WO2020166147A1

WO2020166147A1 - Leakage oil detection apparatus and leakage oil detection method

Info

Publication number: WO2020166147A1
Application number: PCT/JP2019/043721
Authority: WO
Inventors: 莉呂; 六戸　敏昭; 明山岸
Original assignee: 株式会社日立製作所
Priority date: 2019-02-14
Filing date: 2019-11-07
Publication date: 2020-08-20
Also published as: JP2020134189A

Abstract

The purpose of the present invention is to provide a leakage oil detection apparatus and a leakage oil detection method for detecting leakage oil with high sensitivity, even in cases when there is a significant amount of background noise from afternoon sunlight. The present invention is a leakage oil detection apparatus for detecting leakage oil from oil-filled equipment, said leakage oil detection apparatus being characterized by comprising a UV light source that irradiates the oil-filled equipment with UV light, an imaging device that captures images of the oil-filled equipment, a control device that controls the operation of the UV light source and the imaging device, a storage device that stores the images captured by the imaging device, an image processing device that processes the images stored in the storage device, and a display device that displays the processing results from the image processing device, wherein the image processing device: acquires the intensity value differentials for the intensity values of the blue components between a first image obtained by capturing an image of the oil-filled equipment during irradiation by the UV light source and a second image captured of the oil-filled equipment under imaging conditions different from the imaging conditions for the first image to thereby acquire an image expressing the intensity value differentials; adjusts the contrast of the image expressing the intensity value differentials; compares the values against a pre-set threshold value, and identifies sites having a value greater than the threshold value as leakage oil adhesion sites.

Description

Oil leak detection device and oil leak detection method

The present invention relates to an oil leakage detection device and an oil leakage detection method suitable for being applied to oil-filled equipment such as oil storage tanks and transformers.

▽ Conventionally, there was a concern that oil leakage (oil leakage) may occur due to deterioration or accidents in oil-filled equipment such as oil storage tanks and transformers. Since oil leakage may lead to environmental pollution and disasters, a technique for easily and highly accurately detecting oil leakage at the initial stage has been required. In general, there is a technology to detect oil leaks by detecting the fluorescence emitted by irradiating oil with ultraviolet light, but the accuracy of oil leaks decreases when background noise from sunlight during the day is large. To do.

There is a technique described in Patent Document 1 as a technique for solving this problem. This document includes a trigger signal for determining the shutter time width and timing of the visible light camera, and the irradiation timing visible light camera synchronized with the trigger signal captures an image at 30 frames/sec. Then, an image of the excitation light source irradiation in which the control unit generates the light emission operation signal with the gate width of 1.7 ms to 24 ms in synchronization with the trigger signal of either the even-numbered frame or the odd-numbered frame and the image obtained when the irradiation is not performed. There is described a technique for discriminating oil leakage based on a difference in luminance between both images.

Patent No. 5351081

However, with the technique described in Patent Document 1, it is difficult to detect a thin oil film attached to the surface of an object unless a high-intensity excitation light source is used in an environment where sunlight is irradiated. This can be seen from what is shown in FIG. 1 obtained from the experimental results of the inventor. The conditions of this experiment are outdoor: ambient illuminance of 3850 lux, ultraviolet light intensity of 1920 μW/cm 2, and oil film thickness of about 1 mm.

As an experimental method, a commercially available ultraviolet LED light is used to irradiate the surface of the object on which the oil is attached. UV light LED light, fix the camera and the object.

First, when shooting with a camera when the LED light was not illuminated, the image on the left side of the upper part of Fig. 1 was obtained. Next, when the LED light was irradiated and the image was similarly taken by the camera, the image on the right side in the upper part of FIG. 1 was obtained. It was difficult to discriminate oil leakage in the image obtained after calculating and subtracting the luminance for each pixel from these two images, as shown in the lower part of FIG.

An object of the present invention is to provide an oil leakage detection device and an oil leakage detection method that detect oil leakage with high sensitivity even when background noise due to sunlight in the daytime is large.

From the above, in the present invention, "an oil leak detection device for detecting an oil leak of an oil-filled device, which emits ultraviolet light to the oil-filled device, an imaging device for photographing the oil-filled device, and an ultraviolet light source. A control device that controls the operation of the light source and the image pickup device, a storage device that saves an image captured by the image pickup device, an image processing device that processes the image stored in the storage device, and a processing result of the image processing device And an image processing device configured to display a first image obtained when an oil-filled device is photographed by irradiating an ultraviolet light source, and a photographing condition different from the photographing condition of the first image. An image representing the intensity value difference is acquired by acquiring the intensity value difference between the second image obtained by photographing the oil-filled device and the intensity value of the blue component, and the contrast is adjusted with respect to the image representing the intensity value difference. An oil leakage detection device characterized by recognizing a portion larger than the threshold value as an oil leakage adhesion portion as compared with a predetermined threshold value".

Further, in the present invention, "a method for detecting an oil leak in an oil-filled device, which is a first image obtained when an oil-filled device is photographed by irradiating ultraviolet light, and the first image An image representing the intensity value difference is acquired by acquiring the intensity value difference between the second image obtained by capturing the oil-filled device under the image capturing condition different from the image capturing condition described above by the intensity value of the blue component, and the intensity value The contrast of the image representing the difference is adjusted, and a method for detecting an oil leak in which a portion larger than the threshold value is recognized as an oil leak adhesion portion by comparing with a predetermined threshold value".

According to the present invention, it is possible to realize an oil leak detection device and an oil leak detection method capable of detecting oil leak with high sensitivity even when background noise due to sunlight in the daytime is large.

The figure which shows the experimental result at the time of applying the technique of patent document 1. The figure which shows the schematic structural example of the oil leak detection apparatus 100 which concerns on Example 1 of this invention. The figure which shows a series of processing procedures of the oil leak detection method which concerns on Example 1 of this invention. The pouch figure of the spectrum obtained when irradiating the mineral oil for transformers with an ultraviolet light source. The figure which shows the calculation result of the pixel intensity value by (1) Formula. The figure which shows a series of processing procedures of the oil leak detection method which concerns on Example 2 of this invention. FIG. 11 is a diagram showing an example of a tone curve adjustment method used in image processing of the second embodiment. FIG. 11 is a diagram showing an example of a tone curve adjustment method used in image processing of the second embodiment. FIG. 9 is a diagram showing an image of an experimental result obtained by adjusting the tone curve of Example 2; 9 is a flowchart illustrating an oil leak detection process according to the third embodiment. The flowchart explaining the oil leak detection process which concerns on Example 4. The flowchart explaining the oil leak detection process which concerns on Example 5. The flowchart explaining the oil leak detection process which concerns on Example 6. The flowchart explaining the oil leak detection process which concerns on Example 7. The flowchart explaining the oil leak detection process which concerns on Example 8.

Hereinafter, an oil leak detection device and an oil leak detection method according to an embodiment of the present invention will be described with reference to the drawings. The same reference numerals are used for the same components in each embodiment.

In this embodiment, taking an oil-filled transformer in a substation as an example to inspect a thin oil film attached to the surface of an oil-filled device will be described with reference to FIGS. 2 to 5.

FIG. 2 is a diagram showing a schematic configuration example of the oil leakage detection device 100 according to the first embodiment of the present invention. In this case, the inspection object 1 is an oil-filled device such as a transformer, a capacitor, a hydraulic operating device of GIS (gas insulated switchgear), a rectifier, an inverter, a converter, or the like. In this embodiment, the oil-filled transformer arranged in the substation will be described as the inspection object 1.

The oil leak detection device 100 includes an ultraviolet light source 2, an image pickup device 3, a control device 4 that controls the operations of the

devices

2 and 3, a storage device 5 that stores a captured image, and an image processing device 6 that processes the stored image. , A display device 7 for displaying the processing result. In the present invention, the oil leak detection device 100 captures an image of the oil leak 8 adhering to the surface of the inspection object 1 and displays the inspection result on the display device 7. The constituent parts of the oil leak detection device 100 excluding the ultraviolet light source 2, the image pickup device 3, or the display device 7 are often configured by a computer.

Of these, the UV light source 2 can be a black light, an LED UV light source, or a lamp that contains UV light components, such as a visible light cut filter. The image pickup device 3 can be a general-purpose product such as a digital camera or a surveillance camera that shoots visible light. Note that it is desirable to have a color imaging function.

FIG. 3 is a diagram showing a series of processing steps of the oil leakage detection method according to the first embodiment of the present invention. These procedures show a series of procedures performed from the irradiation of the inspection object 1 to the determination of the presence/absence of oil leakage, which is performed using the arithmetic unit of the computer.

In the flowchart of FIG. 3, in the first processing step S11, the inspection object 1 is kept in a state in which the ultraviolet light source 2 is not irradiated with ultraviolet light, and in this state, an image is taken by the image pickup device 3 in the processing step S21, and the obtained image 1 is obtained. Is stored in the storage device 5. The image 1 at this time is taken by, for example, sunlight in the daytime.

In the next processing step S31, the inspection object 1 is irradiated with the ultraviolet light source 2, and in this irradiation state, an image is taken by the image pickup device 3 in the processing step S41, and the obtained image 2 is stored in the storage device 5. Image 2 at this time was taken with ultraviolet light.

It is needless to say that both Image 1 and Image 2 are images of the same inspection object 1 taken under the same conditions such as the same angle and distance, and that each part on the image shows a corresponding part. Further, the two images are stored in the storage device 5 as digital information for each small area obtained by vertically and horizontally dividing the image area into m×n. The information of each small area includes information on the brightness (intensity) of this area, and the brightness information is represented as, for example, 256 gradation information.

Next, in processing step S51, a difference in luminance (intensity) value is obtained for each of the corresponding small areas of the two

images

1 and 2. However, the difference here is the difference for the blue component of the information about the luminance (intensity) in each small area. By doing so, a black-and-white difference image 3 obtained from the difference is created and saved.

Fig. 4 shows a pouch diagram of the spectrum obtained when a transformer mineral oil was irradiated with an ultraviolet light source having a peak value of 365 nm. The strongest peak was observed around 405 nm. Blue is the main component at this wavelength. Therefore, in the color image pickup device, the color information of the actual object can be measured by the three color filters of red (R), green (G), and blue (B) mounted inside. For example, with blue light, the intensity value of blue (B) is the highest on the resulting image. Compared with the parts other than the blue component, the difference in the intensity value of blue (B) is the largest. As described above, the difference between the image 1 and the image 2 can be maximized by using the intensity value of the blue component.

However, as shown in Fig. 1, in the case of a thin oil film, the oil part cannot be confirmed in image 3 after the difference. Therefore, image processing is further performed on the image after the difference.

Returning to FIG. 2, in the processing step S61, the intensity value of each pixel is adjusted for the image 3 using the equation (1). In the equation (1), I is the intensity value of each pixel of the image 3 before adjustment, and I′ is the intensity value of the pixel after adjustment. γ is a coefficient less than 1.
[Equation 1]
I′=255 (I/255) ^γ
γ <1.........(1)
The adjusted image is saved as image 4. At this time, the post-difference image 3 is dark as a whole, so that the image becomes bright as a whole by using the expression (1).

Regarding the constant value 255, there is no particular problem even if it is changed to an invariable numerical value smaller than 255, for example, the maximum value Imax of the intensity value of each pixel on the image, but if 255 is used, it does not adhere to the oil adhesion site. There may be a large difference between the parts.

As shown in FIG. 5, it is possible to increase the difference in strength value between the oil-attached portion A and the non-attached portion B by using the equation (1). For example, when the intensity value of the oil-attached portion pixel A is 3 and the intensity value of the non-attached portion pixel B is 1 before the adjustment, the value of γ is set to 0.2 by substituting in the above formula (1) The calculation result shows that the intensity value of the pixel in the oil-attached portion A is 164, and the intensity value of the pixel in the non-attached portion B is 84. Along with this, the difference becomes large, and it becomes easier to determine the oil-attached portion A and the portion B that does not adhere. Of course, there is a value of γ that is most easy to determine depending on the image, so the optimum γ may be set accordingly. By adjusting the contrast of the image by this method, the detection sensitivity of the oil-leakage-attached portion A can be improved. Note that not only the equation (1) but also other similar equations exist and can be used as appropriate.

For example, as other formulas for adjusting the contrast, there are formula (2) and formula (3).
[Equation 2]
I′=I+K×(I−(Imax+Imin)/2) (2)
However, I>(Imax+Imin)/2
[Equation 3]
I′=I−K×((Imax+Imin)/2−I) (3)
However, I<(Imax+Imin)/2
Basically, by performing a mathematical calculation using the intensity value of each pixel of the image, the difference between the intensity values of the oil adhered portion A and the non-adhered portion B can be increased to adjust the so-called contrast. Good.

Returning to the flowchart of FIG. 3, in processing step S71, the intensity value I′ of each pixel in the adjusted image 4 is compared with a threshold value Ith for determining oil leakage, and pixels having an intensity value I′ larger than Ith are oil leaked. Recognize as a part.

In implementing the present invention, it is not necessary to perform all the steps in FIG. 3, and only some of the steps may be performed depending on the situation of the place where the operation is performed. Moreover, when the wavelength of the irradiation light source and the color of fluorescence differ depending on the type of oil, this method can be applied by selecting an appropriate light source and blue, red, or green of the camera.

The oil leak detection device of this embodiment enables highly accurate oil leak detection even when background noise due to sunlight in the daytime is large.

Next, an oil leakage detection method according to the second embodiment of the present invention will be described with reference to FIGS. 6, 7(a), 7(b) and 8. In the second embodiment, the configuration of the oil leakage detection device 100 is the same as that of the first embodiment, and only the method of the cooperative processing is different from the first embodiment. Therefore, the common points with the first embodiment will not be described repeatedly.

FIG. 6 is a diagram showing a series of processing steps of the oil leakage detection method according to the second embodiment of the present invention. In this series of flows, processing steps S12 to S52 and processing step S72 perform the same processing as processing steps S11 to S51 and processing step S71 of the first embodiment. Therefore, only the processing step S62, which is a difference from the first embodiment, will be described. In processing step S62, the image 3 is emphasized by utilizing the tone curve adjustment.

The concept of tone curve adjustment will be described with reference to two types of examples shown in FIGS. 7A and 7B.
On the left side of FIGS. 7A and 7B, the intensity value I of each small region of the image 3 before adjustment is shown as the input luminance value on the horizontal axis, and the frequency of occurrence in each gradation of each small region is shown. Is shown on the vertical axis as the output luminance value. This characteristic is that when the intensity value I of each small region is represented by 256 gradations, the frequency of occurrence of each intensity gradation is represented by a histogram. According to this histogram notation, the occurrence frequency concentrates on the luminance of low gradation, and the occurrence frequency decreases extremely in the gradation of the middle area to the high area. This means that the width of the detected gradation is extremely small and it is difficult to discriminate by the difference between the gradations. It can be said that the characteristics on the left side of FIGS. 7A and 7B are those in which the correlation between input and output is represented by a linear curve L.

Therefore, in the second embodiment, the gain correction is performed by adjusting the tone curve so that the gain becomes high especially in the low region. Of these, FIG. 7A gives a gain of saturation characteristic as shown on the right side thereof, and FIG. 7B shows a high gain in the low region as shown on the right side thereof, and a gain in the intermediate region and the high region. By providing an S-shaped characteristic that lowers the output, a slight difference in gradation in the low region can be enlarged and output.

As described above, looking at the histogram shown on the left side of FIG. 7A, the intensity values of the respective pixels of the image 3 that have been subtracted before the image processing are gathered near 0, and the points located on the curve L. The input luminance value and the output luminance value of are the same. On the other hand, as shown on the right side of FIG. 7A, the input luminance value is adjusted to a large output luminance value in a region where the input value is small with respect to the main image, that is, a region where the input luminance value is close to 0, and the saturation characteristic is adjusted. When the shape is adjusted to the shape of the curve L', the oil leakage site is clearly observed. For example, before adjusting the M point, both the input brightness value and the output brightness value are m, but if the M point is adjusted to the N point by increasing the gain, the output brightness value at the input brightness value m point is n. Becomes As a result, the concentrated histogram is flattened, and the contrast between the oil-leakage-attached portion and the non-oil-attached portion is increased.

Note that, as shown on the right side of FIG. 7B, the input luminance value may be adjusted to a small output luminance value in a region having a large input value, such as a region where the input luminance value is close to 255. For example, before adjusting the M′ point, both the input brightness value and the output brightness value are m′, and if the M′ point is adjusted to the N′ point, the output brightness value at the input brightness value m′ point becomes n′. .. Of course, it is desirable to adjust to the optimum adjustment value.

Note that it is effective to set the output brightness value to a value that is 5 times the input value or more, and set the so-called tone curve slope to 5 or more. However, since there is an optimum value depending on the intensity of ultraviolet light and the intensity of ambient light, it is preferable to set the optimum value accordingly.

Fig. 8 is an example of an image adjusted using this method. For the image 1 and the image 2, the blue (B) component of each image could not be observed in the black and white difference image 3 in which the difference of the intensity value can be obtained for each pixel. The oil part is clearly observed. Further, the above adjustment values are stored, and automatic adjustment can be expected.

As described above, according to the oil leak detection method of the present embodiment, it is possible to detect the oil leak with high accuracy even when the background noise due to sunlight in the daytime is large as in the case of the first embodiment. Note that the tone curve adjustment is installed in commercially available image processing software, so if you use these software, you can expect to obtain the confirmation of the oil leakage area with simple adjustment without making special software. it can.

The oil leak detection method according to the third embodiment of the present invention will be described with reference to FIG. In the third embodiment, the configuration of the oil leak detection device 100 is the same as that of the first embodiment, and only the oil leak detection method is different from the first embodiment. Therefore, the common points with the first embodiment will not be described repeatedly.

In the flowchart of FIG. 9, in the first processing step S13, the inspection object 1 is irradiated with ultraviolet light from the ultraviolet light source 2. At this time, in process step S23, an image is captured by the image capturing device 3 with the 405 nm filter being worn in front of the image capturing device 2, and the obtained image 6 is stored in the storage device 5. In process step S33, an image is captured by the image capturing device 3 with the 670 nm filter being worn in front of the image capturing device 2, and the obtained image 7 is stored in the storage device 5.

According to the imaging at the above wavelength, as described in Example 1, the central wavelength of fluorescence is 405 nm, so fluorescence is not completely observed at the wavelength of 670 nm. As a result, the same effect as the presence or absence of irradiation with an ultraviolet light source can be obtained. In the processing step S32, the wavelength need not be limited to 670 nm. If fluorescence is not observed, the value is not specified.

In processing step S43, the difference between the intensity values of the blue components of the

images

6 and 7 is calculated for each pixel. As a result, a black and white post-difference image 8 is obtained.

In processing step S53, the intensity value of each pixel is adjusted using the equation (1) for the image 8. The adjusted image is saved as image 9.

In processing step S63, the adjusted image 9 is used to recognize a pixel larger than Ith as an oil leakage site as compared with a threshold value Ith for determining oil leakage.

By performing the above processing, the oil leak detection according to the present embodiment can detect the oil leak with high accuracy even when the background noise due to the sunlight in the daytime is large as in the case of the first embodiment. Moreover, the on/off control of the ultraviolet light source can be eliminated.

The oil leak detection method according to the fourth embodiment of the present invention will be described with reference to FIG. In the fourth embodiment, the configuration of the oil leak detection device 100 is the same as that of the first embodiment, and only the oil leak detection method is different from the first embodiment. Therefore, the common points with the first embodiment will not be described repeatedly.

FIG. 10 is a diagram showing a series of processing steps of the oil leakage detection method according to the fourth embodiment of the present invention. In this series of flows, processing steps S14 to S44 and processing step S64 perform the same processing as processing steps S13 to S43 and processing step S63 of the third embodiment of FIG. Therefore, only the processing step S54, which is the difference from the third embodiment, will be described. In the processing step S54, instead of adjusting the intensity value of each pixel using the expression (1) for the image 8, the image 3 is emphasized by using the tone curve adjustment.

The oil leak detection method of the present embodiment can obtain the same effects as those of the third embodiment, and since tone curve adjustment is incorporated in commercially available image processing software, it is possible to use these software exclusively. It is expected that you can obtain the confirmation of the oil leakage site with a simple adjustment without making the software.

The oil leak detection method according to the fifth embodiment of the present invention will be described with reference to FIG. In the fifth embodiment, the configuration of the oil leak detection device 100 is the same as that of the first embodiment, and only the oil leak detection method is different from the first embodiment. Therefore, the common points with the first embodiment will not be described repeatedly.

FIG. 11 is a diagram showing a series of processing steps of the oil leakage detection method according to the fifth embodiment of the present invention. In this series of flows, processing step S65 and processing step S75 carry out the same processing as processing step S61 and processing step S71 in the first embodiment of FIG. The processing contents from processing step S15 to processing step S55 are different from those of the first embodiment.

In this flow, in the first processing step S15, the ultraviolet light source is irradiated with the intensity P1, and in the processing step S25, an image is taken by the image pickup device 2 and stored as the image 11. Further, next, in a processing step S35, an ultraviolet light source is irradiated with an intensity P2, and in a processing step S45, an image is photographed by the image pickup device 2 and stored as an image 12.

The ultraviolet light source in process step S15 and process step S35 can be electrically controlled to change the intensity. Further, the distance between the ultraviolet light source and the measurement object can be changed. At this time, different light sources may be used.

In processing step S55, the difference between the intensity values of the blue components of the images 11 and 12 is obtained for each pixel. Then, the black-and-white post-subtraction image 13 is obtained.

As described above, in the oil leak detection apparatus of the present embodiment, the same effect as that of the first embodiment can be obtained, and the control of the ultraviolet light source does not need to be turned on/off. Therefore, the apparatus can be simplified. be able to.

The flowchart of the oil leak detection method according to the sixth embodiment shown in FIG. 12 differs from the flowchart of the oil leak detection method according to the fifth embodiment shown in FIG. 11 only in processing step S65.

In the processing step S65 of FIG. 12, the image 13 is processed using the tone curve adjustment.

The oil leak detection apparatus of the present embodiment can obtain the same effects as those of the fifth embodiment, and the tone curve adjustment is installed in commercially available image processing software. It is expected that you can obtain the confirmation of the oil leakage site with a simple adjustment without making the software.

A flowchart of the oil leakage detection method according to the sixth embodiment will be described with reference to FIG. In the sixth embodiment, the configuration of the oil leak detection device 100 is the same as that of the first embodiment, and only the oil leak detection method is different from the first embodiment. Therefore, the common points with the first embodiment will not be described repeatedly.

In the flow of FIG. 13, in the first processing step S17, the image 14 when the oil leak 8 does not adhere to the surface of the object 1 when the ultraviolet light source 2 is irradiated is stored in the storage unit.

After that, the ultraviolet light source 2 is irradiated in processing step S27, an image is taken by the image pickup device 2 in processing step S37, and the image 15 is stored.

In processing step S47, for image 14 and image 15, the intensity value of the blue component of each image is calculated for each pixel. A black and white post-subtraction image 16 is obtained. In the following cooperative processing, the equation (1) is applied.

The oil leak detection apparatus of the present embodiment can obtain the same effect as that of the first embodiment, and in particular, when the image of the same portion can be taken every time by the installation type apparatus, only one image is taken. This simplifies the detection process.

In Example 8 shown in FIG. 14, the cooperative processing by the equation (1) of the processing step S57 of FIG. 13 adopts the cooperative processing by the tone curve adjustment of the processing step S58.

The oil leak detection apparatus of the present embodiment can obtain the same effects as those of the seventh embodiment, and further, since tone curve adjustment is incorporated in commercially available image processing software, if these softwares are used, exclusive use is possible. It is expected that you can obtain the confirmation of the oil leakage site with a simple adjustment without making the software.

The oil leak detection device according to the first to eighth embodiments described above is, in short, "an oil leak detection device for detecting oil leak in oil-filled equipment,
An ultraviolet light source that irradiates the oil-filled device with ultraviolet light, an imager that captures the oil-filled device, a control device that controls the operation of the ultraviolet light source and the imager, and an image captured by the imager. A storage device, an image processing device that processes an image stored in the storage device, and a display device that displays a processing result of the image processing device,
The image processing apparatus controls the oil-filled device under a shooting condition different from a first image obtained when the oil-filled device is shot by irradiating the ultraviolet light source. An image representing the intensity value difference is acquired by acquiring the intensity value difference between the captured second image and the intensity value of the blue component, and the contrast of the image representing the intensity value difference is adjusted to a predetermined value. An oil leak detection device that recognizes a portion that is larger than the threshold value as an oil leak adhesion portion as compared with the threshold value".

Here, the second image of the oil-filled device is the image 1 in processing step S21 in the first embodiment of FIG. 3, the image 1 in processing step S22 in the second embodiment of FIG. 6, and the third embodiment of FIG. Then, either the image 6 in the processing step S23 or the image 7 in the processing step S33, the image 6 in the processing step S24 or the image 7 in the processing step S34 in the fourth embodiment of FIG. 10, and the processing in the fifth embodiment of FIG. Either image 11 in step S25 or image 12 in processing step S45, either image 11 in processing step S26 or image 12 in processing step S46 in the sixth embodiment of FIG. 12, or processing step S17 in the seventh embodiment of FIG. 14 means the image 14 in the processing step S18 in the eighth embodiment of FIG.

1: Inspection object, 2: Ultraviolet light source, 3: Imaging device, 4: Control device, 5: Storage device, 6: Image processing device, 7: Display device, 8: Oil leak, 100: Oil leak detection device

Claims

An oil leakage detection device for detecting oil leakage of oil-filled equipment,
An ultraviolet light source that irradiates the oil-filled device with ultraviolet light, an imager that captures the oil-filled device, a control device that controls the operation of the ultraviolet light source and the imager, and an image captured by the imager. A storage device, an image processing device that processes an image stored in the storage device, and a display device that displays a processing result of the image processing device.
The image processing apparatus controls the oil-filled device under a shooting condition different from a first image obtained when the oil-filled device is shot by irradiating the ultraviolet light source. An image representing the intensity value difference is acquired by acquiring the intensity value difference from the captured second image with the intensity value of the blue component, and the contrast is adjusted with respect to the image representing the intensity value difference, and it is determined in advance. An oil-leakage detecting apparatus, which recognizes a portion larger than a threshold as an oil-leakage-attached portion in comparison with a threshold.
The oil leakage detection device according to claim 1,
The oil leak detection device, wherein the second image is an image of the oil-filled device taken under imaging conditions in which the ultraviolet light source is not illuminated.
The oil leakage detection device according to claim 1,
The second image is obtained by photographing the oil-filled device under a photographing condition in which a transmission frequency of a filter attached in front of the image pickup device is different from a transmission frequency of the filter in the first image. Oil leak detector.
The oil leakage detection device according to claim 1,
The oil leak detection device, wherein the second image is an image of the oil-filled device taken under imaging conditions in which the intensity of the ultraviolet light source is different from the intensity of the ultraviolet light source in the first image.
The oil leakage detection device according to claim 1,
The oil leakage detection device is characterized in that the second image is taken in a state where oil leakage does not occur in the oil-filled device.
The oil leakage detection device according to any one of claims 1 to 5,
An oil leak detection device, wherein the contrast adjustment method is a tone curve adjustment method.
The oil leakage detection device according to any one of claims 1 to 6,
After adjusting the contrast, the input brightness value is adjusted to a large brightness value in a region where the input brightness value in the image representing the intensity value difference is near 0, and the slope of the tone curve is set to 5 or more. Oil leakage detection device.
The oil leakage detection device according to claim 3,
When irradiating the ultraviolet light source, a filer that transmits only fluorescence and a filer that does not completely transmit fluorescence are mounted in front of the imaging device.
An oil leakage detection method for detecting oil leakage in oil-filled equipment,
A first image obtained when the oil-filled device is photographed by irradiating with ultraviolet light; and a second image obtained by photographing the oil-filled device under photographing conditions different from the photographing conditions of the first image. An image representing the intensity value difference is acquired by acquiring the intensity value difference of the blue component intensity value, the contrast is adjusted with respect to the image representing the intensity value difference, and the threshold value is compared with a predetermined threshold value. An oil leak detection method that recognizes a large area as an oil leak adhesion area.