JPS60179882A - Recognizing method of substance - Google Patents

Abstract

PURPOSE:To shorten a processing time by extracting picture groups on several positions specifying an outline through an L filter without tracking the whole periphery of the outline and recognizing the outline. CONSTITUTION:The filter position of the L filter is moved to a starting point PST and whether the scanning position is the feature point of the filter or not is checked. When the scanning position is the lower-most end (PED) of the scanning range, ''no substance to be detected'' on the picture is decided and the picture processing is ended. When the scanning position is located on a position other than the lowermost end, the scanning position is returned to the leftmost end of the scanning range and also moved by a certain distance (DELTAY) in the lower direction (Y direction). Scanning is executed again from the scanning position, the presence or absence of a filter feature point is checked, and when the feature point is detected, ''detection of a substance to be detected'' is decided at the scanning position.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an object recognition method for recognizing an object from a contour line in a television image.

The conventional object recognition method used by this woodcutter focuses on the line points (edges) on each surface that make up the object. Points where the brightness changes suddenly in the image are extracted as line points, and the line points are By connecting K, it is converted into a line drawing (line drawing).

For example, to explain an example of the procedure for recognizing a circular object using the above method with reference to FIGS. Differential processing is performed along the eclipse line to extract one contour candidate point where the brightness suddenly changes (FIG. 1(b)). Next, 1% pixels in the vicinity of this point are subjected to similar differentiation processing, and the pixel with the largest differential value is identified as a point continuous to the contour candidate point mentioned above.By repeating this operation, continuous contours are created. Obtain the point (1lli! Guo 4I + l candidate) (Fig. 1 (Cυ
, Furthermore, when this contour point closes, it is considered as one object. However, in such conventional contour line extraction methods, the following factors hinder the tracking of contour candidate points: (1) Blue light due to metallic luster (see Figure-A2 (a)) (2) Object shape (see Figure 2 (1))) (31'4
B: Unclear image due to rust, dirt, etc. on the surface of the body (4)' Image disturbance due to atmospheric noise, etc., and as a result, there is a problem that 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 an object recognition method that can recognize the outline of a desired object at extremely high speed.

According to this invention, the outline of the object to be detected is known, and the positional relationships of several outline parts that can identify the outline are also known, and each pixel group is aligned with the outline of the object to be detected. The input image of a predetermined field of view in which the object to be detected is present is scanned using a filter means for specifying as one set at least three or more pixel groups in the position corresponding to The presence or absence of contour candidate points whose brightness suddenly changes is detected within each pixel group, and when contour candidate points are detected at the same time in each pixel group, it is recognized that the contour of the object to be detected exists. There is.

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 an apparatus for implementing the object recognition method according to the present invention, in which the object to be detected 1 is
The case where the camera 2 is photographing is shown. Engineering T
The V camera 2 photographs the object 1 in a predetermined field of view and outputs a video composite signal containing brightness signals of the input image to the analog separation circuit 3 and the A/D converter 4. The synchronization separation circuit 3 separates the synchronization signal from the input video composite signal, and generates a random signal based on this synchronization signal.
The address of the access memory array (RAM array) 5 is determined, and the A/Di converter 4 converts the brightness signal of the input video composite signal into true image data with a brightness state of 16 levels, and converts this into the above-mentioned image data. Write to the specified address position.In this way, one screen worth of image data indicating the brightness of the original image shown in FIG. 4 is stored in the RAM array 5.The X and Y addresses of the RAM array 5 In specifying.

More arbitrary image data can be extracted.

- The main program, etc. for carrying out the method of the present invention is stored in the memory 6, and the central processing unit (CPU)
7 executes image processing of the image data in the RAM array 5 based on the main program contents.Next, the image processing procedure of this a pr, r7 will be explained.
il, the principle of the object recognition method according to Reimei Moto will be explained.

The method of the present invention examines the swallowtails of a desired outline depending on whether or not there are features of the ring 4 within a set range of three or more at the same time. That is, as shown in FIG. 5, a range A□ of ΔL1 in which the center position of the reference position P is separated by a distance L1, a range A2 of ΔL2 in which the center position is separated by a distance L2 from the reference position P, . . . ...The center position is separated from the reference position P by a distance Ln [d1 distance Δ1. Set a Buko filter corresponding to the ring and flS line of the object to be detected in the burn, specifying the range An of n as one set,
The input screen is scanned using the above reference position P as the travel position, and points where there are contour melting complement points whose brightness changes at the same time within each range A1゜A2, .=A n (fill feature points) For example, for the detection target object 1 in the non-conventional example, as shown in Fig. 6! 1-3
σ) Set a filter (hereinafter referred to as L-filter) that adds a certain range AI' t A21 A3, respectively,
σ) While moving the filter position of the L-filter (the position indicated by the hot circle), find the filling symptom point σ within the range ``l I A2 P A3 (・ and at the same time the ring v and the complement point σ ] Check for presence 0 In addition, L-Fill interval, ,L
,.

...Ln is determined by the geometrical shape of the object to be detected9, and is uniquely determined by the magnification of the television camera. Therefore, when determining the filter structure, complicated procedures such as learning are not necessary. is unnecessary. Buta, ΔL1゜ΔL2...
...ΔLn is determined by noise, blur components, distortion of the optical system, etc.

Next, the processing procedure of 0PU7 to execute the above contents tt is explained in the seventh section.
A detailed explanation will be given according to the flowchart shown in the figure. First, the filter position (scanning position) of the L-fill described above is moved to the starting point P8T as shown in FIG. 4, and it is checked whether this scanning position is a filter feature point.

If the scanning position is not a filter feature point, the scanning position is moved in the right direction (X direction) by K number of pixels.

This moving chamber is within the L-filter check range ΔL.

ΔL2. It is determined by ΔL.

Then, determine whether the scanning position is at the rightmost end of the scanning range,
If it is not 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 repeated.

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.

If the scanning position is at the lowest end of the scanning range (ie, position pED in FIG. 4), 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 a certain distance (ΔY
) in the downward direction (Y direction) (see Figure 4). Scanning is then performed again from that scanning position to check for the presence or absence of filter feature points.When the screen is scanned with the L-filter as described above and a filter feature point is detected, the "detection target" is detected at that scanning position. "Object 1 detected" 0
In addition, in this L-filter, as shown in FIG. 6, the filter feature point appears when cutting off the section l at the approximate center of the detection parabolic object 1, so the exact position of the detection target object 1 can be specified. I can't.

Therefore, in this case, the L-filter should be moved in the vertical direction (
By moving in the sub-scanning direction) and examining the range in which the filter feature points exist, the accurate position of the detection target object 1 can be specified. Furthermore, if the detection target object 1 is determined to exist only when the tracking length of the filter feature point in the sub-scanning direction is equal to or greater than a certain reference length,
Object recognition with a higher accuracy rate is possible.

Also, the L-filter has a certain range A, ,A,...An
The pixels are not limited to the above embodiment in which the pixels are lined up on the same scanning line, but as shown in FIGS. It can be set as follows.

In addition, since the L-filter has directionality, for example, if the screen is scanned with an L-filter set according to the object to be detected, and it is determined that there is no object to be detected, then the L-filter The object is rotated by a predetermined angle and scanned again by the L-filter, and this is repeated until it is determined that the object to be detected has been detected. It is also possible to infer in which direction the object to be detected exists.0 As described above, according to the present invention, the entire contour is not tracked, but only a few points where the contour can be identified by the L-filter are tracked. The contour line of the object to be detected is recognized when the pixel groups are extracted and contour candidate points are simultaneously detected within each pixel group, which simplifies the contour line recognition algorithm and reduces processing time. can do.

In addition, since the sensitivity to noise is low, the probability of finding the target object is high.

[Brief explanation of the drawing]

Figures 1 (a) to (d) are diagrams used to explain the basics of object recognition using the conventional contour extraction method, and Figure 2 (a)
1 and (b) are diagrams each showing an example of a factor that inhibits conventional contour tracing, FIG. 3 is a schematic configuration diagram showing an example of a device for carrying out the object recognition method according to the present invention, and FIG. is a diagram showing the brightness of image data stored in the RAM array in FIG. 3, FIGS. 5 and 6 are diagrams used to explain the filter according to the present invention, and FIG. A flowchart showing an example of the processing procedure of the central processing unit, and FIGS. 8 to 11 are diagrams showing other embodiments of the L-filter, respectively. 1... Detection target object, 2... I'TV camera,
3... Synchronization separation circuit, 4... A/D converter, 5
...RAM array, 6...Memory, 7...Central processing unit ('OP U). Figure 1 ((1) (b) (() (d) Figure 2 (a) (b) Figure 3 Figure 4 Figure 5 Measures 6 Figure 7

Claims (1)

[Claims] An input image of a predetermined field of view in which the detection target object exists, using a filter means for specifying as one set at least three or more pixel groups in which each pixel group has a positional relationship corresponding to the outline of the detection target object. At the same time, it detects the presence or absence of outline candidate points whose brightness changes suddenly within each pixel group, and when the outline candidate points are simultaneously detected within each pixel group of the set, the outline of the object to be detected is detected. An object recognition method that recognizes the existence of lines.