JPH05209730A - Image processing apparatus - Google Patents

Abstract

PURPOSE:To measure the three-dimensional position and the like of a body by using one image pickup device, and picking up the similar image so that the viewing difference due to monocular vision is eliminated. CONSTITUTION:A zoom lens 2, which expands and reduces the image, is attached to a TV camera 1 for picking up an image. The image having the parallax for the same object are obtained by changing the focal distance of the zoom lens and picking up the image. A body 8 is measured based on this image and each focal distance at the time when the image is picked up.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing device for measuring the three-dimensional position and distance of an object (object to be measured) using an image pickup device (for example, TV camera).

[0002]

2. Description of the Related Art In order to measure a three-dimensional position or distance of an object using a TV camera, two or more images having a common visual field and parallax are required.

For this reason, conventionally, stereoscopic vision using two TV cameras has been generally performed for measuring the three-dimensional position and distance of an object. That is, two TV cameras, each having a lens with the same focal length, are installed at equal heights so that the optical axes are parallel to each other and the depression angle is 0 degree, and the same object (object) is placed in two directions. The image of the object is captured, and the parallax obtained from the two images obtained at this time causes the 3
It was designed to measure dimensional position and distance.

FIG. 4 shows an optical system of this conventional example. Here, a is the distance between the two TV cameras, f is the focal length of each TV camera (common), L is the distance from the imaging surface to the object m, and the optical axes of the TV cameras are parallel to each other. Further, they are installed so that the depression angle is 0 degree and they are at the same height, and the intersections of this optical axis and the perpendicular drawn from the article m toward the optical axis are designated as Q ₁ and Q ₂ .

Then, the position when the object m is imaged on the image pickup surface of _one TV camera is m ₁ , the distance from the center of the lens of this TV camera to this position m ₁ is s ₁ , and the other T
The position when the object m is imaged on the imaging surface of the V camera is m ₂ , and this position m ₂ is from the center of the lens of this TV camera.
If the distance to is s ₂ , the parallax S between them is S = s _{1 to} s ₂ = (line segment Q ₁ m to line segment Q ₂ m) f / (L−f) = a · f / (L -F), and the distance L can be obtained from this formula (for example, Computer Vision 42-4, May 1986 5).
Please refer to the issue on the 22nd of March.

[0006]

However, in the case of the above-mentioned conventional example, two TV cameras are necessary for measuring an object, which is structurally designed so that two TV cameras can be supported. It requires a fairly complex and robust head. Also, because of the difference in the observation points generated by two TV cameras, the density value of the same object at the same point changes when it is made into an image, and the appearance of the object also differs, so that it is not a similar shape There is a problem in that the feature points to be detected when performing is limited.

In view of the above, the present invention uses a single fixed image pickup device (for example, a TV camera), that is, by taking an image of a similar shape in which the difference in appearance is eliminated by a single eye, the three-dimensional position of an object, etc. The purpose is to provide a device that enables measurement of.

[0008]

In order to achieve the above object, an image processing apparatus according to the present invention has a zoom lens for enlarging and reducing an image attached to an image pickup apparatus for picking up an image, and a focal length of the zoom lens. It is characterized in that the object is measured from an image obtained by changing the magnification and the magnification ratio of the same object, and the focal length at the time of this imaging.

[0009]

According to the present invention configured as described above, an image obtained by picking up an image at a certain focal point and a same object obtained by picking up an image at another focal point at the same point. The image with respect to is a similar shape in which the appearance of the object is the same but the size is different, and the unknown number in position measurement can be obtained from both images of this similar shape and the focal lengths at the time of imaging.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to FIGS.

FIG. 1 is a block diagram of this embodiment. A TV camera 1 for picking up an image is provided with a zoom lens 2 for enlarging and reducing the image. The zoom lens 2 is configured such that its focal length can be moved by the focal length moving unit 3, and also has a focal length detecting unit 4 for detecting the focal length.

On the other hand, the TV camera 1 includes an A / D converter 5
Is connected to the A / D converter 5, which is connected to a frame memory 6 capable of storing two images. As a result, the video signal output from the TV camera 1 is converted into a digital signal by the A / D converter 5 and stored in the frame memory 6.

A central control processing unit 7 for reading the data written in the frame memory 6 is provided, and a command is given to the focal length moving unit 3 by the central control processing unit 7 and a zoom lens is also provided. The focal length of 2 can be read by the central control processing unit 7.

Next, an example of use of the above embodiment will be described.

First, the object 8 to be measured is set on the TV camera 1
At this time, the central control processing unit 7 reads the focal length of the zoom lens 2 through the focal length detection unit 4 before the image is captured. Then, the image signal output from the TV camera 1 is picked up by the A / D
It is converted into a digital signal by the converter 5 and stored in the frame memory 6.

Next, the central control processing unit 7 gives a command to the base distance moving unit 3 to change the focal length of the zoom lens 2 and, in the same manner as described above, reads the changed focal length and the TV camera. The image signal from 1 is A /
The data is stored in the frame memory 6 through the D converter 5.

FIG. 2 shows an optical system when the object 8 is imaged by changing the focal length at this time. Here, O represents the central axis of the optical system, h represents the size (height) of the object 8, and L represents the distance from the imaging surface to the object 8.

When the object 8 is first imaged, the size of the object 8 stored in the frame memory 6 is h _s , the focal length obtained by the focal length detector 4 at this time is f _s , and the focal length is And the size of the object 8 stored in the frame memory 6 when the object 8 is imaged for the second time is changed to h _e ,
The focal length obtained by the focal length detector 4 at this time is f
When _e, these _{relationships, h s / h = f s} / (L-f s) h e / h = f e / (L-f s) and it can reveal, the distance L and the object from the equation The size (height) h of 8 can be obtained.

In order to measure the position of the object 8, it is sufficient to detect the characteristic points of the object 8 and determine that they are the same point. From the images of the two common portions, the edges and corner points that are the characteristic points of the object 8 are detected. This can be done by detecting

For example, FIG. 3 shows an image of the object 8 stored in the frame memory 6 before and after changing the focal length, which is subjected to high-pass filter processing by the central processing unit 7 to form a line image. When the corner point P ₁₁ before the focal length change, which is the corner point of the object 8, is used as the feature point, and the corresponding corner point P ₂₁ is detected from the image after the focal length change,
Small area including points <P ₁₂ , P ₁₁ , P ₁₄ of image before focal length change and points <P ₂₂ , P ₂₁ , P ₂₄ of image after focal length change
The point having the highest correlation can be detected as the same point on the object 8 by obtaining the correlation with the small area including the.

As the image pickup device, any device such as a video camera other than a TV camera can be used as long as a zoom lens can be attached and used.

[0022]

Since the present invention has the above-mentioned structure,
It is possible to measure the distance and the three-dimensional position of an object with one image pickup device, that is, with a single eye, without requiring a plurality of image pickup devices. It eliminates the need to be relatively robust with a relatively simple structure for the platform and eliminates the difference in the appearance of the object caused by the difference in the observation points of multiple image capturing devices, and There is an effect that the point can be detected without performing the correction process.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention.

FIG. 2 is a diagram similarly showing an optical system for explaining the operation.

FIG. 3 is a diagram showing an example of images of an object before and after changing a focal length.

FIG. 4 is a diagram showing an optical system for explaining an operation of a conventional example.

[Explanation of symbols]

 1 TV camera (image pickup device) 2 Zoom lens 3 Focal length moving unit 4 Focal length detection unit 5 A / D converter 6 Frame memory 7 Central control processing unit 8 Object (measurement object)

Claims (1)

[Claims] 1. A zoom lens for enlarging and reducing an image is attached to an image pick-up device for picking up an image, and a zoom ratio or a reduction ratio for the same object obtained by changing the focal length of the zoom lens and taking an image is changed. An image processing apparatus characterized in that an object is measured from the image thus obtained and each focal length at the time of image pickup.