JPH06103967B2 - Image processor for overhead line inspection - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial field of application] Since the present invention automatically performs visual inspection of overhead lines such as power lines, an abnormality of the overhead line is automatically detected from an image captured by a television camera. The present invention relates to an image processing device for inspecting overhead lines.

[Conventional technology]

Conventionally, this kind of overhead line has been inspected by a person walking or riding a helicopter and directly visually confirming it with binoculars, but recently, as shown in Japanese Patent Laid-Open No. 59-85909, a television camera is mounted on the helicopter. There is a method of shooting with a VTR and recording it on the VTR, or installing a TV camera on a self-propelled inspection machine that runs on an overhead line as shown in JP-A-61-192614 It is taken. In this case, after returning / collecting, the VTR image is played back to check if there is any abnormality on the person watching the monitor TV.

Abnormalities to be detected include strand breaks, melt marks, foreign matter adherence, and twisted wire unevenness, but these are particularly problematic.
It is a wire break as shown in (c), and if left unattended, it will gradually increase and may develop into a major accident such as a ground fault or short circuit. Therefore, it is necessary to detect and treat it as soon as possible. is there.

[Problems to be Solved by the Invention]

In such a conventional visual inspection of an overhead line, it is necessary for a person to watch the television for a long time to make a judgment, which causes problems such as fatigue and oversight.

The present invention has been made to solve the above problems, and image processing for overhead line inspection that can automatically perform anomaly detection work for overhead lines, which is conventionally performed by a person watching a monitor television. The purpose is to obtain the device.

[Means for Solving the Problems]

An overhead line inspection image processing apparatus according to the present invention processes an image of an overhead line captured by a television camera, extracts an outline of the overhead line from an edge portion of the image, and detects an abnormal change thereof. By doing so, the abnormality of the overhead line is found.

[Action]

The overhead line inspection image processing apparatus according to the present invention detects an abnormal state of the overhead line by extracting the outer contour of the overhead line from the edge portion of the image and detecting the abnormal change.

Example of Invention

An embodiment of the present invention will be described below with reference to the drawings. In FIG. 1, (1) is an overhead line to be inspected, (2) is a helicopter, (3) is an imaging system, (4) is a VTR for reproduction,
(5) is an image processing device, (6) is a monitor television, (7)
Is an indicator.

In this embodiment, an image of the overhead line (1) is displayed on the helicopter (2).
From the TV camera (31), record it with the recording VTR (32), play it back with the playback VTR (4), analyze it with the image processing device (5), and if a defective part is found, display it (7 )
Is displayed, and at that time, a person (not shown) looks at the monitor television (6) to confirm the contents.

An image processing device (5) includes a video AD converter (51) for taking in a video signal a, sampling it at a predetermined resolution and converting it into a two-dimensional digital image, a video memory (52) for storing the converted digital image, An ALU unit (53) that performs the arithmetic processing described below on the images stored in the video memory, a video DA converter (54) for displaying the arithmetic results on a monitor TV, and a VTR control signal b are generated to sequentially generate images. It consists of a control unit (55) that controls a series of operations such as image input, calculation, and display while proceeding.

The minimum constituent unit of the sampled digital image is called a pixel.

Next, the operation of the image processing apparatus (5) will be described by taking the strand break shown in FIG. 5 (a) as an example. FIG. 2 shows the processing flow.

First, the digital image sampled by the video A / D converter (51) is loaded into the video memory (52), and the ALU unit (53) performs a differential operation to remove the edge part (the part where the brightness changes rapidly) of the image. Extract. The details of the differential operation are described in the general image processing technique manual, and therefore omitted. An example of the obtained differential image is shown in FIG. Here, for the sake of simplicity, an image in which the edge portion is binarized to “1” and the others to “0” will be described.

Next, the upper start point is first searched for in the obtained differential image. Therefore, the pixel "1" is searched for from the upper left side of the screen toward the lower side. In the example of FIG. 3, point A is the upper start point.
Then, starting from point A, the upper contour of the overhead line is tracked. In this process, as shown in FIG. 4 (a), eight pixels adjacent to the current pixel are examined in a clockwise direction from directly above, and the pixel which is "1" is first set as the next point and the sequential scanning is repeated. . However, when the next point returns to the right edge of the screen or the previous point, it ends there. In the example of FIG. 3, starting from point A and ending at point A '. If it ends in the middle of the screen, it is considered to be a fake due to noise or the like, and therefore the starting point search is repeated. However, the above process cannot be performed if there is a break in the outline of the overhead line. In that case, make sure that the break does not occur by the enlargement / reduction process after the differential calculation (see the image processing technical manual for details). I need to put it. The differential processing may be omitted if the level of the inside of the overhead line of the input image is uniform and the outline of the overhead line can be extracted by performing binary image conversion at an appropriate slice level.

When tracking is completed up to the right edge of the screen, inflection point detection is performed. For this, for example, there is a method in which the azimuth detected in the tracking of the contour is stored as a sequence and the numerical change of this sequence is examined. In the example of FIG. 3, the azimuth row is as shown in FIG. 3A, and there is a portion where the azimuth is 4 or more in the portion where the strands are broken, so that an abnormal protrusion is detected. As another method, the number of "1" points above the line segment connecting AA 'may be counted, and it may be determined that there is an abnormality when the value is equal to or larger than a certain value.

Next, similarly for the lower contour, this time, the left end of the screen is searched for from the bottom to the top, and eight pixels which are adjacent to each other are sequentially examined from the bottom to the counterclockwise direction as shown in FIG. 4 (b). And tracking. In the example of FIG. 3, starting from point B,
It ends at point B '. In this case, the azimuth example is as shown in FIG. 7B, and there is no particular abnormality.

The above processing is sequentially performed for each screen by the VTR, and if an abnormality is detected, it is displayed on the display (7).

In the above embodiment, the image taken from the helicopter was targeted, but it may be taken from the ground with a telescope or the like, or may be taken with a self-propelled inspection machine. Further, in the above embodiment, the image recorded in the VTR is reproduced and judged, but the image processing device is mounted on a helicopter or a self-propelled inspection machine, or the photographing system and the image processing device are connected by a communication line. Alternatively, the determination may be performed online.

Further, in the above embodiment, when an abnormality is found, the display is immediately notified, but if the abnormal portion is stored in memory and can be confirmed collectively, the work can be performed more efficiently.

Although the above embodiment has described the case where an image other than the overhead line is not included, flying objects (small birds, etc.) may be mixed in, and buildings and trees may be detected in the background. May be erroneously determined to be abnormal.

Therefore, the above erroneous determination can be prevented by performing determination processing for a plurality of consecutive frames of an image and determining an abnormality only when the position of the foreign matter moves in accordance with the moving speed of the screen. .

〔The invention's effect〕

As described above, according to the present invention, the overhead line inspection image processing apparatus is configured to automatically detect an abnormal change by processing the image of the overhead line and extracting the outer contour from the edge portion of the image. Therefore, it is not necessary for a person to watch the television for a long time to make a decision, and there is an effect that a problem such as fatigue and oversight can be avoided.

[Brief description of drawings]

FIG. 1 is a system configuration diagram using an image processing device for overhead line inspection according to an embodiment of the present invention, FIG. 2 is a processing flow chart of the image processing device, and FIG. 3 is an explanatory diagram showing an image processing example. 4 (a) and (b) are outline tracking principle diagrams,
FIGS. 5 (a) to 5 (c) are partial views showing examples of anomalous overhead lines. In the figure, (1) is an overhead line, (2) is a helicopter,
(3) is a photographing system, (4) is a VTR for reproduction, (5) is an image processing device, (6) is a monitor television, and (7) is a display.

Claims (2)

[Claims] 1. An overhead line inspection image processing apparatus for inspecting the overhead line by arithmetically processing an image obtained by shooting the overhead line with a television camera, wherein an outline of the overhead line is provided by an edge portion of the image. An image processing apparatus for inspecting an overhead line, wherein the abnormality in the overhead line is detected by extracting the abnormal state and extracting the abnormal change. 2. The overhead line inspection image according to claim 1, wherein an erroneous determination is prevented by determining an abnormality when the same type of abnormality is detected in a plurality of consecutive frames of the image. Processing equipment.