JPH03664B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for recognizing a substantially circular contour line in a television image at high speed.

Conventional methods for recognizing objects in images include:
One example is a method of tracing the contour of an object. This method focuses on the edges of each surface that makes up the object. Points with sudden changes in brightness in the image are extracted as edge points, and the edge points are connected to create a line drawing. It is used for conversion (line drawing).

For example, an example of the procedure for recognizing a circular object using the above method will be explained with reference to FIGS.
The original image (Fig. 1 a) taken with a television camera is first differentiated along the scanning line to extract one contour candidate point where the brightness suddenly changes (Fig. 1 b). Next,
A similar differentiation process is performed on each pixel in the vicinity of this point, and the pixel with the largest differential value is regarded as a point continuous to the contour candidate point mentioned above.By repeating this operation, continuous contour points (contour line candidate) (Fig. 1c), and when this contour point is closed (Fig. 1d), it is considered as one object.

However, in such conventional contour line extraction methods, there are several factors that impede tracking of contour candidate points: (1) Blaming due to metallic luster (see Figure 2 a) (2) Overlapping objects (see Figure 2 b) (3) Unclear images due to rust, dirt, etc. on the surface of objects; (4) Image disturbances due to electrical noise; as a result, there are problems in which objects that should be present may be missed. Furthermore, in order to solve these problems, recognition algorithms become complex, making real-time processing almost impossible.

The present invention has been made in view of the above-mentioned circumstances, and an object of the present invention is to provide a method for recognizing a substantially circular contour line that can recognize the presence or absence of a desired contour line (substantially circular contour line) at extremely high speed.

According to the present invention, the filter means is used to designate the first pixel group and the second pixel group, which are spaced apart from each other by a center-to-center distance equivalent to the diameter of the substantially circular contour line and are arranged in the main scanning direction, as one set. Then, by scanning the input image of a predetermined field of view in which the substantially circular contour exists, a filter feature point is searched for where two contour candidate points, each of which suddenly changes in brightness, exist simultaneously within each pixel group of the one set. Then, the filter feature points are tracked and scanned in the sub-scanning direction from the searched scanning position, and when this tracking length is greater than or equal to a preset length, a candidate point for the center position of the approximately circular contour is determined from the gravity of the tracking length,
Thereafter, the presence or absence of contour candidate points is detected while turning the filter means by a preset angle around the center position candidate point, and when the proportion of contour candidate points present is equal to or higher than a preset value, a substantially circular contour line is detected. I try to recognize that it exists.

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

FIG. 3 is a schematic configuration diagram showing an example of a device for carrying out the method for recognizing a substantially circular contour line according to the present invention, and shows a case where the ITV camera 2 is photographing a circular object 1 as a detection target object. .
The ITV camera 2 photographs the circular object 1 in a predetermined field of view and outputs a video composite signal containing brightness signals of the input image to the synchronization separation circuit 3 and the A/D converter 4. The synchronization separation circuit 3 separates a synchronization signal from the input video composite signal, specifies the address of a random access memory array (RAM array) 5 based on this synchronization signal, and the A/D converter 4 separates the synchronization signal from the input video composite signal. The bright/dark signal of the video composite signal to be
It is converted into gradation image data and written to the specified address position. In this way RAM
The array 5 stores one screen worth of image data indicating the brightness of the original image shown in FIG. In addition,
By specifying the X and Y addresses of the RAM array 5, arbitrary image data can be extracted.

On the other hand, the memory 6 stores a main program etc. for implementing the method of the present invention, and the central processing unit (CPU) 7 performs image processing of the image data in the RAM array 5 based on the main program contents. Execute.

Next, the processing procedure of this CPU 7 will be explained according to the flowcharts shown in FIGS. 6 and 7.

First, a search is performed for a candidate point for the center position of a circular object. During this search, in order to detect the outline of the diameter portion of the circular object, two pixel groups (each The pixel group consists of several pixels and has a length ΔL), and a filter is set to designate them as one set. This is called an L-filter, and the position dividing each pixel group into two equal parts is called a filter position. According to this L-filter, by specifying the filter position, the above-mentioned one set of pixel groups can be uniquely specified. Note that the interval L between the L-filters is determined by the geometrical shape of the object to be detected, and is uniquely determined by the magnification ratio of the television camera. Therefore, complicated procedures such as learning are not necessary when determining the filter structure. Further, the length ΔL is determined by the diameter of the circular object, noise, blur components, and distortion of the optical system.

Now, as shown in FIG. 4, the filter position (scanning position) is moved to the starting point PST, and it is checked whether this scanning position is a filter feature point. here,
A filter feature point refers to a case where a contour candidate point whose brightness suddenly changes at the same time exists within a set of pixels specified by the scanning position. Therefore, a filter feature point is detected when the interval between two contour candidate points whose brightness changes suddenly is approximately the distance L.

If the scanning position is not a filter feature point, the scanning position is moved to the right (X direction) by several pixels.
This movement amount is the check range ΔL of the L-filter.
Determined by

Then, it is determined whether the scanning position is at the rightmost end of the scanning range, and if it is not at the rightmost end, it is determined whether the current position is a filter feature point, and if it is not a filter feature point, the above process is performed. repeat.

In this way, when the scanning position reaches the rightmost end of the scanning range without any filter feature points, it is determined whether the scanning position is at the bottom end of the scanning range.

The scan position is at the bottom of the scan range (i.e., the fifth
In the case of position PED) in the figure, it is determined that there is no object to be detected on this screen, and image processing ends. On the other hand, if the scanning position is not at the bottom end of the scanning range, the scanning position is returned to the leftmost end of the scanning range and moved downward (in the Y direction) by a certain distance (ΔY) (see FIG. 5). Then, scanning is executed again from that scanning position to check the presence or absence of filter feature points.

Next, a case where filter feature points are detected will be described.

In this case, proceed to the flowchart shown in FIG. 7, and first, as shown in FIG. 8, the filter position (scanning position) when the filter feature point is detected is
pi, the two contour candidate points pa and pb at that time
, and store this point PS as the tracking start position PS in the sub-scanning direction. In other words, by focusing on the fact that the contour lines of a circular object that are symmetrical to the center of the circle are almost parallel in a certain section, we can investigate whether the filter feature points continue for a certain length or more in the sub-scanning direction. , First, the tracking start position PS is stored.

Next, the scanning position is moved to a point directly above the start position PS, and it is determined whether this point is a filter feature point. If this scanning position is a filter feature point, the scanning position is moved to a bisecting point of the outline candidate point, and the scanning position is stored. Next, it is determined whether the distance of the memory position from the start position PS is less than the standard or not, and if it is less than the standard, the flow returns to upward tracking again, and if it exceeds the standard, the flow shown in FIG. Then, the subroutine is exited and the scanning for searching for filter feature points shown in FIG. 6 is started again. In this way, the filter feature points are tracked until no filter feature points exist. Note that the above reference is set to, for example, 2/3 of the radius of the circle, depending on the radius of the circle to be sought. Therefore, if it exceeds this standard, it is determined that the length is too long as a parallel portion of the circle.

On the other hand, if the point directly above the scanning position is not a filter feature point, that is, when the upward tracking of the filter feature point is completed, the scanning position is moved to the tracking start position.
Return to PS and trace the filter feature points downward in the same way as above.

When the tracking of the lower filter feature point ends with the distance from the start position PS to the scanning position being less than the standard, the top point PT and bottom point PB ( (Fig. 8), and it is determined whether this tracking length exceeds a certain standard. If the tracing length is less than the standard, it is determined that the contour line is the desired contour line, exits from this flow, and moves on to scanning to search for filter feature points again as shown in Figure 6. Assuming that there is a high possibility that a contour line of a circular object exists,
The process moves on to the next flow of checking whether it is a circular contour.

In this case, first, the highest point PT and the lowest point of the above trace
Candidate point for the center position of the circular object from the bisecting point with PS
Find PO (Figure 8). Next, move the filter position to the center position candidate point PO, and also move its Lr
A rotation angle θ for rotating the filter by a predetermined angle is set. Next, the L-filter is rotated by a rotation angle θ (see FIG. 9), and it is checked whether the rotated position is a filter feature point. In this way, the presence or absence of filter feature points is sequentially checked until the L-filter rotates 180 degrees, and after the L-filter rotates 180 degrees, the proportion of filter feature points present is determined. If this ratio is less than the standard, the flow is exited and the scan moves to search for filter feature points in Figure 6 again, and if this ratio is greater than the standard, it is determined whether the rotation angle θ is sufficiently small. make that judgment.

Note that the rotation angle θ is determined depending on how detailed the image should be checked, and this depends on how much of the circular feature appears in the image. Therefore, if there are fewer circular features in the image, the rotation angle is made smaller. Further, the reference ratio depends on the nature of the contour line in the image, and if the continuity of the contour line is high, a ratio closer to "1" may be used.

If the rotation angle θ is sufficiently small, it is determined that the object to be detected is detected (FIG. 6), and this image processing ends. If the rotation angle θ is not sufficiently small, the rotation angle θ is made smaller and the filter feature points are checked again using the rotation angle θ.

In this embodiment, the presence or absence of a detection target object is determined based on the ratio of filter feature points, but the presence or absence of a detection target object may also be determined based on the ratio of the total number of contour candidate points. good. Also,
The object to be detected is not limited to a circular object, but may be any object as long as it has at least a circular or approximately circular (original or polygonal shape close to a circular) outline.

As explained above, according to the present invention, the entire circumference of the contour is not traced, but the approximately parallel portion of the center of the approximately circular contour is detected and traced by the L-filter to search for a candidate point for the center position. After that, the L-filter is rotated by a predetermined angle and the existence rate of contour candidate points is determined to recognize the approximately circular contour. Therefore, the recognition algorithm for the approximately circular contour is simplified and the processing time can be shortened. . Furthermore, since the sensitivity to noise is low, the accuracy rate for finding the target object is high.

[Brief explanation of the drawing]

Figures 1 a to d are diagrams used to explain the procedure for object recognition using the conventional contour extraction method, and Figure 2 a
3 and b are diagrams showing examples of factors inhibiting conventional contour tracing, respectively; FIG. 3 is a schematic configuration diagram showing an example of a device for carrying out the substantially circular contour recognition method according to the invention; The figure shows L- according to the present invention.
Figures used to explain the filter; Figure 5 is a diagram showing the brightness of the image data stored in the RAM array in Figure 3; Figures 6 and 7 are the processing of the central processing unit in Figure 3. Flowcharts illustrating an example of the procedure, FIGS. 8 and 9 are diagrams used to explain the flowchart of FIG. 7, respectively. 1... Circular object, 2... ITV camera, 3...
Synchronous separation circuit, 4...A/D converter, 5...
RAM array, 6... memory, 7... central processing unit (CPU).

Claims (1)

[Claims] 1. A filter means for specifying as one set a first pixel group and a second pixel group that are arranged in the main scanning direction and spaced apart by a center-to-center distance equivalent to the diameter of a substantially circular contour line. is used to scan an input image of a predetermined field of view in which the substantially circular contour line exists, and detect the presence or absence of contour candidate points whose brightness suddenly changes within each pixel group of the set of pixels, and simultaneously scan two contour candidate points. When a point reaches a scanning position that is a filter feature point, the filter feature point is tracked and scanned from that scanning position in the sub-scanning direction, and if this tracking length is equal to or greater than a preset length, the filter feature point is specified by the filter means. A candidate center position point of a substantially circular contour line is determined from the center position between the first and second pixel groups and the center position of the tracking length, and then the filter means is moved at a predetermined angle at a time about the center position candidate point. The presence or absence of filter feature points or contour candidate points is detected while rotating, and when the proportion of filter feature points or contour candidate points that exist is greater than or equal to a preset value, it is recognized that a substantially circular contour exists. How to recognize contour lines. 2. The tracking scan of the filter feature point uses, as the tracking start position, a bisecting point of the two contour candidate points when the scanning position that becomes the filter feature point is reached;
The presence or absence of a filter feature point is detected when the center position of the filter means is moved to a position directly above the start position, and if a filter feature point exists, the center position of the filter means is moved to a position directly above the start position. Detect the presence or absence of a filter feature point by moving it again to the position directly above the bisecting point of 2. A method for recognizing a substantially circular contour line according to claim 1, wherein the substantially circular contour line is tracked and scanned by.