CN114878094A - Multi-spectral-segment excited oil mark imaging device and detection method - Google Patents

Abstract

The invention discloses a multi-spectrum-segment excited oil stain imaging device and a detection method, wherein in the multi-spectrum-segment excited oil stain imaging device, a multi-spectrum-segment ultraviolet excitation light source system comprises ultraviolet light source arrays with multiple central wavelengths, and the ultraviolet light source arrays are arranged at the end part of a circular table-shaped light source base and emit ultraviolet light to a target area to be detected in the inspection process; the halogen lamp light source system comprises a halogen lamp arranged on a circular truncated cone-shaped light source base, and emits halogen lamp light with uniform light source intensity distribution in a preset wavelength range; the hyperspectral image acquisition system acquires image data with a wave band interval of a preset length in a preset acquisition wavelength range and synchronously acquires three-dimensional data of a reflection spectrum and an image gray level; the data processing unit is connected with the hyperspectral image acquisition system to analyze three-dimensional data of the reflection spectrum and the image gray level, and fusion of multispectral images is carried out to obtain a detection result of whether oil leakage exists.

Description

Multi-spectral excitation oil mark imaging device and detection method

technical field

The invention belongs to the technical field of oil leakage online, in particular to a multi-spectral excitation oil mark imaging device and a detection method.

Background technique

Oil-filled equipment, such as transformers, reactors, current transformers, etc., will inevitably leak oil in the entire industrial process from production to operation, increasing the risk of insulation failure and polluting the environment.

To prevent oil leakage from equipment, several methods have been proposed to detect oil leakage defects. The most commonly used method is the visual inspection method, which is observed by employees with rich work experience through visual observation. The accuracy and detection efficiency of this method completely depend on the rich work experience of the employees. The second common method is the oil gauge reading method, which directly judges whether an oil leakage accident has occurred through the reading change of the oil gauge. Obviously, this method is only suitable for faults with large oil leakage. The third common method is infrared thermal imaging. The detection object of this method is the abnormal temperature rise of power equipment caused by oil leakage fault. Obviously, this method is mainly suitable for faults with large oil leakage, and cannot detect trace oil leakage. Effective detection of faults.

The above information disclosed in this Background section is only for enhancement of understanding of the background of the invention and therefore it may contain information that does not form the prior art that is already known in this country to a person of ordinary skill in the art.

SUMMARY OF THE INVENTION

In view of the problems existing in the prior art, the present invention proposes a multi-spectral excitation oil mark imaging device and a detection method, which utilizes an ultraviolet light source as a fluorescent excitation light source and combines the hyperspectral imaging technology to subdivide the spectrum capability, which overcomes the problem of low light environment. The fatal disadvantage of complex power equipment itself is the difficulty in imaging texture information, which makes it difficult to quickly locate the oil spill area.

The object of the present invention is to be achieved through the following technical solutions, a multi-spectral excitation oil mark imaging device, comprising:

Circular truncated light source base;

A multi-spectral ultraviolet excitation light source system, which includes a plurality of ultraviolet light source arrays with different central wavelengths, the ultraviolet light source arrays are located at the end of the circular frustum-shaped light source base and emit ultraviolet light to the target area to be detected during the inspection process. ;

A halogen lamp light source system, comprising a halogen lamp arranged on the circular frustum-shaped light source base, which emits a halogen lamp with uniform light source intensity distribution within a predetermined wavelength range;

A hyperspectral image acquisition system, which collects image data with a band interval of a predetermined length within a predetermined acquisition wavelength range and simultaneously obtains reflection spectrum-image grayscale three-dimensional data, wherein the spectral image data with the number of imaging bands higher than 100 bands is defined is a hyperspectral image, and the spectral resolution of the hyperspectral image acquisition system is used as the predetermined length;

A data processing unit, which is connected to the hyperspectral image acquisition system to analyze the reflection spectrum-image grayscale three-dimensional data, and performs multi-spectral image fusion to obtain the detection result of the target area to be detected in the inspection process, wherein the The above results can indicate whether there is oil leakage in the target area to be tested;

a central processing unit, which is connected to the multi-spectral ultraviolet excitation light source system, the halogen light source system, the hyperspectral image acquisition system and the data processing unit, wherein,

The central processing unit is used to adjust the light source intensity of the multi-spectral ultraviolet excitation light source system according to the distance of the target area to be measured during the distance inspection process, and switch the emission wavelength after the excitation wavelength image acquisition is completed;

The central processing unit is also used to adjust the light source intensity of the halogen light source system according to the distance of the target area to be measured in the distance inspection process;

The central processing unit is further configured to control the hyperspectral image acquisition system to collect hyperspectral image data, and control the data processing unit to further perform data analysis on the collected hyperspectral image data.

Preferably, the circular cone-shaped light source base is a hollow circular cone, the top of a multi-spectral ultraviolet excitation light source system, a halogen light source system, and a wide-angle lens group of a hyperspectral imaging system for collecting images of oil-leakage areas and non-oil-leakage areas The surface is flush mounted on the end of the hollow circular frustum.

Preferably, the central wavelengths of the ultraviolet light source array are respectively 255nm, 265nm, 315nm and 365nm, the ultraviolet light source array is connected to the central processing unit and the central processing unit controls the emission wavelength and light source intensity, and the plurality of ultraviolet light source arrays are The wide-angle lens group of the hyperspectral imaging system is the center, and the ultraviolet light sources of different wavelengths are arranged at circular intervals and evenly arranged at the end of the circular frustum-shaped light source base.

Preferably, the halogen light source system is symmetrically distributed on the circular frustum-shaped light source base, and is located between the wide-angle lens group of the hyperspectral imaging system and the multi-spectral ultraviolet excitation light source system, and the predetermined wavelength range of the halogen light source is 400 nm -900nm.

Preferably, the predetermined acquisition wavelength range of the hyperspectral image acquisition system is 400 nm-900 nm, and the spectral resolution is 3 nm.

Preferably, the multi-spectral excitation oil streak imaging device further comprises:

a data storage unit, which is connected to the data processing unit to store the detection results of the target area to be tested in the inspection process;

a display unit, which is connected to the data processing unit to visually display the detection result of the target area to be detected in the inspection process;

The central processing unit is connected to the data storage unit and the display unit.

Preferably, both the multi-spectral ultraviolet excitation light source system and the halogen lamp light source system are connected to a power management module for power supply and energy consumption control, and the power management module is connected to the central processing unit.

Preferably, the multi-spectral ultraviolet excitation light source system and the halogen light source system emit uniformly in the form of point light sources, and reach the target area to be measured during the inspection process in the form of parallel light. During the inspection process, the target area to be measured passes through After reflection, it reaches the wide-angle lens group of the hyperspectral imaging system.

Preferably, the detection result of the target area to be detected during the inspection process includes a multi-spectral ultraviolet excitation reflectance value.

In addition, the present invention also discloses a detection method of a multi-spectral excitation oil mark imaging device, which comprises the following steps:

Step S1, turn on the 255nm wavelength ultraviolet excitation light source and adjust the luminous intensity according to the distance from the multi-spectral excitation oil streak imaging device to the target area to be measured during the inspection process; turn on the halogen light source and excite the oil smear imaging device according to the multi-spectral band Adjust the luminous intensity from the distance to the target area to be tested during the inspection process;

Further, obtain a hyperspectral image of the target area to be measured during the inspection process;

Then obtain the hyperspectral image of the standard whiteboard, obtain the hyperspectral image when the lens cover is closed, and obtain the reflection spectrum-image grayscale 3D data of the 255nm ultraviolet light source after normalization;

In step S2, the 255nm band ultraviolet excitation light and the halogen light reach the target area to be tested during the inspection process in the form of parallel light. A photoelectric response intensity curve DN _255nm (x,y,λ) is formed on each pixel point (x,y) of the photoelectric converter of the hyperspectral imaging system, where (x,y) is the pixel point coordinate of the photoelectric converter and is related to the The spatial positions in the hyperspectral image correspond one-to-one, and λ represents the wavelength dimension of the photoelectric response intensity curve of the hyperspectral image acquisition system;

Step S3: Collect the photoelectric response intensity DN _255nmwhite (x, y, λ) of the hyperspectral image acquisition system when the standard whiteboard is used as the target area to be measured in the inspection process and the hyperspectral image when the lens cover of the hyperspectral image acquisition system is closed Collect the photoelectric response intensity of the system DN _255nmblack (x,y,λ), and obtain the average value of DN _255nmwhite (x,y,λ) and DN _255nmblack (x,y,λ) at all coordinate points respectively:

Step S4: After normalizing the photoelectric response intensity curve DN _255nm (x, y, λ) formed on each pixel point (x, y), it is converted into a reflectance value I _{255 nm} (x, y, λ),

选取最大特征值a_max对应的特征值向量

作为不同波段下的反射率值系数，从而实现W个波段数据融合和降维，计算得到测量空间坐标范围内所有坐标点(x,y)的数据融合和降维后反射率值I_255nm(x,y)；Step S5: Decentralize the reflectance value of any coordinate (x _i , y _i ), and calculate the covariance matrix and its eigenvalue a and eigenvalue vector

Select the eigenvalue vector corresponding to the largest eigenvalue a _max

As the reflectivity value coefficient under different wavelength bands, data fusion and dimensionality reduction of W bands are realized, and the reflectivity value I _255nm (x ,y);

Step S6: Turn off the 255nm band ultraviolet excitation light source, turn on the 265nm band UV excitation light source, turn on the halogen light source, acquire the hyperspectral image of the target area to be tested during the inspection process, then acquire the hyperspectral image of the standard whiteboard, and acquire when the lens cover is closed The hyperspectral image is normalized to obtain the reflection spectrum-image grayscale three-dimensional data of the 265nm ultraviolet light source, and S2-S5 are executed to obtain the data fusion and dimension reduction of all coordinate points (x, y) within the measurement space coordinate range. Reflectance value I _265nm (x,y);

Step S7: Turn off the ultraviolet excitation light source in the 265nm band, turn on the ultraviolet excitation light source in the 315nm band, and adjust the luminous intensity according to the distance from the multi-spectral excitation oil streak imaging device to the target area to be tested during the inspection process; keep the halogen light source in an on state, Obtain the hyperspectral image of the target area to be measured during the inspection process;

Further obtain the hyperspectral image of the standard whiteboard, obtain the hyperspectral image when the lens cover is closed, obtain the reflection spectrum-image grayscale three-dimensional data when the 315nm ultraviolet light source is obtained through normalization, and perform steps S2-S5 to obtain all measurements within the spatial coordinate range of the measurement. The reflectance value I _315nm (x, y) after data fusion and dimension reduction of the coordinate point (x, y);

Step S8: Turn off the 315nm band ultraviolet excitation light source, turn on the 365nm band ultraviolet excitation light source, and adjust the luminous intensity according to the distance from the multi-spectral excitation oil streak imaging device to the target area to be tested during the inspection process, and keep the halogen light source in an on state, Obtain the hyperspectral image of the target area to be tested during the inspection process, then obtain the hyperspectral image of the standard whiteboard, obtain the hyperspectral image when the lens cover is closed, and obtain the reflection spectrum-image grayscale 3D data of the 365nm ultraviolet light source after normalization , perform steps S2-S5, obtain the reflectivity value I _365nm (x, y) after data fusion and dimension reduction of all coordinate points (x, y) in the measurement space coordinate range;

Step S9: After steps S1-S8, the reflectance values after data fusion and dimension reduction under 4 ultraviolet excitation light sources are obtained:

I(x,y)=[ _I255nm (x,y) _I265nm (x,y) _I315nm (x,y) _I365nm (x,y)],

Decentralize I(x _i , y _i ) at any coordinate (x _i , y _i ):

Where m is the wavelength sequence number of the ultraviolet excitation light source, m=1, 2, 3, 4;

及其特征值b和特征值向量

选取最大特征值b_max对应的特征值向量

作为不同波段下的反射率值系数，从而获得多谱段反射率值：Calculate the covariance matrix of I _m (x _i ,y _i )

and its eigenvalue b and eigenvalue vector

Select the eigenvalue vector corresponding to the largest eigenvalue b _max

As the coefficient of reflectance value in different bands, the reflectance value of multi-spectral bands can be obtained:

Step S10: When the value of I(x _i , y _i ) is greater than or equal to 1, it is considered that the coordinate (x _i , y _i ) has a fluorescent effect, and it is marked as an oil spill pixel. When I(x _i , y _i ) When the value is less than 1, it is considered that the coordinate (x _i , y _i ) has no fluorescence effect, and it is marked as a pixel point without oil leakage; all coordinate points (x, y) within the range of the target area to be tested are traversed and calculated during the inspection process. The multi-spectral reflectance value I(x,y)={I(x ₁ ,y ₁ ),I(x ₂ ,y ₂ ),...,I(x _i ,y _i )}, can be realized Pixel-level oil spill area judgment, and further mark the judgment results of coordinate points (x, y) in the image under any wavelength band in the reflection spectrum-image grayscale three-dimensional data, so that pixel-level oil marks can be formed.

Compared with the prior art, the present invention has the following advantages: using the ultraviolet light source as the fluorescent excitation light source and combining the hyperspectral imaging technology to subdivide the spectral capability, the traditional fluorescent oil leakage detection is easily disturbed by the environmental light source and can only be applied to the dark environment. In addition, it is difficult to accurately obtain the corresponding fluorescence band images of unknown oil samples, and it is difficult to image the texture information of complex power equipment itself in low light environment, so it is difficult to quickly locate the oil spill area. The visual reconstruction of information and oil leakage areas is convenient for power operation and maintenance personnel to quickly grasp the rapid positioning of trace oil leakage areas of oil-filled equipment, and accurately formulate operation and maintenance strategies.

Description of drawings

Various other advantages and benefits of the present invention will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings in the description are for the purpose of illustrating the preferred embodiments only, and are not to be considered as limiting the present invention. Obviously, the drawings described below are only some embodiments of the present invention, and for those of ordinary skill in the art, other drawings can also be obtained from these drawings without creative effort. Also, the same components are denoted by the same reference numerals throughout the drawings.

In the attached image:

1 is a schematic structural diagram of a multi-spectral excitation oil mark imaging device according to an embodiment of the present invention;

2 is a schematic structural diagram of a truncated light source base of a multi-spectral excitation oil mark imaging device according to an embodiment of the present invention, wherein 3 is a wide-angle lens group of a hyperspectral imaging system, 4 is a multi-spectral ultraviolet excitation light source system, 5 is the halogen light source system;

3 is a schematic diagram of the arrangement of a multi-spectral ultraviolet excitation light source system and a halogen lamp light source system of a multi-spectral excitation oil mark imaging device according to an embodiment of the present invention, wherein A1, A2, and A3 are the center wavelengths of 255nm. UV light source, B1, B2, B3 are UV light sources with a central wavelength of 265 nm, C1, C2, C3 are UV light sources with a central wavelength of 315 nm, D1, D2, D3 are UV light sources with a central wavelength of 365 nm, L1, L2, L3 It is a halogen light source;

Fig. 4 shows the installation method of a circular frustum light source base, a multi-spectral ultraviolet excitation light source system, a halogen lamp light source system and a hyperspectral image acquisition system of a multi-spectral excitation oil mark imaging device according to an embodiment of the present invention, Among them, 1 is the hyperspectral image acquisition system, 2 is the conical light source base, 3 is the wide-angle lens group of the hyperspectral imaging system, 4 is the multi-spectral ultraviolet excitation light source system, and 5 is the halogen light source system;

5 is an image data acquisition step of a multi-spectral excitation oil streak imaging device according to an embodiment of the present invention;

6 is a calculation result of the reflectivity value I _oil (λ) in the oil area and the reflectivity value I _non (λ) in the non-oil area in a multi-spectral excitation oil mark imaging method according to an embodiment of the present invention;

Fig. 7 is the reflectivity value after data fusion and dimension reduction of all coordinate points (x, y) in the measurement space coordinate range obtained by calculation in a multi-spectral excitation oil mark imaging method according to an embodiment of the present invention;

8 is a multi-spectral ultraviolet excitation reflectance value of all coordinate points (x, y) in the measurement space coordinate range obtained by calculation in a multi-spectral excitation oil mark imaging method according to an embodiment of the present invention;

Fig. 9(a) to Fig. 9(d) are the results of detecting oil leakage on the surface of different power equipment for plant-type and mineral-type transformers by a multi-spectral excitation oil mark imaging method according to an embodiment of the present invention; Fig. 9 (a) is the detection result of vegetable transformer oil on the surface of the transformer metal box; Figure 9(b) is the detection result of the mineral transformer oil on the surface of the composite insulator silicone rubber shed; Figure 9(c) is the surface of the pillar insulator porcelain shed The detection result of mineral transformer oil; Figure 9(d) is the detection result of mineral transformer oil on the surface of transformer epoxy resin bushing shed;

FIG. 10 is a schematic structural diagram of a hyperspectral imaging system of a multispectral excitation oil mark imaging device according to an embodiment of the present invention;

11 is a schematic diagram of a three-dimensional data cube of a multi-spectral excitation oil mark imaging method according to an embodiment of the present invention;

12 is a schematic diagram of the field layout and optical path of a method for forming a multi-spectral excitation oil mark according to an embodiment of the present invention;

13 is a schematic diagram of the photoelectric response intensities of a standard white image and a standard black image of a multispectral excitation oil mark imaging method according to an embodiment of the present invention;

FIG. 14 is a schematic diagram of a calculation method for normalizing a reflection spectrum curve of a multi-spectral excitation oil mark imaging method according to an embodiment of the present invention.

The present invention will be further explained below in conjunction with the accompanying drawings and embodiments.

Detailed ways

Specific embodiments of the present invention will be described in more detail below with reference to FIGS. 1 to 14 . While specific embodiments of the present invention are shown in the drawings, it should be understood that the present invention may be embodied in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided so that the present invention will be more thoroughly understood, and will fully convey the scope of the present invention to those skilled in the art.

It should be noted that certain terms are used in the description and claims to refer to specific components. It should be understood by those skilled in the art that the same component may be referred to by different nouns. The description and the claims do not use the difference in terms as a way to distinguish components, but use the difference in function of the components as a criterion for distinguishing. As referred to throughout the specification and claims, "comprising" or "including" is an open-ended term and should be interpreted as "including but not limited to". Subsequent descriptions in the specification are preferred embodiments for implementing the present invention, however, the descriptions are for the purpose of general principles of the specification and are not intended to limit the scope of the present invention. The scope of protection of the present invention should be determined by the appended claims.

To facilitate the understanding of the embodiments of the present invention, the following will take specific embodiments as examples for further explanation and description in conjunction with the accompanying drawings, and each accompanying drawing does not constitute a limitation to the embodiments of the present invention.

For better understanding, as shown in FIGS. 1 to 9( d ), the multi-spectral excitation oil mark imaging device includes a frustum-shaped light source base 10 ;

The multi-spectral ultraviolet excitation light source system 1 includes a plurality of ultraviolet light source arrays with different central wavelengths, and the ultraviolet light source arrays are located at the end of the circular frustum light source base 10 and emit light to the target area to be detected during the inspection process. Ultraviolet light; it should be noted that, if there is an oil leakage area, the present invention actually utilizes the fluorescence characteristics generated by the ultraviolet light to excite the oil leakage area;

A halogen lamp light source system 2, which includes a halogen lamp arranged on the circular frustum-shaped light source base 10, which emits a halogen lamp with uniform light source intensity distribution in a predetermined wavelength range;

A hyperspectral image acquisition system 3, which collects image data within a predetermined wavelength range with the collection band interval of a predetermined length and synchronously obtains reflection spectrum-image grayscale three-dimensional data;

A data processing unit 4, which is connected to the hyperspectral image acquisition system 3 to analyze the reflection spectrum-image grayscale three-dimensional data, and perform multi-spectral image fusion to obtain an oil leakage area detection result;

The central processing unit 7 is connected to the multi-spectral ultraviolet excitation light source system 1, the halogen lamp light source system 2, the hyperspectral image acquisition system 3 and the data processing unit 4, and the central processing unit 7 adjusts the multi-spectral ultraviolet excitation The luminous wavelength and light source intensity of the light source system 1 and the light source intensity of the halogen light source system 2 and the hyperspectral image acquisition system 3 and the data processing unit 4 are controlled.

In the preferred embodiment of the multi-spectral excitation oil mark imaging device, the circular truncated light source base 10 is a hollow circular truncated body, a multi-spectral ultraviolet excitation light source system 1, a halogen lamp light source system 2, and a light source system for collecting oil leakage areas. The top surface of the wide-angle lens group 9 of the hyperspectral imaging system and the image of the non-oil leakage area are installed flush with the end of the hollow circular frustum.

In the preferred embodiment of the multi-spectral excitation oil mark imaging device, the central wavelengths of the ultraviolet light source array are respectively 255nm, 265nm, 315nm and 365nm, and the ultraviolet light source array is connected to the central processing unit 7 and is controlled by the central processing unit. 7. Control the emission wavelength and the intensity of the light source, a plurality of ultraviolet light source arrays are centered on the wide-angle lens group 9 of the hyperspectral imaging system, and the ultraviolet light sources of different wavelengths are arranged at circular intervals and evenly arranged on the circular frustum light source base 10 At the end of the UV light source, the intensity of the multiple ultraviolet light sources is consistent and the light-emitting areas overlap.

In the preferred embodiment of the multi-spectral excitation oil mark imaging device, the halogen light source system 2 is symmetrically distributed on the circular frustum light source base 10, and is located in the wide-angle lens group 9 and the multi-spectral imaging system of the hyperspectral imaging system. Between the ultraviolet excitation light source system 1, the predetermined wavelength range of the halogen light source is 400nm-900nm.

In a preferred embodiment of the multi-spectral excitation oil mark imaging device, the predetermined collection wavelength range is 400nm-900nm, and the spectral resolution is 3nm. It can be understood that the specific parameters here are all examples. Generally, a hyperspectral image acquisition system is a spectral image acquisition system with a wavelength range of 400-1000 or 2500 nm and a spectral resolution below 5 nm. In principle, the predetermined acquisition wavelength range should cover the aforementioned predetermined wavelength range, for example, the halogen lamp is 400-900 nm, and the hyperspectral acquisition is 300-1000 nm, and so on.

In a preferred embodiment of the multi-spectral excitation oil smear imaging device, the multi-spectral excitation oil smear imaging device further comprises:

a data storage unit 5, which is connected to the data processing unit 4 to store the detection result of the oil leakage area,

The display unit 6, which is connected to the data processing unit 4 to visually display the detection result of the oil leakage area,

The central processing unit 7 is connected to the data storage unit 5 and the display unit 6 .

In the preferred embodiment of the multi-spectral excitation oil mark imaging device, the multi-spectral ultraviolet excitation light source system 1 and the halogen lamp light source system 2 are both connected to the power management module 8 for power supply and energy consumption control, so The power management module 8 is connected to the central processing unit 7 .

In the preferred embodiment of the multi-spectral excitation oil mark imaging device, the multi-spectral ultraviolet excitation light source system 1 and the halogen lamp light source system 2 emit uniformly in the form of point light sources, and reach the detection area in the form of parallel light. The non-oil leaking area reaches the wide-angle lens group 9 of the hyperspectral imaging system after reflection.

In the preferred embodiment of the multi-spectral excitation oil mark imaging device, the detection result of the oil region includes the multi-spectral excitation reflectance value.

In one embodiment, a multi-spectral excited oil streak imaging device includes,

The circular frustum light source base 10,

The wide-angle lens group 9 of the hyperspectral imaging system is arranged on the top surface of the circular frustum light source base 10 to collect image signals of the oil leakage area and the non-oil leakage area,

The multi-spectral ultraviolet excitation light source system 1 includes a plurality of ultraviolet light source arrays with different central wavelengths, and: the central wavelengths of the ultraviolet light source arrays are respectively 255 nm, 265 nm, 315 nm and 365 nm, and the ultraviolet light source array is connected to the central processing unit 7 And the central processing unit 7 controls the emission wavelength and the intensity of the light source, a plurality of ultraviolet light source arrays are centered on the wide-angle lens group 9 of the hyperspectral imaging system, and the ultraviolet light sources of different wavelengths are arranged at circular intervals and evenly arranged in the circle. At the end of the table-shaped light source base 10, the intensity of the multiple ultraviolet light sources is consistent and the light-emitting areas overlap;

The halogen lamp light source system 2 is composed of halogen lamps with uniform light source intensity distribution in the wavelength range of 400nm-900nm. The halogen lamp light source system 2 is connected with the central processing unit 7 and the central processing unit 7 controls the light source intensity, wherein the halogen lamp The light source system 2 is used to: provide a light source for imaging the surface texture information of the power equipment under indoor or dark light conditions, and can be turned off under outdoor conditions;

The hyperspectral image acquisition system 3 is controlled by the central processing unit 7 to collect data, wherein the hyperspectral image acquisition system 3 is used for: collecting data images with a wavelength interval of 3 nm in the wavelength range of 400nm-900nm, and synchronously obtaining the reflection spectrum- Image grayscale 3D data;

A data processing unit 4, which is connected to the central processing unit 7, wherein the data processing unit 4 is used for: analyzing the reflection spectrum-image grayscale three-dimensional data, and performing multi-spectral image fusion;

a data storage unit 5, which is connected to the data processing unit 4, wherein the data storage unit 5 is used for: storing and analyzing the detection results of the oil leakage area;

A display unit 6, which is connected to the data processing unit 4, wherein the display unit 6 is used to: visually display the detection result of the oil leakage area;

The central processing unit 7, which is connected with the other units, wherein the central processing unit 7 is used to: realize data collection, data analysis, task assignment, and manual interaction;

The power management module 8 is connected with the other units, wherein the power management module 8 is used for power supply and energy consumption control of each unit.

In the multi-spectral excitation oil mark imaging device, the circular truncated light source base 10 is a hollow truncated truncated body.

In the preferred embodiment of the multi-spectral excitation oil mark imaging device, the circular frustum-shaped light source base 10 is provided with mounting holes for installing the wide-angle lens group 9 of the hyperspectral imaging system and the multi-spectral ultraviolet excitation light source system 1. , the top surfaces of the wide-angle lens group 9 of the hyperspectral imaging system, the multi-spectral ultraviolet excitation light source system 1 and the halogen lamp light source system 2 are flush. The multi-spectral ultraviolet excitation light source system 1 includes 4 groups of 12 ultraviolet light sources with different central wavelengths, namely, ultraviolet light sources A-1, A-2, and A-3 with a central wavelength of 255 nm, and ultraviolet light sources with a central wavelength of 265 nm. Light sources B-1, B-2, B-3, UV light sources C-1, C-2, C-3 with a central wavelength of 315 nm, UV light sources D-1, D-2, D-3 with a central wavelength of 365 nm , wherein 12 ultraviolet light sources are arranged at circular intervals and evenly arranged at the end of the truncated light source base 10, that is, the order of arrangement is A-1, B-1, C-1, D-1 , A-2, B-2, C-2, D-2, A-3, B-3, C-3, D-3, at the same time, the intensity of the light source at the center wavelength of each light source is the same, and the radiation field of view coincides with each other.

In a preferred embodiment of the multi-spectral excitation oil streak imaging device, the halogen lamp light source system 2 includes three halogen lamp light sources L-1, L- 2. L-3, wherein three halogen light sources are axially symmetrically distributed on the circular frustum light source base 10, and are located between the wide-angle lens group 9 of the hyperspectral imaging system and the multispectral ultraviolet excitation light source system 1.

In a preferred embodiment of the multi-spectral excitation oil mark imaging device, the hyperspectral image acquisition system 3 can collect data images with a wavelength interval of 3 nm in the wavelength range of 400 nm-900 nm, and obtain reflection spectrum-images synchronously. Grayscale 3D data.

The multi-spectral excitation oil mark imaging device also includes a data processing unit 4 for analyzing the reflection spectrum-image grayscale three-dimensional data and performing multi-spectral image fusion, so as to store and analyze the detection results of the oil leakage area. The data storage unit 5, the display unit 6 used to visually display the detection results of the oil leakage area, the central processing unit 7 used to realize data collection, data analysis, task allocation, and manual interaction, used for power supply and energy consumption control of each unit. Power management module 8.

The detection method of the multi-spectral excitation oil mark imaging device comprises the following steps:

Step S1, turn on the 255nm band ultraviolet excitation light source, turn on the halogen light source, obtain the hyperspectral image of the target area to be measured during the inspection process (see S2 for the detailed process), and then obtain the standard whiteboard hyperspectral image (see S3 for the detailed process), obtain When the lens cover is closed, the hyperspectral image (see S3 for the detailed process) is normalized to obtain the 255nm excitation reflection spectrum-image grayscale three-dimensional data (see S3 for the detailed process);

Among them, the hyperspectral imager we selected is a push-broom hyperspectral imaging system, as shown in Figure 10, which collects a slit image (that is, a spatial dimension, x) in each imaging cycle, and then passes through the dispersive element. The light is dispersed into spectral information (λ) and detected on a photodetector array. The position of the entrance slit is controlled by a stepper motor, and scanning is performed along another spatial dimension (y) to obtain a three-dimensional spatial spectral cube (x, y, λ). The spectral range of the hyperspectral imager is 400-900nm with a total of 176 bands, and the photodetector pixels are 1936×1456, namely x∈[1,1936], y∈[1,1456].

Among them, the hyperspectral data structure is a three-dimensional data cube, that is, it includes the plane space position information (x, y) and the spectral information (λ), thereby realizing the simultaneous acquisition of 176 images, and contains the position of each pixel (x, y) and spectral information (λ). y), as shown in Figure 11.

In step S2, the 255nm-band ultraviolet excitation light and the halogen light reach the detection area in the form of parallel light. After the target area to be detected is reflected during the inspection process, as shown in FIG. 12 ,

A photoelectric response intensity curve DN _255nm (x, y, λ) is formed on each pixel point (x, y) of the photoelectric converter of the hyperspectral imaging system through the wide-angle lens group of the hyperspectral imaging system, which exemplarily adopts the following Tabular form:

Where (x, y) is the pixel coordinates of the photoelectric converter and corresponds to the spatial position in the hyperspectral image, and λ represents the wavelength of light;

和Step S3: Collect the photoelectric response intensity DN _255nmwhite (x, y, λ) of the standard whiteboard and the photoelectric response intensity DN _255nmblack (x, y, λ) under the lens cover of the hyperspectral image acquisition system with the hyperspectral image acquisition system closed, respectively. DN _255nmwhite (x,y,λ) and DN _255nmblack (x,y,λ) are averaged to obtain:

and

表征了全反射条件下各个波长下光电探测器的响应值，

则表示了在光电探测器在热噪声等作用下的暗计数结果,如图13所示。in,

The response value of the photodetector at each wavelength under the condition of total reflection is characterized,

Then it shows the dark counting result of the photodetector under the action of thermal noise, etc., as shown in Figure 13.

Step S4: After normalizing the photoelectric response intensity curve DN (x, y, λ) formed on each pixel point (x, y), it is converted into a reflectance value I _{255 nm} (x, y, λ),

Exemplarily, I _255nm (x, y, λ) takes the following tabular form:

After normalization, I _255nm (x, y, λ) is transformed into a reflection spectrum curve with physical meaning. Due to the oil leakage, the fluorescence effect in the oil area shows different characteristics from the base material. The difference of the reflectance spectrum curve can realize the identification of oil area and non-oil area.

As shown in Figure 14, it can be seen that the reflectance, that is, the fluorescence, decreases with the increase of the wavelength during oil leakage, and the fluorescence reflectance of the oil trace is higher than that of the non-oil area in some wavelength bands, and in the visible to near-infrared wavelength range On the other hand, due to the thin thickness of the oil trace, the shape of the fluorescence reflectance curve is mainly determined by the substrate material.

选取最大特征值a_max对应的特征值向量

作为不同波段下的反射率值系数，从而实现W个波段数据融合和降维，计算得到测量空间坐标范围内所有坐标点(x,y)的数据融合和降维后反射率值I_255nm(x,y)；Step S5: Decentralize the reflectance value of any coordinate (x _i , y _i ), and calculate the covariance matrix and its eigenvalue a and eigenvalue vector

Select the eigenvalue vector corresponding to the largest eigenvalue a _max

As the reflectivity value coefficient under different wavelength bands, data fusion and dimensionality reduction of W bands are realized, and the reflectivity value I _255nm (x ,y);

Because the fluorescence effect in the oil spill area shows different characteristics from the base material, the reflectance spectrum curves of the oil area and the non-oil area are different. After data fusion and dimensionality reduction, the reflectance value of the oil spill area will be significantly different. higher than the non-oil area, thus realizing oil mark imaging, as shown in Figure 7, which shows the reflectance values of all coordinate points (x, y) in the measurement space coordinate range before data fusion, and after fusion and dimensionality reduction reflectance value.

Step S6: Turn off the 255nm band UV excitation light source, turn on the 265nm band UV excitation light source, keep the halogen lamp light source on, obtain the hyperspectral image of the target area to be tested during the inspection process (see S2 for the detailed process), and then obtain the standard whiteboard height Spectral image (see S3 for the detailed process), obtain the hyperspectral image when the lens cover is closed (the acquisition results of S11 can be directly applied), and obtain the 265nm excitation reflection spectrum-image grayscale 3D data after normalization (see S3 for the detailed process), execute S2-S5, obtain the reflectivity value I _265nm (x, y) after data fusion and dimension reduction of all coordinate points (x, y) in the measurement space coordinate range;

Step S7: Turn off the 265nm band UV excitation light source, turn on the 315nm band UV excitation light source, keep the halogen lamp light source on, obtain the hyperspectral image of the target area to be tested during the inspection process (see S2 for the detailed process), and then obtain the standard whiteboard height Spectral image (see S3 for the detailed process), obtain the hyperspectral image when the lens cover is closed (the acquisition results of S11 can be directly applied), and obtain the 315nm excitation reflection spectrum-image grayscale three-dimensional data after normalization (see S3 for the detailed process), and execute S2-S5, obtain the reflectivity value I _315nm (x, y) after data fusion and dimension reduction of all coordinate points (x, y) in the measurement space coordinate range;

Step S8: Turn off the 315nm band UV excitation light source, turn on the 365nm band UV excitation light source, keep the halogen lamp light source on, acquire the hyperspectral image of the target area to be tested during the inspection process (see S2 for the detailed process), and then acquire the standard whiteboard height Spectral image (see S3 for the detailed process), obtain the hyperspectral image when the lens cover is closed (the acquisition results of S11 can be directly applied), and obtain the 365nm excitation reflection spectrum-image grayscale 3D data after normalization (see S3 for the detailed process), execute S2-S5, obtain the reflectivity value I _365nm (x, y) after data fusion and dimension reduction of all coordinate points (x, y) in the measurement space coordinate range;

Step S9: After steps S1-S8, the reflectance values after data fusion and dimension reduction under 4 ultraviolet excitation light sources are obtained:

I(x,y)=[ _I255nm (x,y) _I265nm (x,y) _I315nm (x,y) _I365nm (x,y)],

Decentralize I(x _i , y _i ) at any coordinate (x _i , y _i ):

Where m is the wavelength sequence number of the ultraviolet excitation light source, m=1, 2, 3, 4;

及其特征值b和特征值向量

选取最大特征值b_max对应的特征值向量

作为不同波段下的反射率值系数，从而获得多谱段反射率值：Calculate the covariance matrix of I _m (x _i ,y _i )

and its eigenvalue b and eigenvalue vector

Select the eigenvalue vector corresponding to the largest eigenvalue b _max

As the reflectance value coefficient under different bands, the multi-spectral reflectance value can be obtained:

Step S10: When the value of I(x _i , y _i ) is greater than or equal to 1, it is considered that the coordinate (x _i , y _i ) has a fluorescent effect, and it is marked as an oil spill pixel. When I(x _i , y _i ) When the value is less than 1, it is considered that the coordinates (xi, yi) have no fluorescence effect, and it is marked as a pixel point without oil leakage; The spectral reflectance value I(x, y)={I(x ₁ , y ₁ ), I(x ₂ , y ₂ ), ..., I(x _i , y _i )}, the pixel level can be realized The oil leakage area is judged, and the judgment result of the coordinate point (x, y) is further marked in the image under any wavelength band in the reflection spectrum-image grayscale three-dimensional data, so that pixel-level oil marks can be formed.

Field Application Verification

1. Measurement of oil leakage on the metal box surface of vegetable oil transformer: use the device and method of the present invention to test the actual plant test transformer, Figure 9 (a) is the detection result, and the result shows that the test result of the device and method of the present invention is accurate and effective;

2. Measurement of the leakage of mineral transformer oil on the surface of the silicone rubber material: using the device and method of the present invention, the silicone rubber material is tested after being leaked by the mineral transformer oil. The results are accurate and effective;

3. Measurement of leakage of mineral transformer oil on the surface of ceramic material: using the device and method of the present invention, the ceramic material is tested after being leaked by mineral transformer oil. Figure 9(c) is the test result, which shows that the test result of the device and method of the present invention is accurate efficient;

4. Measurement of mineral transformer oil leakage on the surface of epoxy resin material: Using the device and method of the present invention, the epoxy resin material was tested after being leaked by mineral transformer oil. Figure 9(d) is the test result, which shows the device and method of the present invention. The test results are accurate and valid.

The above tests are all carried out under strong light conditions, and the results show that the device and method of the present invention can effectively overcome the difficulty of detecting unknown oil sample components under strong light conditions.

Although the embodiments of the present invention have been described above with reference to the accompanying drawings, the present invention is not limited to the above-mentioned specific embodiments and application fields, and the above-mentioned specific embodiments are only illustrative and instructive, rather than restrictive . Those of ordinary skill in the art can also make many forms under the inspiration of this specification and without departing from the scope of protection of the claims of the present invention, which all belong to the protection of the present invention.

Claims (10)

1. A multi-spectral-band excitation oil mark imaging device comprises,

a truncated cone-shaped light source base;

the multispectral ultraviolet excitation light source system comprises a plurality of ultraviolet light source arrays with different central wavelengths, wherein the ultraviolet light source arrays are arranged at the end part of the circular truncated cone-shaped light source base and emit ultraviolet light to a target area to be detected in the inspection process;

a halogen lamp light source system including a halogen lamp provided on the circular truncated cone-shaped light source base, which emits halogen lamp light having a uniform light source intensity distribution in a predetermined wavelength range;

the hyperspectral image acquisition system acquires image data with wave band intervals of a preset length in a preset acquisition wavelength range and synchronously acquires reflection spectrum-image gray three-dimensional data, wherein the spectral image data with the imaging wave band number higher than 100 wave bands is defined as a hyperspectral image, and the spectral resolution of the hyperspectral image acquisition system is taken as the preset length;

the data processing unit is connected with the hyperspectral image acquisition system to analyze the three-dimensional data of the reflection spectrum and the image gray level, and performs multispectral image fusion to obtain a detection result of a target area to be detected in the inspection process, wherein the result can indicate whether the target area to be detected leaks oil or not;

a central processing unit connected with the multi-spectrum ultraviolet excitation light source system, the halogen lamp light source system, the hyperspectral image acquisition system and the data processing unit,

the central processing unit is used for adjusting the light source intensity of the multi-spectrum ultraviolet excitation light source system according to the distance of a target area to be detected in the distance inspection process and switching the light-emitting wavelength after the excitation wavelength image is acquired;

the central processing unit is also used for adjusting the light source intensity of the halogen lamp light source system according to the distance of a target area to be detected in the distance inspection process;

the central processing unit is also used for controlling the hyperspectral image acquisition system to acquire hyperspectral image data and controlling the data processing unit to further perform data analysis on the acquired hyperspectral image data.

2. The multi-spectral excitation oil stain imaging apparatus according to claim 1, wherein preferably the frustum-shaped light source base is a hollow frustum body, and the top surfaces of the multi-spectral ultraviolet excitation light source system, the halogen lamp light source system and the hyperspectral imaging system wide-angle lens group for collecting images of oil leakage area and non-oil leakage area are flush mounted on the end of the hollow frustum body.

3. The multi-spectral-band excitation oil stain imaging device according to claim 2, wherein the central wavelengths of the ultraviolet light source arrays are 255nm, 265nm, 315nm and 365nm respectively, the ultraviolet light source arrays are connected with the central processing unit and the central processing unit controls the light-emitting wavelength and the light source intensity, the ultraviolet light source arrays are centered on the hyperspectral imaging system wide-angle lens group, the ultraviolet light sources with different wavelengths are arranged at circular intervals and are uniformly arranged at the end of the circular truncated cone-shaped light source base, and the light sources of the ultraviolet light sources have the same intensity and the light-emitting areas are overlapped.

4. The multi-spectral excitation oil stain imaging device according to claim 3, wherein the halogen lamp light source system is centrally symmetrically distributed on the circular truncated cone-shaped light source base and is located between the hyperspectral imaging system wide-angle lens group and the multi-spectral excitation light source system of ultraviolet light, and the predetermined wavelength range of the halogen lamp light source is 400nm-900 nm.

5. The multi-spectral-band excitation oil stain imaging device according to claim 1, wherein the predetermined acquisition wavelength range of the hyperspectral image acquisition system is 400nm-900nm, and the spectral resolution is 3 nm.

6. The multi-spectral excitation oil stain imaging apparatus according to any one of claims 1-5, wherein the multi-spectral excitation oil stain imaging apparatus further comprises,

the data storage unit is connected with the data processing unit to store the detection result of the target area to be detected in the inspection process;

the display unit is connected with the data processing unit to visually display the detection result of the target area to be detected in the routing inspection process;

the central processing unit is connected with the data storage unit and the display unit.

7. The multi-spectral excitation oil stain imaging device according to claim 1, wherein the multi-spectral ultraviolet excitation light source system and the halogen lamp light source system are both connected to a power management module for power supply and energy consumption control, and the power management module is connected to the central processing unit.

8. The multi-spectral-segment excited oil stain imaging device according to claim 1, wherein the multi-spectral-segment ultraviolet excitation light source system and the halogen lamp light source system are uniformly emitted in a point light source form, and reach a target area to be measured in an inspection process in a parallel light mode, and reach the hyperspectral imaging system wide-angle lens group after the target area to be measured in the inspection process is reflected.

9. The multi-spectral-band excited oil stain imaging device according to claim 1, wherein the detection result of the target area to be detected in the inspection process comprises a multi-spectral-band ultraviolet excited reflectance value.

10. The method for detecting a multi-spectral-band excitation oil stain imaging device according to any one of claims 1-9, comprising the steps of,

step S1, starting a 255nm wave band ultraviolet excitation light source and adjusting the luminous intensity according to the distance from the multi-spectrum band excitation oil trace imaging device to a target area to be detected in the inspection process; starting a halogen lamp light source and adjusting the luminous intensity according to the distance from the multi-spectrum section excitation oil mark imaging device to a target area to be detected in the inspection process;

further, acquiring a hyperspectral image of a target area to be detected in the inspection process;

then acquiring a hyperspectral image of the standard whiteboard, acquiring the hyperspectral image when a lens cover is closed, and acquiring three-dimensional data of a reflection spectrum-image gray scale when a 255nm ultraviolet light source is obtained through normalization processing;

step S2, enabling the 255nm wave band ultraviolet excitation light and the halogen lamplight to reach a target area to be detected in the inspection process in a parallel light mode, after the target area to be detected in the inspection process is reflected, forming a photoelectric response intensity curve DN on each pixel point (x, y) of a photoelectric converter of the hyperspectral imaging system through the hyperspectral imaging system wide-angle lens group _255nm (x, y, lambda), wherein (x, y) is the pixel point coordinate of the photoelectric converter and corresponds to the spatial position in the hyperspectral image one by one, and lambda represents the wavelength dimension of the photoelectric response intensity curve of the hyperspectral image acquisition system;

step S3: photoelectric response intensity DN of hyperspectral image acquisition system when standard white boards are respectively acquired as target areas to be detected in inspection process _255nmwhite (x, y, lambda) and the photoelectric response intensity DN of the hyperspectral image acquisition system when the lens cover of the hyperspectral image acquisition system is closed _255nmblack (x, y, λ), and DN for all coordinate points, respectively _255nmwhite (x, y, λ) and DN _255nmblack (x, y, λ) is averaged to obtain:

step S4: forming a photoelectric response intensity curve DN for each pixel point (x, y) _255nm (x, y, lambda) is converted into a reflectance value I after normalization treatment _255nm (x，y，λ)，

Step S5: for any coordinate (x) _i ，y _i ) The reflectivity value of the optical fiber is processed by decentralization, and covariance matrix and eigenvalue a and eigenvalue vector thereof are calculated

Selecting the maximum characteristic value a _max Corresponding eigenvalue vector

As reflectivity value coefficients under different wave bands, thereby realizing W wave band data fusion and dimension reduction, and calculating to obtain the reflectivity value I after the data fusion and the dimension reduction of all coordinate points (x, y) in the measurement space coordinate range _255nm (x，y)；

Step S6: closing a 255nm wave band ultraviolet excitation light source, opening a 265nm wave band ultraviolet excitation light source, opening a halogen lamp light source, obtaining a hyperspectral image of a target area to be detected in the inspection process, then obtaining a standard whiteboard hyperspectral image, obtaining a hyperspectral image when a lens cover is closed, obtaining reflection spectrum-image gray three-dimensional data when the 265nm ultraviolet light source is obtained through normalization processing, executing S2-S5, and obtaining data fusion and a reflectivity value I after dimensionality reduction of all coordinate points (x, y) in a measurement space coordinate range _265nm (x，y)；

Step S7: closing a 265nm wave band ultraviolet excitation light source, opening a 315nm wave band ultraviolet excitation light source, and adjusting the luminous intensity according to the distance from a multi-spectrum band excitation oil trace imaging device to a target area to be detected in the inspection process; keeping a halogen lamp light source in an open state, and acquiring a hyperspectral image of a target area to be detected in the inspection process;

further acquiring a standard white board hyperspectral image, acquiring a hyperspectral image when a lens cover is closed, acquiring three-dimensional data of a reflection spectrum and an image gray level when a 315nm ultraviolet light source is obtained through normalization processing, executing steps S2-S5, and obtaining a reflectance value I after data fusion and dimension reduction of all coordinate points (x, y) in a measurement space coordinate range _315nm (x，y)；

Step S8: closing a 315nm wave band ultraviolet excitation light source, opening a 365nm wave band ultraviolet excitation light source, adjusting the luminous intensity according to the distance from a multi-spectrum band excitation oil trace imaging device to a target area to be detected in the inspection process, keeping a halogen lamp light source in an opening state, obtaining a hyperspectral image of the target area to be detected in the inspection process, then obtaining a standard white board hyperspectral image, obtaining a hyperspectral image when a lens cover is closed, obtaining three-dimensional data of reflection spectrum-image gray scale when the 365nm ultraviolet light source is obtained through normalization processing, executing steps S2-S5, and obtaining data fusion of all coordinate points (x, y) in a measurement space coordinate range and a reflectivity value I after dimensionality reduction _365nm (x，y)；

Step S9: after the steps S1-S8, the reflectivity values after data fusion and dimensionality reduction under 4 ultraviolet excitation light sources are obtained:

I(x，y)＝[I _255nm (x，y) I _265nm (x，y) I _315nm (x，y) I _365nm (x，y)]，

for any coordinate (x) _i ，y _i ) I (x) of _i ，y _i ) Performing decentralized processing:

wherein m is the serial number of the wavelength of the ultraviolet excitation light source, and m is 1,2,3, 4;

calculating I _m (x _i ，y _i ) Covariance matrix of

And its eigenvalue b and eigenvalue vector

Selecting the maximum characteristic value b _max Corresponding eigenvalue vector

As a reverse at different wavebandsIndex coefficient, thereby obtaining multi-spectral reflectance values:

step S10: when I (x) _i ，y _i ) When the value is 1 or more, the coordinate (x) is considered _i ，y _i ) The fluorescence effect occurs, and the fluorescence effect is marked as an oil leakage pixel point when I (x) _i ，y _i ) If the value is less than 1, the coordinate (x) is considered _i ，y _i ) Marking the pixel points without oil leakage as the fluorescence effect does not occur; traversing and calculating multi-spectral-segment reflectivity values I (x, y) { I (x, y) } of all coordinate points (x, y) in the target region range to be detected in the routing inspection process ₁ ，y ₁ )，I(x ₂ ，y ₂ )，...，I(x _i ，y _i ) And judging the pixel-level oil leakage area, and further marking the judgment result of the coordinate point (x, y) on an image under any wave band in the three-dimensional data of the reflection spectrum-image gray scale, so that pixel-level oil stain imaging can be realized.

Images