JPH09247376A - Picture inclination detecting method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To precisely detect the inclination of a picture by respectively obtaining the inclination of more than three edges of a picture frame and detecting the inclination of the picture through the use of more than two edges of similar inclination among these. SOLUTION: Based on picture information stored in a picture memory part, CPU retrieves a point where data constituting the picture information changes from black to white or white to black to recognize the picture frame. Next, CPU obtains a vertex from the picture frame. For example, points at four corners A(x0, y0), B(x1, y1), C(x2, y2) and D(x3, y3) are obtained. A CPU respectively obtains inclination α, β, γ and δ to the x-direction of sides (a) to (d) from each coordinate of the obtained vertexes A to D. For example, the inclination γ of the side (c) is obtained from γ=(y2-y1)/(x2-x1).

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image inclination detecting method used in an image processing apparatus having a function of automatically correcting an inclination of an image projected on a microfilm to obtain an erect image.

[0002]

2. Description of the Related Art In a microfilm projection device such as a microfilm reader, a microfilm reader printer or a microfilm scanner, which projects image information recorded on a microfilm onto a screen, a photoconductor, an image sensor, etc. Occasionally, the original image is taken while it is tilted, or the film is tilted due to transport failure during film transport by a roll film carrier, etc., and the image projected on the screen or image sensor may be tilted. .

In order to correct the inclination of such an image, in a recent microfilm projection apparatus, the inclination of the image is automatically corrected by rotating a prism arranged in the optical path, and the projected image is erected. There are things I tried to do. In order to correct the tilt of the projected image, first, the projected image is read by an image sensor or the like, the tilt of the image is calculated based on the image information, and the prism is rotationally driven by a predetermined angle according to the tilt. .

As a method for detecting the inclination of an image, there is a method of detecting the inclination of an image from the inclination of any one side of an image frame, as disclosed in JP-A-7-20570 and JP-A-7-72549. Proposed. Japanese Patent Laid-Open No. 7-2057
In the detection method disclosed in No. 0, the tilt of the image is obtained from the tilt of the lower side of the projected image, and in the detection method disclosed in Japanese Patent Laid-Open No. 7-72549, a certain range at the left and right ends of the projected image is excluded and left. The inclination of the image is obtained by using one side of the broken portion.

[0005]

As in the conventional image inclination detecting method, a method of first recognizing one side of an image, obtaining the inclination of the side, and regarding the inclination of this side as the inclination of the projected image is not provided. , If a part of the image, especially the recognized side is missing or dirty, the tilt of the recognized side cannot be accurately detected, and a tilt different from the actual tilt of the projected image may be detected. is there. As described above, since the recognized inclination of the side does not always match the inclination of the projected image, there is a problem that the accuracy of image inclination detection decreases.

Therefore, the present invention is to provide an image inclination detecting method capable of accurately detecting the inclination of an image even if a part of the image has some cracks or stains.

[0007]

According to the present invention for achieving the above object, in an image inclination detecting method for obtaining the inclination of an image frame of a rectangular projection image, the inclinations of three or more sides of the image frame are obtained respectively. The image inclination detecting method is characterized in that the inclination of the image is detected by using two or more sides having similar inclinations.

When there is a side whose inclination cannot be recognized correctly due to chipping or dirt, the inclination of this side is significantly different from the inclinations of the other sides. Therefore, the method according to the present invention excludes this side. , The inclination of an image is detected using the inclinations of a plurality of other sides whose inclinations are close to each other. As a result, the inclination of the image can be accurately detected even if the original is not a perfect rectangle such as a part of the image having a chip or a stain.

[0009]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below. FIG. 1 is a schematic configuration diagram of an optical path system of a microfilm projector.

The illustrated microfilm projection device projects the image information projected on the microfilm onto a CCD line sensor, and projects the image information as follows. The light emitted from the light source 1 is condensed by the condenser lens 2 and then folded by the mirror 3 to irradiate the microfilm F loaded in a carrier (not shown) from below to pass through to the projection lens 4. . After that, the light that has passed through the tilt correction prism 5 is bent by the scan mirror 6 and further bent by the plurality of folding mirrors 7, 8 and 9 to form a CCD.
It is guided to the line sensor 10 and forms an image on the sensor 10. As a result, the image information of the microfilm F becomes CC
It is projected and exposed on the D line sensor 10 to be read. The image signal photoelectrically converted by the CCD line sensor 10 is input to a laser printer (not shown) or the like, and copying or the like is performed based on this image signal.

The tilt correction prism 5 is connected to a prism drive motor 14 via a pair of timing pulleys 11, 12 and a timing belt 13. The prism drive motor 14 is composed of, for example, a pulse motor or the like so that the rotation angle can be detected. The prism 5 is rotationally driven in a predetermined direction by a predetermined angle by transmitting the rotational power of the prism drive motor 14. As a result, the projected image on the CCD line sensor 10 is rotated,
The tilt of the image can be corrected.

The image is scanned by the CCD line sensor 10 in the main scanning direction. Further, the scan mirror 6 is rotationally driven by the mirror drive motor 15, and the image is scanned in the sub scanning direction. The sub-scan may be performed by driving the CCD line sensor 10 or the microfilm F, in addition to the mirror scanning for driving the scan mirror 6.

FIG. 2 is a schematic block diagram of a control system of the microfilm projector.

The A / D converter 30 converts the image information of the microfilm F captured as an analog electric signal in the CCD line sensor 10 by the mirror scanning as described above into a digital image signal based on the conversion signal from the CPU 31. Convert. The converted digital image signal is input to the processing circuit 32, subjected to predetermined image processing such as gamma correction based on the signal output from the CPU 31, and then 1
It is sampled at a mm pitch and stored in the image memory unit 33. The CPU 31 calculates the sampling data read from the image memory unit 33 based on the program stored in the EPROM 34, obtains the tilt of the image, and corrects the tilt. The CPU 31 outputs a prism rotation signal for rotating the tilt correction prism 5, while the CPU 31 receives a fixed position signal of the prism 5. The EPROM 34 stores a program or the like for detecting an image inclination described later, and the RAM 35 temporarily stores calculation results and the like when the CPU 31 performs various processes. is there.

Next, the operation contents of the image inclination detection will be described with reference to the flowchart shown in FIG. The operation of this flowchart will be described with reference to FIG.

The image information printed on the microfilm F is whiter than the surroundings in the case of a positive film and blacker than the surroundings in the case of a negative film. Image inclination detection processing can be performed on any of the films.

First, the CPU 31 searches the sampling data (image information) stored in the image memory unit 33 for a point at which the data forming the image information changes from black to white or from white to black. The image frame is recognized (step S1). Next, the CPU obtains a vertex from the image frame. For example, in FIG. 4, points A (x0, y0), B (x1, y1), C (x2, y) at the four corners from the image frame are displayed.
2), D (x3, y3) is calculated (S1).

The vertices of the image frame are obtained as follows. A line in which the data forming the image information changes from black to white to black or from white to black to white is searched in the X direction and the Y direction shown in FIG. 4, and a point having the smallest change distance, that is, a minimum point is obtained. . This is because the above black-and-white or black-and-white change distance is minimized at the apex. It should be noted that the method for obtaining the vertices is not limited to this, and can be modified as appropriate.

Next, the CPU 31 calculates a side a, a side b, a side c, and a side from the obtained coordinates of the vertices A, B, C, and D.
Slopes α, β, γ, δ of the side d from the x direction shown in FIG.
Are calculated respectively (S2). For example, the slope γ of side c is
γ = (y2-y1) / (x2-x1)

Here, as in the prior art, one with an image
In the method of obtaining the inclination of the side and using this as the inclination of the image, when a part of the image is missing or dirty, the image frame may be recognized as a trapezoidal shape as shown in FIG. The inclination of the detected image greatly differs depending on which side is focused, and as a result, there is a problem that the detection accuracy of the image inclination is significantly reduced.

Therefore, in this embodiment, the inclination of the image is obtained as follows. Since the inclinations of the sides b and d are 90 ° rotated with respect to the sides a and c, the inclinations of the sides b and d are (90 ° -β), (90
° -δ). The angle α obtained in this way,
(90 ° -β), γ, and (90 ° -δ) should all have the same value if the image frame is a perfect rectangle. However, since the image frame shown in FIG. 4 is recognized as a trapezoid, the inclinations α, (90 ° -β), and (90 ° -δ) of three sides a, b, and d that intersect at a right angle are Although the values are almost the same, the slope γ of the side c is a value greatly different from those of the other three sides. Then, in this case, the inclination θ of the image is determined based on the inclinations of the three sides a, b and d excluding the side c.

Explaining with reference to the flow chart shown in FIG. 3, in steps S3 to S6, four angles α, (90
Approximate states of (° -β), γ, (90 ° -δ) are sequentially determined, and four angles α, (90 ° -β), γ, (90 °-
If the value also substantially coincides with δ), step S3
Is determined to be “Y”, and the inclination θ of the image is determined to be θ = α (S7). If the absolute value of the difference between the two angles is within about 1 degree, it is assumed that the two angles are close to each other.

When the angles α, (90 ° -β) and γ are substantially the same and the angle (90 ° -δ) of the side d is significantly different from these angle values, the step It is determined to be “Y” in S4, and the image inclination θ is determined as θ = α based on the inclinations of the three sides a, b, and c excluding the side d (S7).

Further, as shown in FIG. 4, the angles α,
If (90 ° -β) and (90 ° -δ) substantially match, and the angle γ of the side c is a value greatly different from these angle values, it is determined as "Y" in step S5. , The image inclination θ based on the inclinations of the three sides a, b, and d excluding the side c
= Α is determined (S7).

Further, the angles α, (90 ° -δ) and γ are almost the same, and the angle (90 ° -β) is obtained with respect to these angle values.
If the values are significantly different, “Y” is determined in step S6.
It is determined that the inclination θ of the image is three sides a excluding the side b,
Based on the inclinations of c and d, θ = α is determined (S7).

Further, the angles (90 ° -β), γ, and (90 ° -δ) substantially coincide with each other, and the angle α corresponds to these angle values.
If the values are significantly different, “N” is determined in step S6.
It is determined that the inclination θ of the image is three sides b excluding the side a,
Based on the slopes of c and d, θ = β is determined (S8).

Then, the CPU 31 drives the prism drive motor 14 to rotate the prism 5 so that the inclination θ of the image thus obtained becomes zero, and the projected image becomes an erect image.

As described above, in the present embodiment, the side (side c in FIG. 4) whose inclination cannot be correctly recognized due to chipping or stains has the inclination (γ) of the other side. It is largely different with respect to the inclination, so it is excluded when detecting the inclination of the image, and the inclination of the image is detected using the inclinations of the other plural sides (a, b, d) whose inclinations are close to each other. To be done. Therefore, the inclination of the image can be accurately detected even if the original is not a perfect rectangle such as a part of the image having a chip or a stain.

In the flow chart of the present embodiment shown in FIG. 3, for example, when it is determined to be "Y" in step S5, the value of the inclination θ of the image determined in step S7 is (90 ° -β) or (90
Deg .- [delta]) or an average value of [alpha], (90 [deg.]-[Beta]) and (90 [deg.]-[Delta]). Further, although the inclinations of the four sides are obtained, the inclinations of the image may be detected by obtaining the inclinations of at least three sides. Further, the case where the value of the inclination of the image is determined based on the inclinations of the approximated three sides has been described, but the inclination of the image may be detected based on the approximated inclinations of at least two sides.

[0030]

According to the image inclination detecting method of the present invention, the inclinations of three or more sides of the image frame are obtained, and the inclination of the image is detected by using two or more sides of which the inclinations are close. Therefore, even if a part of the image is chipped or soiled, when detecting the tilt of the image, the side for which the tilt is not correctly recognized is removed, thereby accurately detecting the tilt of the image. It becomes possible.

[Brief description of drawings]

FIG. 1 is a schematic configuration diagram of an optical path system of a microfilm projector.

FIG. 2 is a schematic block diagram of a control system of the microfilm projector.

FIG. 3 is a flowchart showing an operation content of image inclination detection.

4 is a diagram for explaining a calculation process of the flowchart shown in FIG.

[Explanation of symbols]

 5 ... Prism 10 ... CCD line sensor 31 ... CPU F ... Microfilm a, b, c, d ... Side α, β, γ, δ ... Side inclination

Claims (1)

[Claims] 1. An image inclination detecting method for obtaining an inclination of an image frame of a rectangular projection image, wherein inclinations of three or more sides of the image frame are respectively obtained, and two or more sides of which the inclinations are close to each other are used. An image tilt detecting method, which comprises detecting the tilt of an image.