17 upward or downward, the distance between the color document and the solid-state image sensors 6 and 7 changes, and the size of the image of the color document 1 changes. . Then, when the position of the camera head 17 changes, the rack gear 39 is slid by the cam plate 41, so that the sector gear 38. Ribbon 37. Gear 36.
The lens barrel 31 connects to the optical path 2 via the interlocking mechanism of the rack gear 31a.
Move along 5. As a result, the focus of the reduction optical system 4 is automatically adjusted, and the image of the color original 1 is formed on the solid-state image sensor 6.7.

The solid-state image sensor 6.7 is attached to a moving table 44. The moving table 44 is threaded onto a feed screw shaft 45 and is loosely fitted into a guide 46 . These feed screw shaft 45 and guide 46 are supported by a pair of bearings 47. In addition, the feed screw shaft 45 is equipped with a sub-scanning pulse motor 4.
8 are connected to each other, and a sub-scanning pulse motor 48 rotates by a constant angle with a pulse from a motor drive circuit 49.

The intermittent rotation of the feed screw shaft 45 causes the movable table 44 to move in the sub-scanning direction.

The number of pulses input to the sub-scanning pulse motor 48 is counted by a counter 50, and from the contents of this counter 50, the current positions of the solid-state image sensors 6 and 7 in the sub-scanning direction can be known. The contents of this counter 50 are sent to the microcomputer 27.

As mentioned above, the solid-state image sensor 6.7 moves in the sub-scanning direction for each color filter and performs three-color separation photometry, so in order to eliminate positional deviation of color signals, the same reading range is always used. It is necessary to read the color information. This reading range is defined by a reading start point B and a reading end point C, and these points B and O are the number of scanning lines from the movement start point A, that is, the solid-state image pickup devices 6 and 7 are defined by Bl.
It is determined by the number of pulses of the sub-scanning pulse motor 48 required to move to C. Whether or not this point B has reached 0 can be easily determined from the contents of the counter 50.

Therefore, in order to eliminate positional deviation and perform accurate three-color separation photometry, the movement start point A must be detected with high precision and the movement stage 4 must be
4 must be positioned at this point A. For this purpose, a limit switch 51 that detects the position of the moving table 44 in the moving direction and a potentiometer 52 that detects the rotational position of the feed screw shaft 45 are provided. This limit switch 51 roughly detects the position of the moving table 44, and the potentiometer 52 detects the rotational position of the feed screw shaft 45 within one rotation with high precision. The output signals of this limit switch 51 and potentiometer 52 are transmitted to a position detector 53.
is being sent to. The position detector 53 includes, for example, a potentiometer that specifies a reference rotational position, a comparator that determines whether the output voltage of this potentiometer matches the output voltage of the potentiometer 52, and an output signal of the comparator and a limit switch 51. It consists of an AND circuit that calculates a logical product with an output signal. When the output voltage of the potentiometer 52 reaches a predetermined value after the limit switch 51 is turned on, it is determined that the moving table 44 is located at the movement starting point A.

Furthermore, in order to move the movable table 44 with high precision, the amount of movement of the movable table 44 per pulse is made small, and the movable table 44 is moved one step using a plurality of pulses. Note that, instead of the limit switch 5, a proximity switch such as a reed switch, a photo sensor consisting of a light source section and a light receiving section, etc. can be used. Limit switch 54
is for detecting that the moving table 44 has reached the end point of its movement.

The solid-state image sensing devices 6 and 7 are electrically scanned by a drive circuit 55 controlled by a microcomputer 27, read the image recorded on the color original 1 in a line shape, and output it as a time-series signal.

The color signal of 1 rai/min outputted from the solid-state sensor 7 for clear images passes through an analog switch 57 controlled by a microcomputer 27, and then passes through an analog switch 57 for blue color.

Gain/offset setting circuit 5 for green, red, and black and white
It is sent to any one of numbers 8 to 61.

These gain/offset setting circuits 58 to 61 are provided with a variable resistor 62 for gain adjustment in their feedback circuit, and a variable resistor 63 for offset adjustment on one of the input terminals.

Similarly, the color signals read from the solid-state image sensor 6 for blurred images pass through an analog switch 64 for blue, green, and blue signals.
Gain/offset setting circuits 65 to 68 for red and black and white
sent to one of them.

An analog switch 69 is also connected to the output terminals of the gain/offset setting circuits 58 to 61, and the output signal of any one of the predetermined gain/offset setting circuits 58 to 61 is taken out in conjunction with the analog switch 57. . Similarly, an analog switch 70 is also connected to the output terminals of the gain/offset setting circuits 65 to 68.

The output signal of the analog switch 69 is input to the non-inverting input terminal of the operational amplifier 71, and the output signal of the analog switch 70 is input to the inverting input terminal, where they are subtracted.

The output signal of the operational amplifier 71 is sent to a sample hold circuit 72 and sampled in synchronization with the drive circuit 55. The color signal of each pixel extracted by the sample and hold circuit 72 is converted into a digital signal by an A/D converter 73.

The color signal converted into a digital signal is sent to the ROM 75 or RAM 76 selected by the changeover switch 74. These ROM 75 and RAM 76 are called logarithmic conversion table memories, and addresses are designated by digitized color signals, and data obtained by logarithmically converting the color signals is stored in each address. Here R
The OM75 stores general logarithmic conversion curve data and is commonly used for each color. The RAM 76 is capable of writing a desired logarithmic conversion curve, and includes four memory banks for storing logarithmic conversion curves for blue, green, red, and black and white, respectively.
It is read from the host computer 77 and written to the RAM 76. Note that this RAM 76 may be used commonly for each color.

Said logarithmic conversion table memo! The concentration level signal for one line logarithmically converted by J 75 and 76 is sent to one of the line buffer memories 79 and 80 selected by the selector 78.
be memorized. These line buffer memories 79.
Two selectors 81 . It is accessed by the host computer 77 via the interface 82, and one row's worth of density level signals are read out. A minicomputer is used as the host computer 77, which stores the captured density level signal for one line in an image memory and forms image files for each color. In addition, in order to reduce the number of image files, it is possible to read out the already stored density level signal and combine it with the density level signal of another color pixel by pixel to form a three-color combination signal and store it again. is desirable.

The host computer 77 and the microcomputer 2
An interface 83 is provided to synchronize with the computer 7 and to send and receive necessary commands and data.

Next, the operation of the above embodiment will be explained. The lighting device 2 is turned on and its direction etc. are adjusted so that the color original 1 is uniformly illuminated. Next, the camera head 17 is moved vertically to set the reduction ratio of the color original 1. When the camera head 17 moves, the reduction optical system 4 is extended along the optical path 25 to automatically adjust the focus. After setting the height of this camera head 17,
Adjust the aperture 33 by operating key 1. This aperture 33
An optical wedge is used for the adjustment, and the adjustment is made so that when the reflected light from the optical wedge is read by the solid-state image sensors 6 and 7, the output is not saturated and density levels can be distinguished. Next, the gains and offsets of the gain/offset setting circuits 58-61 and 65-68 are adjusted. At this time, the gain is adjusted so that the coefficients α and β in equation (1) become desired values. For example α−
In the case of 2°β-1/2, the gains of the gain/offset setting circuits 58 to 61 are set to be twice that of the gain/offset setting circuits 65 to 68.

Finally, operate the keys on the operation panel 21 to select the reading starting point B.
and set the reading end point C.

After performing these preparatory operations, the color original 1 is scanned in a field-sequential manner by operating operation keys to read a color image. First, the blue filter 8 is inserted into the optical path 25-1, and the moving base 44 for the blue filter 8 is intermittently moved from the movement start point A to the reading end point 0.

When the movable table 44 moves intermittently, the reflected light from the color original 1 that has passed through the blue filter 8 enters the solid-state image sensor 6.7, and the color original 1 is read in a line shape. The solid-state image sensor 6.7 is electrically scanned by a drive circuit 55, and the solid-state image sensor 6 for blurred images outputs an unsharp image signal, and the solid-state image sensor 7 for sharp images outputs a sharp image signal. Ru.

The /key 1 image signal is amplified by the blue gain/offset setting circuit 58 and then input to the operational amplifier 71.

On the other hand, the unsharp image signal is amplified by the blue gain/offset setting circuit 65 and then input to the operational amplifier 71, where subtraction is performed. This operational amplifier 71
A signal obtained by performing edge enhancement processing on the blue signal is outputted from the blue signal and sent to the sample hold circuit 72. A sample and hold circuit 72 samples and holds a peak value for each pixel, and this peak value is converted into a digital signal by an A/D converter 73.

The blue signal converted into a digital signal is sent to the logarithmic conversion table memory 75 or 7 selected by the changeover switch 74.
Logarithmically transformed at 6. After this logarithmic conversion, the line buffer memory 7 that is not occupied by the host computer 77
9 or 80, and a blue density level signal for one line is written here. This blue density level signal for one line is taken into the host computer 77 and stored in the image memory.

As described above, the color original 1 is read in a line, subjected to edge enhancement processing, A/D conversion processing, and logarithmic conversion processing, and then stored in the host computer 77.

When the scanning of the color original 1 using the blue filter 8 is completed, the moving stage 44 returns to the movement starting point A, and the green filter 9 is then inserted into the optical path 25. The color original 1 is scanned again using this green filter 9. In this case, the blue gain/offset setting circuits 59 and 66 are selected. Finally, the red filter 10 is inserted into the optical path 25, and the color original 1 is scanned again.

In the present invention having the above configuration, the outline enhancement process is performed independently for each color, so that the sharpness of each color is maintained and the sharpness of the reproduced color image is improved.

Further, since only one or two solid-state image sensors are required and there is no need to provide a solid-state image sensor for each color as in a three-color separation photometry system, there are advantages such as a simpler apparatus.

[Brief explanation of the drawing]

Fig. 1 is a graph showing the principle of contour enhancement processing, Fig. 2 is an explanatory diagram showing the optical configuration of the present invention, Fig. 3 is a perspective view showing the appearance of the device of the present invention, and Fig. 4 is a diagram showing the appearance of the device of the present invention. FIG. 2 is a block diagram showing the configuration. 1... Color original 2... Illumination device 3... Filter turret 4... Reduction optical system 5... Nof mirror 6...
- Solid-state imaging device for blurred images 7... Solid-state imaging device for clear images 8... Blue filter 9... Green filter 10...
・Red filter ]1...ND filter A...Movement start point B...Reading start point C...Reading end point
D...Movement end point 17 ・ ・ ・Camera head 20
・ ・ ・Roll box 21...Operation panel 24...Motor 25 ・・・Optical path 27-...Microcomputer 29...Aperture 44...Moving table 48...
Sub-scanning pulse motors 51, 54...Limit switch 52...Botejon meter 53...Position detector 55...Drive circuit 57, 6
4, 69, 70...Analog switches 58-61
, 65 to 68... Gain/offset setting circuit 71... Operational amplifier 72... Sample hold circuit 73... A/D converter 75, 76... Logarithmic conversion table memory 77...
Host computers 78, 81...Selectors 79, 80...Line buffer memories 82, 83
...Interface 27- Fig. 1 Japanese Unexamined Patent Publication No. 58-27145 (9)

Claims (3)

[Claims] (1) When performing three-color separation photometry on a color image recorded on an M document, sharp image signals and unsharp image signals are extracted and combined for each color, so that edge enhancement processing is performed for each color signal. A method for processing contour enhancement of a color image signal, characterized in that: (2) One or two solid-state image sensors that have a large number of elements arranged in a line and capture a clear image and a blurred image of a color original, and between the color original and the solid-state image sensor. A reduction optical system and blue, green, and red filters arranged, a color filter conversion means for selecting any one of the plurality of color filters and inserting it on the optical path, and a source element array for each color filter. means for moving the solid-state image sensor in a substantially perpendicular direction; means for synthesizing sharp image signals and unsharp image signals for each color output from the solid-state image sensor to perform edge enhancement processing; An A/D converter that converts the processed color signal into a digital signal, and a logarithmic conversion table in which addresses are specified by the output signal of this A/D converter and logarithmically converted data is stored in each address. A color image scanning device comprising a memory and a line buffer memory for storing one line of density level signals sequentially logarithmically converted by the logarithm conversion table memory. (3) The color image scanning device according to claim 2, wherein two solid-state image sensors are used, one of which reads a clear image of a color original, and the other reads a blurred image. .