CN114419416A - Temperature indicating paint temperature automatic interpretation method and device based on BP neural network - Google Patents

Abstract

The invention provides a temperature indicating paint temperature automatic interpretation method and device based on a BP neural network, wherein a color CCD camera is used for collecting color images; calibrating a camera, and establishing a nonlinear mapping relation between a color image RGB value acquired by a CCD camera and a standard color tristimulus value based on a BP neural network, namely color space conversion; establishing a nonlinear mapping relation between temperature indicating paint color data obtained by color space conversion and BP neural network based on the combination of the color data and standard temperature data; and identifying the temperature of the temperature indicating paint image to be tested, which is acquired by the CCD camera, through the color space conversion and the nonlinear mapping relation of color and temperature. The invention has more accurate temperature identification and high resolution ratio, and ensures the environmental adaptability and accuracy.

Description

Temperature indicating paint temperature automatic interpretation method and device based on BP neural network

Technical Field

The invention relates to the technical field of temperature indicating paint, in particular to a temperature automatic interpretation method and device of temperature indicating paint based on a BP neural network.

Background

With the development of temperature measurement technology, temperature measurement technology of temperature indicating paint is researched. The temperature indicating paint is a paint with a special formula, and can change the color of the paint along with the rise of the external temperature. A determined color temperature relation exists between the color change and the corresponding temperature, and the size and the distribution of the temperature field can be reversely deduced by utilizing the color change rule. The technology is widely used for measuring the surface temperature of hot end parts of the aircraft engine at present. The temperature indicating paint testing technology has the advantages of both contact and non-contact temperature measuring methods. It can be attached to the surface of a temperature measuring device without a lead wire and without a test window. Contact can thus be made without damaging the part being measured, which is not possible with conventional thermocouple thermometry. In addition, the temperature indicating paint coating is very thin, and the structure and the airflow in the engine cannot be influenced after the temperature indicating paint coating is coated. Therefore, the temperature indicating paint test has the advantages of noninterference and noninvasiveness, and can be used in the interior of an engine rotating at a high speed. The temperature testing environment of the aircraft engine is relatively severe, and the aircraft engine is faced with a complex structure of high temperature, high pressure and high-speed rotation, and has a narrow internal space. Due to the characteristics of the temperature indicating paint, the temperature indicating paint can be applied to the surface of an extremely complex structure, accurate and visual temperature distribution is obtained, and a plurality of problems in wall surface measurement are solved.

For many years, temperature interpretation of the temperature paint test has been accomplished by manual operation and visual inspection with the naked eye. The manual interpretation method is very susceptible to the lighting conditions, the visual condition of the testers and psychological factors. These problems adversely affect the accuracy, repeatability and efficiency of the temperature indicating paint test. In addition, the manual testing process is highly dependent on the skill and experience of operators, and the application and popularization of the temperature indicating paint testing technology are not facilitated. Due to various defects of the manual interpretation method, related scholars at home and abroad make many attempts to the automatic interpretation method based on the temperature of the temperature indicating paint in recent years, such as a temperature-indicating paint temperature interpretation method (see patent document: CN109087311A, temperature-indicating paint temperature interpretation method) and a temperature-indicating paint region temperature interpretation method (see patent document: CN108986175A, temperature-indicating paint region temperature interpretation method) proposed by Shenyang engine research in China, and an object surface temperature field three-dimensional reconstruction method based on the temperature indicating paint test technology (see patent document: CN108181022A, and an object surface temperature field three-dimensional reconstruction method based on the temperature indicating paint test technology) proposed by Beijing power machinery research, a color camera is used for shooting a temperature indicating paint sample plate image, the color characteristics of the temperature indicating paint are extracted, and then establishing a temperature interpretation model of the temperature indicating paint, wherein the accuracy of the method is greatly related to the image processing of the temperature indicating paint. Besides, the Shenyang engine design research institute of China aviation industry group company also provides an automatic temperature-indicating paint color identification method (see patent document: CN102313607A, an automatic temperature-color-changing identification method of temperature-indicating paint), the color temperature curve of the temperature-indicating paint is used as an interpretation model, but a fixed linear transformation matrix is adopted in the color space conversion process of the temperature-indicating paint image, namely the camera calibration process, and the method can only be applied to specific environments. These methods suffer more or less from the problems of susceptibility to ambient light and image processing accuracy, complex detection and poor temperature resolution.

Disclosure of Invention

The invention aims to provide a temperature indicating paint temperature automatic interpretation method and device based on a BP neural network, which have high identification precision and strong environment adaptability.

In order to solve the technical problem, according to an aspect of the present invention, there is provided a temperature-indicating paint temperature automatic interpretation method based on a BP neural network, comprising the steps of:

step 1, collecting a color image by using a color CCD camera;

step 2, calibrating the camera, and establishing a nonlinear mapping relation between a color image RGB value acquired by the CCD camera and a standard color tristimulus value based on a BP neural network, namely color space conversion;

step 3, establishing a nonlinear mapping relation between the color data of the temperature indicating paint obtained by the color space conversion in the step 2 and the standard temperature data of the temperature indicating paint based on a BP neural network;

and 4, identifying the temperature of the temperature indicating paint image to be tested, which is acquired by the CCD camera, through the color space conversion in the step 2 and the nonlinear mapping relation of the color and the temperature in the step 3.

According to another aspect of the invention, the invention provides a temperature indicating paint temperature automatic interpretation device based on a BP neural network, which adopts the following technical scheme:

the image acquisition module consists of a color CCD camera and is used for acquiring a color image;

the camera calibration module is used for calibrating the camera, and establishing a nonlinear mapping relation between an RGB value measured by the CCD camera and a standard color tristimulus value based on a BP neural network, namely establishing a color space conversion model of the camera;

the temperature identification model module obtains color data of the temperature indicating paint after color space conversion from the camera calibration module, and establishes a nonlinear mapping relation between the temperature indicating paint and the temperature indicating paint based on a BP neural network by combining the standard temperature data of the temperature indicating paint;

and converting the color space of the temperature indicating paint image to be tested acquired by the color CCD camera through the camera calibration module, and identifying the temperature of the temperature indicating paint image by the temperature identification model module.

The invention provides a temperature indicating paint temperature automatic interpretation method/device based on a BP neural network, which can well realize the color space conversion precision of a temperature indicating paint color image and the temperature interpretation precision of the temperature indicating paint by utilizing the better nonlinear mapping capability of the neural network. According to the camera calibration and temperature identification module/method, a user can adjust the color space conversion and temperature identification of the system through the two modules/methods according to the change of the surrounding environment, so that the temperature identification result is more accurate.

Compared with the prior art, the invention adopts the higher nonlinear mapping capability of the neural network to ensure that the temperature identification is more accurate and the resolution ratio is high; camera calibration and temperature identification model establishment are carried out again at any time according to needs, and environment adaptability and accuracy are guaranteed.

Drawings

The accompanying drawings, which are included to provide a further understanding of the embodiments of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention. It is obvious that the drawings in the following description are only some embodiments of the invention, and that for a person skilled in the art, other drawings can be derived from them without inventive effort.

FIG. 1 is a general framework schematic diagram of a temperature-indicating paint temperature automatic interpretation device based on a BP neural network;

fig. 2 is a flow chart of an automatic temperature-indicating paint temperature interpretation device based on a BP neural network according to an embodiment of the invention.

Detailed Description

It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict. The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

The temperature indicating paint temperature automatic interpretation method based on the BP neural network comprises the following steps:

step 1, collecting a color image by using a color CCD camera.

And 2, calibrating the camera, and establishing a nonlinear mapping relation between the RGB value of the color image acquired by the CCD camera and the tristimulus value of the standard color based on the BP neural network, namely color space conversion.

And 3, establishing a nonlinear mapping relation between the color data of the temperature indicating paint obtained by the color space conversion in the step 2 and the standard temperature data of the temperature indicating paint based on a BP neural network.

And 4, identifying the temperature of the temperature indicating paint image to be tested, which is acquired by the CCD camera, through the color space conversion in the step 2 and the nonlinear mapping relation of the color and the temperature in the step 3.

In some embodiments of the present invention in some embodiments,

in the step 1, a CCD camera collects a standard color card, a standard white board and a standard temperature indicating paint sample plate.

In the step 2, the RGB values of the standard color chart and the standard whiteboard image collected in the step 1 and the corresponding standard color tristimulus values thereof are used as input and output, and a BP neural network is used to train color space conversion, that is, a mapping relationship between the RGB values of the color image collected by the CCD camera and the standard color tristimulus values is established.

In the step 3, the standard temperature indicating paint sample plate image collected in the step 1 is subjected to color conversion according to the step 2, and a nonlinear mapping relation between the standard temperature indicating paint sample plate image and the temperature value is established based on a BP neural network in combination with the corresponding temperature value.

As a further aspect of the invention, the invention provides a temperature indicating paint temperature automatic interpretation device based on a BP neural network.

As shown in fig. 1, the temperature indicating paint temperature automatic interpretation device based on the BP neural network comprises: the device comprises an image acquisition module, a camera calibration module and a temperature identification model module.

The image acquisition module is composed of a color CCD camera and is used for acquiring color images.

The camera calibration module is used for calibrating the camera, namely establishing a color space conversion model of the camera. Because the RGB values measured by different CCD cameras are different, the RGB values measured by the same CCD camera are different even under different lighting environments. Therefore, a nonlinear mapping relation between the RGB value measured by the CCD camera and the standard color tristimulus value is established based on the BP neural network, namely a color space conversion model of the camera is established.

The temperature identification model module is used for establishing a temperature identification model of the temperature indicating paint. The temperature indicating paint can present different specific colors at different temperatures, and the temperature identification of the temperature indicating paint is carried out according to the nonlinear mapping relation of the color and the temperature. And obtaining color data of the temperature indicating paint after color space conversion is carried out on the camera calibration module, and establishing a nonlinear mapping relation between the color data and the standard temperature data based on a BP neural network by combining the color data and the standard temperature data to establish a temperature identification model.

The temperature identification model module is used for identifying the temperature of the temperature indicating paint image to be detected, which is acquired by the color CCD camera, through color space conversion of the camera calibration module.

As shown in fig. 2, in some embodiments of the invention,

and a CCD camera in the image acquisition module acquires a standard color card, a standard white board and a standard temperature indicating paint template.

The camera calibration module takes the acquired RGB values of the standard color card and the standard white board image and the corresponding three stimulus values of the standard color as input and output, and trains a color space conversion model by using a BP neural network.

And the temperature identification model module performs color conversion on the collected standard temperature indicating paint sample plate image according to the color space conversion model, and trains the temperature identification model by using a BP neural network in combination with a corresponding temperature value.

For a calibrated CCD camera, the temperature can be recognized only by shooting the temperature indicating paint to be recognized by the camera and passing the obtained image through a color space conversion model and a temperature recognition model. Under different lighting environments, the RGB values of images obtained by shooting the same object by the CCD camera are different, so that the color space conversion model changes, and finally the temperature identification model changes. If the lighting environment changes without recalibrating the camera, errors in the final temperature identification result may result. The camera calibration module and the temperature identification model module can well solve the problem, and the camera calibration and the temperature identification model establishment are carried out again at any time according to the requirements so as to ensure the accuracy of the result.

The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (4)

1. A temperature indicating paint temperature automatic interpretation method based on a BP neural network is characterized by comprising the following steps:

step 1, collecting a color image by using a color CCD camera;

step 2, calibrating the camera, and establishing a nonlinear mapping relation between a color image RGB value acquired by the CCD camera and a standard color tristimulus value based on a BP neural network, namely color space conversion;

step 3, establishing a nonlinear mapping relation between the color data of the temperature indicating paint obtained by the color space conversion in the step 2 and the standard temperature data of the temperature indicating paint based on a BP neural network;

and 4, identifying the temperature of the temperature indicating paint image to be tested, which is acquired by the CCD camera, through the color space conversion in the step 2 and the nonlinear mapping relation of the color and the temperature in the step 3.

2. The temperature-indicating paint temperature automatic interpretation method based on the BP neural network as claimed in claim 1,

in the step 1, a CCD camera collects a standard color card, a standard white board and a standard temperature indicating paint sample plate;

in the step 2, the RGB values of the standard color card and the standard whiteboard image collected in the step 1 and the corresponding standard color tristimulus values thereof are used as input and output, and a BP neural network is used to train color space conversion, that is, a mapping relationship between the RGB values of the color image collected by the CCD camera and the standard color tristimulus values is established;

in the step 3, the standard temperature indicating paint sample plate image collected in the step 1 is subjected to color conversion according to the step 2, and a nonlinear mapping relation between the standard temperature indicating paint sample plate image and the temperature value is established based on a BP neural network in combination with the corresponding temperature value.

3. A temperature indicating paint temperature automatic interpretation device based on a BP neural network is characterized by comprising an image acquisition module, a camera calibration module and a temperature identification model module;

the image acquisition module consists of a color CCD camera and is used for acquiring a color image;

the camera calibration module is used for calibrating the camera, and establishing a nonlinear mapping relation between an RGB value measured by the CCD camera and a standard color tristimulus value based on a BP neural network, namely establishing a color space conversion model of the camera;

the temperature identification model module obtains color data of the temperature indicating paint after color space conversion from the camera calibration module, and establishes a nonlinear mapping relation between the temperature indicating paint and the temperature indicating paint based on a BP neural network by combining the standard temperature data of the temperature indicating paint;

and converting the color space of the temperature indicating paint image to be tested acquired by the color CCD camera through the camera calibration module, and identifying the temperature of the temperature indicating paint image by the temperature identification model module.

4. The temperature-indicating paint temperature automatic interpretation device based on the BP neural network as claimed in claim 3,

a CCD camera in the image acquisition module acquires a standard color card, a standard white board and a standard temperature indicating paint template;

the camera calibration module takes the acquired RGB values of the standard color card and the standard white board image and the corresponding tri-stimulus values of the standard color as input and output, and trains a color space conversion model by using a BP neural network;

and the temperature identification model module performs color conversion on the collected standard temperature indicating paint sample plate image according to the color space conversion model, and trains the temperature identification model by using a BP neural network in combination with a corresponding temperature value.

Images