JP2001169300A - Color shift detector - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a color shift detector which is capable of properly correcting a color shift quantity due to the difference of the number of pixels of a CCD in a still picture and the difference of an enlargement ratio in the still picture or correcting discoloration where a subject in the still picture is in an area of a fixed brightness or brighter than it. SOLUTION: In an electronic endoscope device, subject images subjected to R, G and B surface sequential illumination are sequentially picked up by the CCD 5 being an image-pickup means and the respective R, G and B images which are sequentially picked up are stored in a still picture memory 25 being a memory means and outputted to be displayed on a monitor 3, etc. The endoscope device has a discrimination means 21 for discriminating the kind of the CCD 5, an area setting means 27 for setting a specific area (size) in the range of each image to the respective successively picked up images in accordance with the discriminated kind (the number of pixels) of the CCD 5, and color shift information detecting means (26, 28 and 29) for detecting a minimum color shift image by obtaining color shift information between the respective images stored in the memory 25 based on the set area.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a color shift detector for detecting a color shift occurring between image signals of respective colors which are sequentially imaged by an image pickup means for imaging an object sequentially illuminated with illumination light of different colors of different colors. Related to the device.

[0002]

2. Description of the Related Art In an electronic endoscope apparatus for displaying a subject image in color based on an image pickup signal from a solid-state image pickup device (hereinafter referred to as a CCD) by a frame sequential illumination system, an R, G, and R filter is rotated to rotate R, G, and R. The subject is irradiated with each color light of B sequentially, and images are taken in time series for each color.

For this reason, conventionally, when a CCD is driven by a frame sequential method, image signals of each color of R, G, and B are fetched in a time series, and a time lag occurs until an entire image is formed. Therefore, if one or both of the subject and the imaging side move during shooting, a color shift occurs when the image signals of these colors are superimposed and displayed as a still image.

[0004] For this reason, proposals for obtaining a still image with less image deterioration such as color misregistration are described in Japanese Patent Application Nos. Hei 5-30460 and Hei 2-286125.

[0005] The apparatus described in Japanese Patent Application No. Hei 5-30460 converts an input image signal obtained when a minimum value is obtained from a corrected amount of motion within a predetermined time after activation by the still image instruction means into a still image. It is stored as a picture. According to the publication, in order to obtain an image with little color shift, after selecting a still image, an image with the least color shift is selected within a predetermined time and output to a monitor.

Further, the apparatus described in Japanese Patent Application Laid-Open No. 2-286125 temporarily stores an image signal in an image storage means based on a freeze instruction signal, and when the color shift amount of the stored image signal is within a set value. Holds the image signal, cancels the storage when the value is equal to or more than the set value, and performs the same operation on the image signal after the cancellation to obtain the image signal having the color shift amount within the set value. Is stored in the image storage means as a still image, and is output to and displayed on a monitor.

[0007]

However, in any of the apparatuses described in the above publications, an image with little color shift output to the monitor is not configured to correct the color shift. It did not go away. In addition, since the generated color shift differs depending on the type of CCD or the number of pixels, the color shift does not completely disappear in the conventional method. For example, a CCD having a larger number of pixels has a larger color shift amount.

Further, the conventional color misregistration correction device does not always perform color misregistration correction at the region of interest of the observer, but performs color misregistration correction outside the region of interest of the observer. For example, when there is a dark subject in the center of the screen,
The color shift is corrected for a dark subject, and when the region of interest of the observer is a bright surrounding object, the color shift for the bright surrounding object cannot be corrected. Also,
When the enlargement ratio of an image is different optically or electrically, the amount of color shift differs according to the enlargement ratio. Therefore, the conventional method cannot correct the color shift, and the larger the enlargement ratio, the larger the amount of color shift. was there.

In view of the foregoing, an object of the present invention is to provide a color misregistration detecting device capable of appropriately correcting a color misregistration amount due to a difference in the type of CCD or the number of pixels in a still image. It is assumed that.

It is another object of the present invention to provide a color misregistration detecting device capable of correcting a color misregistration in a region where a subject in a still image has only a constant brightness.

It is a further object of the present invention to provide a color misregistration detecting device capable of appropriately correcting the amount of color misregistration due to a difference in the enlargement ratio at the time of a still image.

[0012]

According to a first aspect of the present invention, there is provided a color misregistration detecting apparatus, comprising: an image pickup means for picking up a subject image in a frame-sequential manner; Memory means for storing each image forming an image, determination means for determining the type of the imaging means,
According to the type of the imaging unit determined by the determination unit, for each of the images sequentially imaged,
An area setting means for setting a specific area within the range of each image; and a color shift detecting color shift information between the images stored in the memory means based on the area set by the area setting means. And information detecting means.

According to the first aspect of the present invention, the CC at the time of a still image
It is possible to appropriately correct the amount of color shift due to the difference in the type of D and the number of pixels.

According to a second aspect of the present invention, there is provided a color misregistration detecting apparatus, wherein: an image pickup means for picking up a subject image in a frame-sequential manner;
Memory means for storing each image forming one image, brightness detection means for detecting a portion having a brightness level equal to or higher than a predetermined value among the imaging signals from the imaging means, and detection by the brightness detection means. Area setting means for setting a predetermined area including the part within the range of each of the images sequentially imaged in accordance with the part, and storing in the memory means based on the area set by the area setting means. And color shift information detecting means for detecting color shift information between the respective images.

According to the second aspect of the present invention, it is possible to correct a color shift in a region where a subject at the time of a still image has only a constant brightness.

According to a third aspect of the present invention, there is provided a color misregistration detecting apparatus, wherein: an image pickup means for picking up a subject image in a frame-sequential manner;
Memory means for storing each image forming one image, enlargement magnification change means for changing the enlargement magnification of the video imaged by the imaging means and displayed on the display means, and the enlargement magnification in the enlargement magnification change means. Accordingly, for each of the images sequentially imaged sequentially, an area setting means for setting a specific area within the range of each image, and the memory means based on the area set by the area setting means. Color shift information detecting means for detecting the stored color shift information between the images.

According to the third aspect of the present invention, it is possible to appropriately correct the amount of color misregistration due to the difference in the enlargement ratio at the time of a still image.

[0018]

Embodiments of the present invention will be described with reference to the drawings. Before explaining the first and second embodiments of the present invention with reference to FIGS. 1 to 7, an electronic endoscope apparatus to which an apparatus for detecting and correcting a color shift of the present invention is applied with reference to FIG. Will be described.

FIG. 8 shows the overall configuration of the photographing system of the electronic endoscope apparatus. As shown in FIG. 8, the endoscope apparatus according to the present embodiment includes a processor 1 and an endoscope 2. The processor 1 includes a CCD driving circuit 6, a signal processing unit 8, and a surface sequential light source device 14. The endoscope 2 includes a light guide 9 and an illumination lens 10, an objective lens 4, a CCD 5, and a buffer amplifier 7. A light guide 9 for transmitting illumination light from a surface-sequential light source device 14 is inserted through an insertion portion of the endoscope 2 (in FIG. 8, the length is considerably shorter than the actual length).
Further, at the image forming position of the objective lens 4, CC
D5 is provided. On the other hand, the light source device 14 in the processor 1 is provided with a lamp 17 that emits light in a wide band from ultraviolet light to fluorescence observation. As the lamp 17, a general xenon lamp or strobe lamp is used, which emits not only visible light but also a large amount of ultraviolet light and fluorescent light. The lamp 17 is supplied with power from a power supply unit (not shown). An RGB rotation filter 19, which is driven to rotate by a motor 20, is provided in front of the lamp 17. The rotation filter 19 includes:
Filters that transmit light of each wavelength region of red (R), green (G), and blue (B) are arranged along the circumferential direction. A stop 1 is provided between the lamp 17 and the rotary filter 19.
Reference numeral 5E is provided, and the aperture amount is controlled by the aperture control means 15 so that the amount of light emitted from the lamp 20 can be adjusted. The rotation of the motor 20 is controlled by the RGB rotation filter control means 16 and the motor 20 is driven. Transmitted through the RGB rotation filter 19,
The light divided in time series into light of each wavelength region of R, G, and B is a condensing lens 1 that collects illumination light of the surface-sequential light source device 14.
The light is incident on an incident end of the light guide 9 via the light guide 8, is emitted from the illumination lens 10 via the light guide 9, and illuminates the observation site. The light from the observation site (subject) illuminated by the illumination light forms an image on the CCD 5 by the objective lens 4,
It is designed to be photoelectrically converted. A driving pulse from the CCD driving means 6 in the processor 1 is applied to the CCD 5 via a signal line.
The charge is read and transferred. Then, the video signal read from the CCD 5 is current-amplified via a buffer amplifier 7 arranged in the endoscope 2, subjected to various image processing by a signal processing unit 8, and output to the monitor 3. The signal processing unit 8 includes a signal processing unit 11 that performs processes such as enlargement and contour enhancement, and a time-series R,
There is a frame sequential signal synchronizing means 12 for temporarily storing and synchronizing the G and B video signals in the respective R, G and B synchronizing memories via the selectors, and there are instructions for the respective R, G and B still images. It is composed of a still image signal processing means 13 for processing the R, G, and B outputs of the synchronizing means 12 with a still image.

[First Embodiment] [Structure] FIG. 1 shows a structure diagram of an apparatus for detecting and correcting color misregistration according to a first embodiment of the present invention, and FIGS.
The principle of color shift detection will be described.

In the overall configuration diagram shown in FIG. 8, every time the number or type of pixels of the CCD 5 provided at the end of the endoscope 2 is different, color shift detection and correction are performed to correct color shift peculiar to frame sequential. An embodiment in which the color shift search area is made variable by an apparatus for performing the above will be described with reference to FIGS.

As shown in FIG. 1, the signal processing means 11 shown in FIG.
DS circuit, clamp circuit (not shown) for clamping an optical black (OB) unit, A / D circuit 22 for converting an analog signal to a digital signal, enlargement circuit (not shown) for enlarging an image, enhancement (not shown) for edge enhancement of an image It is composed of circuits and the like.

The plane-sequential synchronizing means 12 shown in FIG. 8 is a synchronizing means for converting a plane-sequential signal from the CCD 5 as an image pickup means into R, G, B synchronizing signals, as shown in FIG. A circuit 23 is provided.

The still picture signal processing means 13 shown in FIG.
As shown in FIG. 1, a freeze instructing circuit 24 for instructing a freeze start when a still image (hereinafter, freeze) is selected;
CCD discriminating circuit 21 for discriminating the number or type of pixels of CD5
And a color misregistration search area setting circuit 27 for switching a color misregistration search area range in accordance with a result of the discrimination from the CCD discrimination circuit 21. Freeze memory) 25
Based on a correction freeze start signal from a correction freeze selection means (not shown), a part (for example, about 1/4) of the area stored in the freeze memory 25 is reduced to R, G, Based on the blur correction memory 26 stored for each B and the color shift search area from the color shift search area setting circuit 27, the blur correction memories R, G, By storing the readout address of the blur correction memory of the R, G, B image with the least color shift from the B output signal,
Minimum color shift detection circuit 29 for detecting information on the minimum color shift
And the freeze memory 25 and the shake correction memory 2
6 memory R / W for controlling read / write (R / W)
A controller 28; switching means 30 for switching and outputting a moving image from the synchronization circuit 23 and a still image from the freeze memory 25; and D / A conversion for converting image data into analog signals for each of R, G, and B. Means 31.

The memory R / W controller 28, together with the blur correction memory 26 and the color shift minimum detection circuit 29, stores the areas of the blur correction memory 26 based on the color shift search area set by the color shift area setting circuit 27 into R and G. , B, and retrieves the retrieved data for each of R, G, and B into the color misregistration minimum detection circuit 29 to detect color misregistration information between the R, G, and B images. ). Further, the memory R / W controller 28 reads out and controls the freeze memory 25 based on the color misregistration minimum value information from the color misregistration minimum detection circuit 29 to correct the color misregistration of R,
It has a color misregistration correction function (means) for outputting G and B signals.

The color misregistration search area setting circuit 27 includes C
In accordance with the size (the number of pixels) of the CD5, an area to be searched for a color shift is set.
Based on the D determination information, search area information of a size corresponding to the number of CCD pixels is sent to the memory R / W controller 28. The memory R / W controller 28 sets a search area for the image data recorded in the blur correction memory 26 based on the input search area information. This makes it possible to set a color shift area according to the size (the number of pixels) of the CCD 5.

The blur correction memory 26 stores a part (for example, about 1/4) of the area stored in the freeze memory 25 for each of R, G, and B as described above. Since the color misregistration is detected only in a partial area of about 1/4, the color misregistration search time is reduced as compared with the case where the color misregistration is detected in all the areas stored in the freeze memory 25.

The color misregistration minimum detection circuit 29 determines the difference between the G image waveform and the R image waveform (or B image waveform) for the R, G, and B image data read from a certain read address in the blur correction memory 26. When the R image (or B image) is delayed or advanced with respect to the G image, the color misregistration frequency is detected as a large number, and the R image (or B image) is detected with respect to the G image. The smaller the delay / advance of the image, the smaller the color misregistration frequency is detected as a smaller number, and R (or B) from the R (or B) memory based on the G signal stored in the G memory of the blur correction memory 25. B) The signal read address is relatively shifted, and the difference (color shift frequency) between the R (or B) signal read for each address and the reference (address fixed) G signal.
R (or B) for detecting and obtaining the minimum color shift frequency
It is configured to detect a signal read address (ie, color misregistration minimum image information).

Next, a configuration example of the color misregistration minimum detection circuit 29 will be described with reference to FIGS. First, the overall configuration of the color misregistration minimum detection circuit 29 will be described with reference to FIG.

R, output from the blur correction memory 26,
The G and B signals are input to input terminals 291, 292, and 293, respectively. The input R signal is directly input to the comparison circuit 297, and is also delayed by one clock in a one-clock (1CK) delay circuit 294 and input to the comparison circuit 297, where
Then, a difference signal ΔR between the two input signals is obtained.
Assuming that the input signals of the comparison circuit 297 are R and R ', the comparison result .DELTA.R is three output signals of R>R', R = R 'and R <R', that is, a ternary signal (for example, +, 0). ,-). Similarly, for the input G and B signals, a difference signal ΔG between the G signal and a signal obtained by delaying the G signal by one clock by a one-clock (1CK) delay circuit 295 is obtained by a comparison circuit 298, and the B signal and this signal are obtained by a comparison circuit 299. Is delayed by one clock (1CK) delay circuit 296 to obtain a difference signal ΔB. The difference signals ΔG and ΔB of the G and B comparison circuits 298 and 299 are also output as ternary signals.

Each of the difference signals ΔR, ΔR of the comparison circuits 297, 298
G is input to the non-coincidence circuit 300, and the coincidence or non-coincidence of ΔR with ΔG is detected. Similarly, comparison circuits 298 and 29
The difference signals ΔG and ΔB of FIG. 9 are input to the mismatch circuit 310, and the match and mismatch of ΔB with ΔG are detected. The detection signals FL1 and FL2 of the non-coincidence circuits 300 and 310 are ORed by a two-input OR circuit 320, the correlation between the R, G, and B signals is obtained, and the correlation is output as a color shift frequency. . Then, the color shift frequency in the color shift search area is counted by a color shift frequency calculation circuit 321 composed of a counter,
The count output is supplied to the color misregistration minimum image detection circuit 322. The color misregistration minimum image detection circuit 322 obtains the minimum color misregistration frequency among the color misregistration frequencies of the R (or B) signals counted in the color misregistration search area, and reads the R (or B) image signal at that time. The address is output as color shift minimum image information.

That is, the color misregistration minimum detection circuit 29 calculates the difference between the signal waveform of the G image and the R image (or B image), and determines whether the R image (or B image) is behind the G image. Or, the color misregistration frequency is accumulated according to the difference (that is, the difference between R and G, the difference between B and G). Then, in the color misregistration search area, the respective areas of the R image and the B image are shifted in the horizontal (H) direction and the vertical (V) direction based on the G image, and the color misregistration is detected for each of the shifted R and B areas. In the accumulated color misregistration frequency data,
The color misregistration frequency minimum value that minimizes the color misregistration is selected, and the read start position of the freeze memory 25 is controlled in the H and V directions at that time under the control of the memory R / W controller 28.

FIG. 3 shows a specific circuit configuration example of the mismatch circuits 300 and 310 in FIG. Since the outputs ΔR, ΔG, and ΔB of the respective comparison circuits 297, 298, and 299 of R, G, and B in FIG. 2 take ternary signals (+, 0, and −) as described above,
This ternary state can be expressed, for example, in two bits [for example, “+” is expressed as (10), “0” is expressed as (00), and “−” is expressed as (01)]. In the case of 2-bit data as described above, generally, the first bit of ΔR is ΔR
1 and the second bit are represented by ΔR2. Similarly, ΔG represents the first bit as ΔG1, the second bit as ΔG2, and ΔB as the first bit as ΔB1 and the second bit as ΔB2. The non-coincidence circuits 300 and 310 can be configured as shown in FIG. That is, the mismatch circuit 300 includes two EX-OR circuits 301 and 302 and an OR circuit 303, and the mismatch circuit 310 includes two EX-OR circuits 311 and 312.
And an OR circuit 313.

FIG. 4 shows an example of a method for detecting a color shift (color shift detection signal generation) in the color shift minimum value detecting circuits of FIGS. 2 and 3.

Reference symbols R, G, B, ΔR, Δ in FIG.
G, ΔB, FL1, FL2, and FL3 are shown in FIG. 2 (and FIG. 3).
Same as above. For the time shown on the horizontal axis,
The levels of the image signals R, G, and B shown on the vertical axis change with time. That is, when the R, G, and B levels move with respect to the time axis, changes (+, 0, −) occur in the difference signals (ΔR, ΔG, ΔB) at a predetermined sample interval. In general, R, G, and B are shifted from each other by a plurality of sample intervals (a plurality of clock times), but the sample interval shown in FIG. This corresponds to a photographing signal delay time (one clock delay time) corresponding to each rotation time difference.
There is a shift of one sample interval between R and G and between B and G in the color shift search area. When R and B are shifted by one sample interval with G as a reference (center), matching is achieved, that is, the color shift is minimized. Can be. Exclusive OR of ΔR and ΔG (EX-O
R) The output FL1 indicates the state of color misregistration between R and G, and Δ
The exclusive OR (EX-OR) output FL2 of B and ΔG is B
And G color shift states. Therefore, FL1 and FL
A logical sum (OR) output FL3 of two is output as information indicating the total color shift state of R, G, and B.

[Operation] FIG. 5 is an operation timing chart from when a normal freeze instruction is issued to when a color shift minimum detection output is obtained. 5A shows a normal freeze selection instruction, FIG. 5B shows a state of the freeze memory 25, and FIG.
Indicates a freeze-selection instruction after color shift correction, and (d) indicates a color shift minimum image detection output from the color shift minimum detection circuit 29.

When a freeze is selected, writing and reading are performed from the memory R / W controller 28 based on a normal freeze instruction signal (see FIG. 5A) output from a freeze selecting means (not shown) in the endoscope 2. Control the behavior,
An image having the least motion within a predetermined time after the freeze is selected is stored as a still image for one screen in the freeze memory 25 (see FIG. 5B) and output from the monitor 3.

Thereafter, if a color misregistration is still seen in the still image, a color misregistration correction freeze instruction is issued from the freeze instruction circuit 24 in response to an instruction from a correction freeze selection means (not shown) in the endoscope 2. 5 (see FIG. 5C), the memory R / W controller 28 based on the color misregistration search area set by the color misregistration search area setting circuit 27 in accordance with the discrimination signal output from the CCD discrimination circuit 21. Searches the still image area written in the blur correction memory 26 for each of R, G, and B, and inputs the searched data of R, G, and B to the color misregistration minimum detection circuit 29.

The color misregistration minimum detection circuit 29 calculates the color misregistration frequency, which is the color misregistration value of the R image and the B image with respect to the reference G image, and stores the R (or B) in the blur correction memory until the color misregistration is minimized. The shift of the read address of the memory is sequentially repeated, and the read signal at the shift address of the R (or B) memory is compared with the G signal, and the memory read address for the R and G images with the least color shift is stored. I do. That is, step S1 in the flowchart of FIG.
According to steps S7 to S7, the R (or B) signal is relatively shifted with respect to the G signal. The read address of the R (or B) memory is shifted, and the R signal (or B signal) of the shifted address matches the G signal, and the R (or B) of the blur correction memory until the color shift is minimized. B) The shift of the read address of the memory is sequentially repeated, and the read signal at the shift address of the R (or B) memory is compared with the G signal. Then, the memory R / W controller 28 stores the memory read addresses for the R and G images with the least color misregistration and outputs them to the memory R / W controller 28, so that the memory R / W controller 28 , B can be read out from each other and output to the monitor 3 via the D / A conversion means 31.

In this embodiment, the search pixel interval for R and G is set to one pixel. Five locations in the horizontal direction and five locations in the vertical direction, for a total of 25 locations (that is, when the search pixel interval is 1
A color misregistration search is performed in a pixel area (25 pixels).
When the number of pixels of the CCD 5 increases, the amount of color misregistration (the number of color misregistration pixels) also increases. Therefore, the search pixel interval is widened and changed to about four pixels. Therefore, when the number of pixels of the CCD 5 is increased and the search pixel interval is set to 4 pixels, a color misregistration search is performed at a total of 25 locations, 5 locations in the horizontal direction and 5 locations in the vertical direction.
The color misregistration search area is set to be approximately 16 times larger.

In this way, even when the type of CCD 5 or the number of pixels used is different, for example, when the number of CCD pixels is increased, the color shift search area is enlarged and the search pixel interval is widened to detect the color shift. Then, control can be performed so that color misregistration is minimized with the same number of search points. Also,
By setting the size of the color misregistration search area according to the type of the illumination light source device (for example, whether or not observation is performed by fluorescent illumination), control is performed such that the color misregistration is minimized in the illumination mode at that time. It is good also as a structure which performs. Further, the freeze memory 25 and the blur correction memory 26 may be configured as the same memory. Note that the color misregistration detection method may be performed by a method of calculating a color variance value. Further, a plurality of color misregistration search areas having different sizes may be set to perform color misregistration detection and correction.

[Effect] As the number of pixels of the CCD increases, the amount of color misregistration increases. Therefore, by setting the size of the color misregistration search area in accordance with the type of CCD or the number of pixels, even if the number of pixels of the CCD increases, good color misregistration detection and correction can be performed with the same number of search locations.

[Second Embodiment] FIG. 7 shows a second embodiment of the present invention.
1 is a configuration diagram of an apparatus for detecting and correcting color misregistration according to the embodiment. The principle of color misregistration detection is the same as that shown in FIGS.

[Configuration] In the overall configuration diagram shown in FIG. 8, an embodiment in which a search area for a color shift peculiar to frame sequential is made variable based on a shift vector in accordance with the brightness of the screen is shown in FIG. This will be described with reference to FIGS.

The plane sequential synchronizing means 12 shown in FIG.
As shown in FIG. 7, a synchronizing circuit 23 for converting a frame sequential signal from the CCD 5 as an image pickup means into R, G, B synchronizing signals is provided.

The still picture signal processing means 13 shown in FIG.
7, a freeze instructing circuit 24 for instructing a freeze start when a freeze is selected, a screen brightness detecting circuit 33 for detecting a portion whose brightness level (brightness level) is equal to or more than a predetermined value, and a screen brightness detecting A color misregistration search area setting circuit 27 for switching a color misregistration search area range including the part in accordance with a detection signal from the circuit 33, and a still image memory (hereinafter, referred to as a memory) for storing image signals of R, G, and B colors during a freeze. Based on the freeze memory 25 and a correction freeze start signal from a correction freeze selection means (not shown), a part of the area (for example, about 1/4) stored in the freeze memory 25 is shortened in order to shorten the color misregistration search time. R,
Based on a blur correction memory 26 for storing each of G and B and a color shift search area from the color shift search area setting circuit 27,
By storing the read addresses of the R, G, and B images with the least color shift from the output signals of the shake correction memories R, G, and B obtained by searching the area of the shake correction memory 26 for color shift. , A color shift minimum detection circuit 29 for detecting information on a color shift minimum value, and reading / writing (R) of the freeze memory 25 and the blur correction memory 26.
/ W), a switching means 30 for switching and outputting a moving image from the synchronization circuit 23 and a still image from the freeze memory 25, and image data for each of R, G and B. D / to convert to analog signal
A conversion means 31 and a shift vector calculation circuit 34 for calculating a shift vector of R (or B) with respect to G of each color of R, G, B in order to further shorten the color shift search time as compared with the first embodiment. It is composed of

The memory R / W controller 28, together with the shake correction memory 26 and the color shift minimum detection circuit 29, stores the areas of the shake correction memory 26 based on the color shift search area set by the color shift area setting circuit 27 into R and G. , B, and retrieves the retrieved data for each of R, G, and B into the color misregistration minimum detection circuit 29 to detect color misregistration information between the R, G, and B images. ). Further, the memory R / W controller 28 reads out and controls the freeze memory 25 based on the color misregistration minimum value information from the color misregistration minimum detection circuit 29 to correct the color misregistration of R,
It has a color misregistration correction function (means) for outputting G and B signals.

[Operation] When the freeze is selected, the write operation is controlled by the memory R / W controller 28 based on a still image instruction signal output from a still image selecting means (not shown) in the endoscope 2 (not shown). Then, an image with the least motion within a predetermined time is stored as a still image for one screen in the freeze memory 25 and output from the monitor 3.

Thereafter, if the still image still shows color misregistration, the color misregistration search area setting circuit 27 is operated based on the freeze instruction circuit 24 in the endoscope 2 to perform color misregistration search. The area is set, and the memory R / W controller 28 controls the writing operation from the freeze memory 25 to the blur correction memory 26 to store the image data of the color misregistration search area in the blur correction memory 26.

In this embodiment, the screen brightness detection circuit 3
The image data in a predetermined area corresponding to the part detected in step 3 is read out from the freeze memory 25 and
To memorize. Then, in the color misregistration minimum detection circuit 29, R
For G and G, the search pixel interval in the color shift search area is set to 5 pixels, and the approximate color shift minimum value is obtained. Then, the search pixel interval is set to one pixel, and the minimum color shift value is selected by a method of obtaining an appropriate minimum color shift value. As a result, a more accurate color shift minimum image can be found.

The color shift minimum detection circuit 29 converts the image data stored in the blur correction memory 26 into R (or B) based on the G signal based on the shift vector output from the shift vector calculation circuit 40. While relatively shifting the signals, the read address of the blur correction memory 26 is shifted in the delay direction (or the advance direction) by an address shifter (not shown) for each screen, and the address of the address thus shifted is shifted. R signal (or B signal) and G
Until the signals match, the shift of the read address is sequentially repeated, and the read R (or B) signal at the shifted address is compared with the G signal. Thus, R
When the signal (or B signal) is registered with the G signal, the address of the R (or B) signal with the least deviation from the G read signal at that time is stored, and finally, Memory R / W controller 2
The R, G, and B images that minimize the color misregistration are read from the freeze memory 25 by the read control of step 8.

The operation of the color misregistration minimum detection circuit 29 is shown in FIGS.
This is as described in FIG. That is, the difference between the signal waveform of the G image and the R image (or the B image) is calculated, and if the R image (or the B image) is behind or ahead of the G image, the difference (that is, the difference) The difference between R and G,
The color misregistration frequency is accumulated according to the difference between B and G).
Then, in the color misregistration search area, the R image,
Each region of the B image is moved in the horizontal (H) direction and the vertical (V) direction, and the color misregistration frequency that minimizes the color misregistration in the color misregistration frequency data integrated for each of the shifted R and B regions. A value is selected, and the reading start position of the freeze memory 25 in the H direction and the V direction at that time is controlled by the control of the memory R / W controller 28.

By setting a color shift search area corresponding to a portion where the brightness level (brightness level) in the screen is equal to or greater than a predetermined value in this manner, the color shift in an image of a constant brightness which is noticed by the user. Is controlled to be minimum.

[Effect] A color misregistration in a region in the screen of the user's attention can be detected and corrected with a simple configuration.

[Third Embodiment] FIG. 9 shows a third embodiment of the present invention.
1 is a configuration diagram illustrating an electronic endoscope apparatus to which an apparatus for detecting and correcting color misregistration according to an embodiment is applied.

[Configuration] As shown in FIG. 9, the electronic endoscope apparatus according to the third embodiment has a light source device 41 for emitting light for observation and a scope 42 for insertion into a body cavity.
, A processor 43 that performs signal processing of image signals obtained by the imaging means, a monitor 44 that displays an image, and a photographing device 45 that records the image on a photographic film.

The light source device 41 includes a lamp 46 for emitting light.
And a rotary filter plate 47 provided on the illumination light path of the lamp 46 to limit the transmission wavelength.

The scope 42 includes a light guide 48 for passing illumination light, a CCD 49 as an image pickup means for picking up light from a subject, a lens 50, and an actuator 51 for moving the lens 50. The operation unit for operating the scope 42 includes a freeze switch 52 for instructing a still image (freeze), a release switch 53 for instructing recording on the photographing device 45, and an actuator in a position where the user can easily press it. An enlargement switch 54 which is an enlargement ratio changing means for instructing the movement of 51 to the "enlargement" side.
And a wide-angle switch 54 for instructing the actuator 51 to move to the “wide-angle” side.

The processor 43 includes a CCD driving circuit 56
A pre-processing circuit 57, an A / D conversion circuit 58,
Selector 59 and three synchronization memories 60a-60c
And three freeze memories 61a to 61c as memory means for still images, and three blur correction memories 62a to 62
c, three D / A conversion circuits 63a to 63c, an actuator control circuit 64, and a memory address control circuit 65
A freeze control circuit 66 and a color misregistration minimum search circuit 67
And

The memory address control circuit 65, together with the blur correction memories 62a to 62c and the color misregistration minimum search circuit 67, controls the blur correction memories 62a to 62c based on the color misregistration search area set based on the enlargement ratio of the enlargement switch 54. Color shift information detection for searching for each of R, G, and B, taking the searched data for each of R, G, and B into a color shift minimum search circuit 67 to detect color shift information between the R, G, and B images. It constitutes a function (means). Also, the memory address control circuit 6
A color shift correction function 5 controls reading of the freeze memories 61a to 61c based on the color shift minimum value information from the color shift minimum search circuit 67 and outputs R, G, and B signals with corrected color shift. )have.

[Operation] The light emitted from the lamp 46 of the light source device 41 passes through the rotary filter plate 47 and enters the light guide 48 of the scope 42. The rotary filter plate 47 is driven to rotate at a predetermined speed by a motor (not shown), so that an R filter, a G filter (not shown),
The B filters are sequentially placed on the optical path, and red, green, and blue lights are sequentially transmitted.

The light incident on the light guide 48 of the scope 42 is applied to the subject such as the digestive tract from the distal end of the scope. The light scattered and reflected by the subject forms an image on the CCD 49 at the tip of the scope. The CCD 49 is driven by a CCD drive circuit 56 in synchronization with the rotation of the rotary filter plate 47, and picks up an image of the CCD 49 when illuminated by light transmitted through each filter of the rotary filter plate 47 such as an R filter, a G filter, and a B filter. The signal is sent to the processor 4 sequentially.
3 is output.

The imaging signal input to the processor 43 is
First, it is input to the pre-processing circuit 57. The pre-processing circuit 57 extracts an image signal (pixel signal) from an image pickup signal of the CCD by a process such as CDS (correlated double sampling). The signal output from the pre-processing circuit 57 is converted from an analog signal to a digital signal by the A / D conversion circuit 58.

The signal output from the A / D conversion circuit 58 is sorted by the selector 59, and the image when the R filter, G filter, and B filter are inserted is output from the synchronization memory R60a, the synchronization memory G60b, and the synchronization memory, respectively. It is stored in the memory B60c. Each synchronization memory 60a-6
The images stored in 0c are read out at the same time, so that the frame sequential images are synchronized. The synchronized image is temporarily stored in the freeze memories 61a to 61c and then read out to adjust the position to be displayed on the monitor 44. This adjustment is performed in the freeze memory 61a-
This is performed by controlling the write timing and read timing of 61c. When the freeze operation is not performed, a conversion for correcting the gamma characteristic of the monitor 44 is performed in a gamma correction circuit (not shown), and is converted into an analog signal by the D / A conversion circuits 63a to 63c. Is displayed. On the monitor 44, a moving image is normally obtained in real time, and a still image can be observed when the operator performs a freeze operation.

When the operator presses the enlargement switch 54, an enlargement instruction signal is input to the actuator control circuit 64, and the actuator control circuit 64 drives the actuator 51 to move the lens 50 to the enlargement side. When the wide-angle switch 55 is pressed, a wide-angle instruction signal is input to the actuator control circuit 64, and the actuator control circuit 64 drives the actuator 51 to move the lens 50 to the wide-angle side. In this way, the operator can observe on the monitor 44 at various magnifications. The actuator control circuit 64 includes an actuator position detection circuit (not shown), and generates and outputs an enlargement ratio signal based on the position detection of the actuator 51. The enlargement ratio signal is sent to the memory address control circuit 65.

Next, the operation at the time of freeze will be described. When the operator presses the freeze switch 52, a freeze instruction signal is input to the freeze control circuit 66. The freeze control circuit 66 compares the color shifts of the R, G, and B image signals synchronized by the synchronization memories 60a to 60c, and stops writing to the freeze memories 61a to 61c at the timing with the least color shift. As a result, the colors R, G, and G that have the least color shift within a predetermined time after the freeze instruction signal is input.
A still image of the B combination is output. Freeze control circuit 66
Then, after the freeze image is determined, a correction instruction signal is output to the memory address control circuit 65. When the correction instruction signal is sent to the memory address control circuit 65, the memory address control circuit 65 controls the freeze memory 6
The data of 1a to 61c is read and the blur correction memory 62
a to 62c.

The color misregistration minimum search circuit 67 performs a color misregistration minimum address search on the image areas of the blur correction memories 62a to 62c in accordance with the flowchart of FIG. Although the color misregistration minimum search is performed for the R signal and the G signal, and for the G signal and the B signal, FIG. 10 shows only the operation for the R signal and the G signal.

First, as shown in step S11 of FIG. 10, the actuator control circuit 64 based on the above-described enlargement instruction is used.
Memory address control circuit 6 according to the enlargement ratio signal from
In 5, the search pixel interval c related to the color misregistration search area is determined and set for the blur correction memories 62a to 62c. Here, when the enlargement ratio is large, the search pixel interval c is set large, and when the enlargement ratio is small, the search pixel interval c is set small. In the present embodiment, 25 color misregistration searches are performed at five locations every c pixels in the horizontal direction (variable i) and at five locations every c pixels in the vertical direction (variable j). Therefore, when the enlargement factor is large, the color misregistration search area on the blur correction memories 62a to 62c is set large, and when the enlargement factor is small, the color misregistration search area is set small. In steps S12 to S14, for the color misregistration search area set in the R-blur correction memory 62a, sample search is performed at, for example, 25 places every c pixels after setting 0 → i, 0 → j, and an image is read. This is similar to the sample search described in FIG.

Then, as shown in step S15, R
In order to search for the minimum color shift once in the image area of the blur correction memory 62a, the G and B blur correction memories 62b and 62c are used.
Is fixed, the read address of the R shake correction memory 62a is shifted, and the image is read.

Next, in step S16, an image is read from the G shake correction memory 62b. For reading this image, a sample search is performed at, for example, 25 locations every c pixels, and the image is read in the same manner as described above. Then, steps S17 and S17
In S18, a color shift between the R image read in S12 to S14 and the plurality of R images read by address shifting in S15 and the G image read in S16 is detected, and based on the color shift information. Then, the evaluation value of the color shift is calculated, and the read address of the R-blur correction memory 62a where the color shift is minimized is stored. The above is the color shift minimum search operation by coarse search detection every c pixels. However, when the enlargement ratio is large, the above-described search method may not find the minimum color shift. At this time, by performing S19 to S26 in FIG. 10 (fine search detection, here, detection every other pixel), a more accurate color shift minimum image can be found. In the flowchart of FIG. 10, the minimum address of the color misregistration between the R signal and the G signal is searched, but this is similarly performed for the G signal and the B signal.

When the search by the color misregistration minimum search circuit 67 is completed, the memory address control circuit 65 changes the color misregistration minimum search circuit 6
7, the read addresses of the freeze memories 61a to 61c are controlled in accordance with the color misregistration minimum address retrieved in step 7, so that the color misregistration-corrected image is stored in the freeze memory 61a
To 61c and output to the monitor 44. By this shift correction, the color shift left in the still images with little color shift stored in the freeze memories 61a to 61c can be further reduced and displayed.

Next, the operation at the time of release will be described. When the operator presses the release switch 53, a release instruction signal (recording instruction to the photographing device 45) is input to the freeze control circuit 66. The operation of the freeze control circuit 66 is almost the same as that at the time of freeze. However, after the time required for generating the color misregistration corrected image has elapsed since the output of the correction instruction signal, the image recording instruction signal is sent to the photographing device 45. Sent. The photographing device 45 receives the image recording instruction signal, and photographs the input image signal on a film. The freeze control circuit 66 automatically restarts writing to the freeze memories 61a to 61c after outputting the image recording instruction signal, and the moving image is displayed again on the monitor 44.

In this embodiment, optical magnification is performed, but this may be applied to electronic zoom.

[Effect] At the time of enlargement, the color shift becomes larger than at the time of non-enlargement. Therefore, by setting the size of the color misregistration search area according to the enlargement ratio, appropriate color misregistration detection and correction can be performed even during enlargement.

(Supplementary Note 1) Image pickup means for picking up a subject image in a frame-sequential manner, memory means for storing each image which is image-sequentially picked up by the image pick-up means and combining them to form one image, and the image pick-up means Determining a specific area within the range of each of the images sequentially imaged according to the type of the imaging unit determined by the determining unit; Area setting means for performing color shift information detection means for detecting color shift information between the respective images stored in the memory means based on the area set by the area setting means; and A color shift correcting unit for correcting the color shift based on the detected color shift information.

(Supplementary Note 2) Imaging means for imaging a subject image in a frame-sequential manner, memory means for storing each image which is image-sequentially captured by the imaging means and combined to form one image, and the imaging means Brightness detection means for detecting a portion having a brightness level equal to or greater than a predetermined value, and a predetermined area including the portion is determined according to the portion detected by the brightness detection means. Area setting means for setting within the range of each image sequentially taken, and color shift information for detecting color shift information between the images stored in the memory means based on the area set by the area setting means A color shift correction device, comprising: a detection unit; and a color shift correction unit that corrects the color shift based on the color shift information detected by the color shift information detection unit.

(Supplementary Note 3) Image pickup means for picking up a subject image in a frame-sequential manner, memory means for storing each image which is picked up in a frame-sequential manner by the image pick-up means and combining them to form one image, and the image pick-up means The image is captured by, the magnification change means for changing the magnification of the image displayed on the display means, according to the magnification in the magnification change means, for each of the images sequentially imaged,
An area setting means for setting a specific area within the range of each image; and a color shift detecting color shift information between the images stored in the memory means based on the area set by the area setting means. A color misregistration correction device comprising: information detection means; and color misregistration correction means for correcting the color misregistration based on the color misregistration information detected by the color misregistration information detection means.

[0078]

As described above, according to the present invention, CC
Since the size of the color misregistration search area is set in accordance with the type of D or the number of pixels, the greater the number of pixels, the easier it is to detect and correct the more misregistration of the image.

Further, in an endoscope image composed of a plurality of images obtained in a time series, an area having a constant brightness is set as a color shift search area, so that the user's attention can be obtained. Detection and correction of color misregistration in an area having a certain brightness can be easily realized.

Further, since the color misregistration at the time of enlargement is larger than that at the time of non-magnification, the size of the color misregistration search area is set in accordance with the enlargement ratio. Color misregistration can be easily detected and corrected.

Further, in an endoscopic image composed of a plurality of images obtained in a time series, a color shift of the endoscopic image is detected and corrected based on a relative positional relationship between the plurality of images. Therefore, the circuit scale is small, and it is possible to detect and correct a color misregistration satisfactorily with respect to a color misregistration image generated in a conventional still image.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an apparatus for detecting and correcting color misregistration according to a first embodiment of the present invention.

FIG. 2 is a block diagram showing a configuration example of a color misregistration minimum detection circuit in FIG. 1;

FIG. 3 is a circuit diagram showing a configuration example of a mismatch circuit in FIG. 2;

FIG. 4 is a view for explaining an example of a color misregistration detection method in FIGS. 2 and 3;

FIG. 5 is a timing chart schematically illustrating the operation of the embodiment of FIG. 1;

FIG. 6 is a flowchart illustrating the operation of the color misregistration minimum detection circuit in FIG. 1;

FIG. 7 is a block diagram showing an apparatus for detecting and correcting color misregistration according to a second embodiment of the present invention.

FIG. 8 is an overall configuration diagram of an electronic endoscope apparatus to which the apparatus for detecting and correcting color misregistration according to the first and second embodiments is applied.

FIG. 9 is an overall configuration diagram of an electronic endoscope apparatus to which a device for detecting and correcting color misregistration according to a third embodiment of the present invention is applied.

FIG. 10 is a flowchart illustrating a color shift minimum address search operation of the color shift minimum search circuit in FIG. 9;

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Processor 2 ... Endoscope 3 ... Monitor 5 ... CCD 6 ... CCD driving means 8 ... Signal processing means 9 ... Light guide 11 ... Signal processing means 12 ... Scene sequential signal synchronizing means 13 ... Still image processing means 14 ... Surface Sequential light source device 15 ... Aperture control means 16 ... RGB rotation filter control means 17 ... Lamp 19 ... RGB rotation filter 20 ... Motor 21 ... CCD discrimination circuit 22 ... A / D conversion circuit 23 ... Synchronization circuit 24 ... Freeze indication circuit 25 ... Still image memory (freeze memory) 26 blur correction memory 27 color shift search area setting circuit 28 memory R / W controller 29 color shift minimum detection circuit 31 D / A conversion means 33 screen brightness detection circuit 34 Shift vector calculating means 41 ... Light source device 42 ... Scope 43 ... Processor 44 ... Monitor 45 ... Photographing device 46 ... La Step 47 ... Rotating filter plate 48 ... Light guide 49 ... CCD 51 ... Actuator 52 ... Freeze switch 53 ... Release switch 54 ... Enlargement switch 55 ... Wide angle switch 56 ... CCD drive circuit 57 ... Pre-process circuit 58 ... A / D conversion circuit 59 ... Selectors 60a-60c ... Synchronization memories 61a-61c ... Freeze memories 62a-62c ... Blur correction memories 63a-63c ... D / A conversion circuits 64 ... Actuator control circuits 65 ... Memory address control circuits 66 ... Freeze control circuits 67 ... Colors Minimum deviation search circuit

Continued on the front page (72) Inventor Kei Takasugi 2-43-2 Hatagaya, Shibuya-ku, Tokyo Inside Olympus Optical Industrial Co., Ltd. (72) Katsuichi Imaizumi 2-34-2 Hatagaya, Shibuya-ku, Tokyo Olympus Optical Industry Co., Ltd. (72) Inventor Fumiyuki Okawa 2-43-2 Hatagaya, Shibuya-ku, Tokyo Olympus Optical Industry Co., Ltd. F-term (reference) 2H040 GA02 GA06 GA12 4C061 AA00 BB00 CC06 DD00 LL02 MM03 NN07 NN05 NN07 QQ02 QQ04 SS09 SS10 SS21 TT01 TT13 TT20 WW01 WW03 YY05 5C065 AA04 BB25 BB41 CC03 CC07 CC08 DD02 DD17 GG01 GG18 GG24 GG30 GG32 GG44

Claims (3)

[Claims] 1. An image capturing means for capturing a subject image in a frame-sequential manner, a memory means for storing each image which is taken in a frame-sequential manner by the image capturing means and combining them to form one image, and a type of the image capturing means Determination means for determining the type of the imaging means determined by the determination means, for each of the images sequentially imaged,
An area setting means for setting a specific area within the range of each image; and a color shift detecting color shift information between the images stored in the memory means based on the area set by the area setting means. A color shift detecting device, comprising: information detecting means. 2. An image pickup means for picking up a subject image in a frame-sequential manner, a memory means for storing each image which is picked up in a frame-sequential manner by the image pick-up means and combining them to form one image, Brightness detection means for detecting a portion having a brightness level equal to or higher than a predetermined value in the imaging signal; and, in accordance with the portion detected by the brightness detection means, a predetermined region including the portion is image-sequentially imaged. Area setting means for setting within the range of each set image, and color shift information detecting means for detecting color shift information between the images stored in the memory means based on the area set by the area setting means A color shift detection device, comprising: 3. An image pickup means for picking up an image of a subject in a frame-sequential manner; The magnification change means for changing the magnification of the image displayed on the display means, according to the magnification in the magnification change means,
Area setting means for setting a specific area within the range of each image for each of the images sequentially imaged, based on the area set by the area setting means, stored in the memory means A color misregistration information detecting means for detecting color misregistration information between the images.