Background technology
When the electric field intensity of overhead transmission line road surfaces exceedes the free intensity of air molecule, air molecule will be ionized, at this moment can hear the discharging sound of " Ci Ci ", smell the smell of ozone, at night, can also see the bluish violet fluorescence sending around wire, this phenomenon is called " corona discharge " or is called for short " corona ".Corona will consume electric energy, and the pulse electromagnetic wave producing during corona discharge can produce interference to radio and high-frequency communication.In addition, corona also can make conductive line surfaces corrode, thus the serviceable life of having reduced wire.
At present, the line data-logging of overhead transmission line is mainly taked far infrared telescope, ultrasonic corona detector and artificial visual inspection.Because the target of corona discharge is little, weak strength, visual being difficult to observed; And in sunshine, contain very strong infrared ray, with metascope, observe false drop rate higher.The response speed of infrared inspection instrument is slow in addition, is unsuitable for taking photo by plane.Someone once attempted to carry out the measurement of corona by the method that radio detector is set along the line, and its unique advantage is can continuous detecting, but expense is high, and accuracy is low, and poor stability is fatal weakness, is difficult to realize.Ultraviolet spectrum Detection Techniques are a kind of new detection methods, use it can under daylight, obtain ultraviolet light image clearly, and reliably convenient, rate of false alarm is low.
The ultraviolet spectrum of corona discharge is mainly at 200~400nm wave band.In air, the peak wavelength of corona discharge is at 300~360nm.But at 300~360nm wave band, earth's surface solar radiation is more much better than than corona, and want much weak in the intensity of " day is blind " ultraviolet band (240~280nm) corona discharge, but now the sun is zero in the background radiation on earth's surface, therefore, be conventionally chosen in non-solar-blind band and carry out corona discharge detection.
The image quality of the ultraviolet imagery system detecting for corona discharge depends primarily on its core component ultraviolet imaging enhancer and " day is blind " ultraviolet filter.The high-end device of this core component there is high-performance, but buy these high performance high-end devices, be subject to strict restriction always.Therefore, usually have to select low side ultraviolet imaging enhancer that performance is lower and low side " day blind " ultraviolet filter as core component.The ultraviolet imaging enhancer noise of low side is large, and image quality is less better; The performance index of ultraviolet filter do not reach high-end advanced level yet in addition.Therefore, under existence conditions, the image quality of system is under some influence, and noise is larger, when signal is weak, and the position that the direct observation of human eye cannot Judging fault point.
Therefore, a kind of ultraviolet image localization of fault treating apparatus need to be proposed, the position of trouble spot can be clearly located.
Utility model content
The problem proposing for solving the utility model, the purpose of this utility model is to provide a kind of system that can ultraviolet dynamic image be analyzed and be processed, and effectively the thermonoise of Background suppression noise and ultraviolet imaging enhancer is also clearly seen position, trouble spot.
Described ultraviolet image localization of fault disposal system comprises spectroscope, day blind filter, visible ray lens, visual light imaging, ultraviolet lens, ultraviolet imagery detector, graphics processing unit and image synthesis unit, spectroscope is divided into transmitted light and reflected light two-way light by target light radiation signal, transmitted light enters visual light imaging CCD through visible ray lens, reflected light first passes through a day blind filter, through ultraviolet lens, enter afterwards, the ultraviolet image of ultraviolet imagery detector is after graphics processing unit carries out image processing, together with the normal image forming with visual light imaging CCD, output to image synthesis unit, processing through image synthesis unit obtains synthetic output image.
Described ultraviolet imagery detector comprises ultraviolet imagery camera lens, day blind ultraviolet filter, photocathode, microchannel plate, video screen and optical fiber cone Coupled CCD Camera, light beam is after ultraviolet imagery camera lens, again by a day blind ultraviolet filter, shine on photocathode, light beam forms photoelectron at photocathode, photoelectron is after microchannel plate 204, and signal is enhanced and amplifies and be converted into visible light signal output by video screen, then by optical fiber cone Coupled CCD Camera, obtains ultraviolet image.
Described ultraviolet imagery detector is identical with the field range of described visual light imaging CCD.
The utility model is analyzed and is processed ultraviolet dynamic image, the effectively thermonoise of Background suppression noise and ultraviolet imaging enhancer, it is originally more difficult fault is positioned in the situation that to make before processing, adopts after this method processes, and can clearly see position, trouble spot.
Embodiment
Fig. 1 is the ultraviolet image localization of fault disposal system structural representation according to an embodiment of the utility model.As shown in Figure 1, ultraviolet imagery system comprises spectroscope 101, day blind filter 102, visible ray lens 103, visual light imaging CCD104, ultraviolet lens 105, ultraviolet imagery detector 106, graphics processing unit 107 and image synthesis unit 108.
As shown in Figure 1, target light radiation signal is divided into transmitted light and reflected light two-way light after 101 points of light actions of spectroscope, and transmitted light enters visual light imaging CCD104 through visible ray lens 103.Reflected light first passes through a day blind filter 102, enters ultraviolet imagery detector 106 afterwards through ultraviolet lens 105.According to an embodiment of the present utility model, can in optical design, make ultraviolet imagery detector 106 identical with the field range of visual light imaging CCD104.
In ultraviolet imagery detector 106 light paths, ultraviolet filter filtering the UV radiation of natural light (280nm~400nm), retained the UV radiation of corona, therefore can detect the corona of solar blind.Visual light imaging CCD104 can the detection of a target normal image.The normal image that the corona image that ultraviolet imagery detector 106 obtains and visual light imaging CCD104 obtain, after image synthesis unit synthesizes, presents a final coloured image, for identification and location to generation corona position.According to an embodiment of the present utility model, the corona ultraviolet image that ultraviolet imagery detector 106 is exported also will carry out after image processing through graphics processing unit 107, output to again image synthesis unit 108, after image synthesis unit 108 is synthetic, obtain a final colored output image with the normal image of visual light imaging CCD104 output.
Fig. 2 is the structural representation of ultraviolet imagery detector 106.Ultraviolet imagery detector comprises ultraviolet imagery camera lens 201, day blind ultraviolet filter 202, photocathode 203, microchannel plate 204, video screen 205 and optical fiber cone Coupled CCD Camera 206.Light beam is after ultraviolet imagery camera lens 201, and some bias light, by filtering, has another part bias light still to exist.Light beam by " day is blind " ultraviolet filter 202, shines on photocathode 203 again.Light beam forms photoelectron at photocathode 203, the photoelectron of launching from photocathode 203 is after microchannel plate 204, signal is enhanced and amplifies and be converted into visible light signal output by video screen 205, then by optical fiber cone Coupled CCD Camera 206, obtains ultraviolet image.According to an embodiment of the present utility model, ultraviolet image is after graphics processing unit 107 is processed through image, finally synthetic with the normal image that visual light imaging CCD104 obtains at image synthesis unit 108, form output image, output image outputs to observed and recorded equipment.
The existing ultraviolet imagery system detecting for corona discharge, because the image quality that affects that is subject to ground unrest and detector thermonoise reduces, noise is larger, when signal is weak, eye-observation cannot Judging fault point position, the utility model detects the feature of ultraviolet imagery system diagram image signal for corona discharge, design a kind of ultraviolet image disposal route detecting for corona discharge, the thermonoise of Background suppression noise and ultraviolet imaging enhancer fast and effectively.Adopt this method, effectively improved the accuracy to effluve localization of fault.
The utility model ultraviolet imagery detector 106 utilizes CCD camera to survey and obtains dynamic image, dynamic image is analyzed and is processed through special algorithm at graphics processing unit 107, the effectively thermonoise of Background suppression noise and ultraviolet imagery detector 106, its treatment scheme as shown in Figure 4.The treatment scheme of graphics processing unit comprises the following steps:
Step 301, input gray level image.
Step 302, adaptive threshold binaryzation.Particularly, gray level image is carried out to adaptive threshold binaryzation, bianry image is joined to the image queue that the length establishing is in advance N again, if queue length is greater than N, delete the 1st frame, what this image queue was stored so is nearest N width bianry image.
Step 303, area of space probability statistics.N width bianry image is analyzed, and the statistics probability that different coordinate points 1 occur in appointed area, obtains a width probabilistic image.By 3X3 fritter, view picture bianry image is divided into a lot of zoness of different, as shown in Figure 4.
In the segment of specifying, occur that 1 just thinks that this region has signal to occur, counting adds 1, in statistics N width bianry image, there is the number of times of signal in appointed area, when occurrence number is greater than threshold value, calculate the signal probability of occurrence of each pixel, when occurrence number is less than threshold value, the signal probability of occurrence of all pixels in this region is zero.Its algorithm is as follows:
Step 304, image filtering.In probabilistic image, the value of each coordinate points represents the probability that on this position, signal occurs, the gray-scale value of every bit pixel of current gray level image and the probability of occurrence of respective coordinates point are multiplied each other, obtain the image that can export, due to the effective thermonoise of Background suppression noise and image intensifier after treatment, can clearly see position, trouble spot.
Step 305, output image.
In test process, video image size is 352*224, frame frequency 25s-1, and bianry image queue length N=8, region unit size is 3*3.After adopting this method processing, most of noise is by filtering, and it is clear that image becomes, and effluve trouble spot is high-visible.
It is a kind of effectively high-tension apparatus fault detect means that corona discharge ultraviolet imagery detects, be subject to the impact of ground unrest and noise of detector, the ultraviolet image Quality Down of output, is not easy to find trouble spot by eye-observation, especially when signal is weak, is just difficult to observe trouble spot.The utility model is analyzed and is processed ultraviolet dynamic image, the effectively thermonoise of Background suppression noise and ultraviolet imaging enhancer, image is through after processing, can clearly observe high-tension apparatus corona discharge picture signal, thereby effectively improve the accuracy to effluve localization of fault.Adopt this method, only need to increase less cost, can realize the real-time processing to ultraviolet image, there is good economic benefits.