CN112683459A - Camera aperture adjusting method and system based on artificial intelligence in air tightness detection process - Google Patents
Camera aperture adjusting method and system based on artificial intelligence in air tightness detection process Download PDFInfo
- Publication number
- CN112683459A CN112683459A CN202011467133.1A CN202011467133A CN112683459A CN 112683459 A CN112683459 A CN 112683459A CN 202011467133 A CN202011467133 A CN 202011467133A CN 112683459 A CN112683459 A CN 112683459A
- Authority
- CN
- China
- Prior art keywords
- aperture
- water body
- image
- turbidity
- optimal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
Images
Landscapes
- Studio Devices (AREA)
Abstract
The invention provides a camera aperture adjusting method and system based on artificial intelligence in an air tightness detection process, and relates to the field of artificial intelligence; the method comprises the steps of collecting an initial image and multiple frames of water body images, carrying out background modeling on the initial image to obtain a background image, and carrying out differential operation on the background image and each frame of water body image to obtain a differential image; superposing the difference images frame by frame, analyzing the bubble obvious degree of the difference images, and determining the superposed frame number when the bubbles are obvious; reflecting the turbidity of the water body according to the illumination intensity of the light source and the illumination intensity of the light after the light passes through the water body, and acquiring a turbidity index, thereby establishing an aperture adjusting model, acquiring the optimal aperture parameter under the minimum superposition frame number, and adjusting the aperture parameter of the camera; and predicting the optimal aperture parameter of the future time period according to the optimal aperture parameter obtained in the historical time period. The method can be used for obtaining the optimal aperture parameter according to the turbidity of the water body in a targeted manner to adjust the aperture parameter of the camera, and reduce data errors, so that obvious bubble characteristics can be accurately acquired.
Description
Technical Field
The invention relates to the field of artificial intelligence, in particular to a camera aperture adjusting method and system based on artificial intelligence in an air tightness detection process.
Background
In the existing air tightness detection method, a commonly used method is that after a sealing device is pressurized, the device is placed in a water body, and the sealing performance of the device is judged by observing generated bubbles. Whether the air tightness is good or not is judged by observing the existence of the air bubbles, and information such as leakage rate is obtained by observing the change of the air bubbles.
In the air tightness detection process, the characteristics of the air bubbles are weak, and especially when the turbidity of water is high, the air bubbles are directly detected, so that detection omission easily occurs. A common bubble feature enhancement method in air tightness detection is a multi-frame superposition method, and obvious features can be obtained by accumulation of bubble features obtained by multiple continuous frames in a time sequence. The traditional automatic aperture method is to automatically adjust the size of the aperture after comparing the brightness of a target area with standard brightness; however, continuous bubbles are generated at the air leakage position in the air tightness detection process, the position and the shape of the bubbles in the water body are continuously changed, the water body is also continuously changed, namely the characteristics of the characteristic area are continuously changed, and therefore the obvious degree of the bubble characteristics acquired by acquiring the water body image by adjusting the diaphragm by using the traditional automatic diaphragm method is greatly different from the ideal obvious degree.
Disclosure of Invention
In order to solve the technical problems, the invention provides a camera aperture adjusting method and system based on artificial intelligence in the air tightness detection process, and the adopted technical scheme is as follows:
in a first aspect, an embodiment of the present invention provides an artificial intelligence based method for adjusting an aperture of a camera in an air tightness detection process, where the method includes the following steps:
acquiring an initial image and a plurality of frames of water body images, wherein the initial image is the water body image acquired when the water body is stable;
performing background modeling on the initial image to obtain a background image, and performing difference operation on the background image and each frame of water body image to obtain a difference image;
superposing the difference images frame by frame, analyzing the bubble obvious degree of the difference images, and determining the superposed frame number when the bubbles are obvious;
reflecting the absorbance of the water body according to the natural logarithm of the difference between the illumination intensity of the light source and the illumination intensity of the light after passing through the water body, and obtaining a turbidity index by taking the absorbance as the turbidity of the water body;
adjusting the aperture size of the camera according to the turbidity index to minimize the number of superposed frames so as to obtain an optimal aperture parameter;
and predicting the optimal aperture parameter of the future time period according to the optimal aperture parameter obtained in the historical time period.
Preferably, the turbidity index refers to an absolute value of a difference value between the turbidity of the water body and the standard turbidity.
Preferably, the optimal aperture parameter is obtained according to an aperture adjustment model, and the size C of the optimal aperture parameter is:
wherein, a and b are undetermined coefficients, and f is a turbidity index.
Preferably, the undetermined coefficient is obtained by fitting by using the aperture size and the turbidity index with obvious bubble characteristics as sample data.
Preferably, the step of obtaining the optimal aperture parameter comprises:
acquiring a plurality of frames of second water body images by using the camera adjusted by the optimal aperture parameter;
and detecting whether the second water body image meets the condition of obvious bubble characteristics, if the bubble characteristics are still not obvious under the optimal aperture parameters, linearly increasing the aperture size until the bubble characteristics are obvious, obtaining the second optimal aperture parameters, and adjusting and updating the optimal aperture parameters according to the difference value of the second optimal aperture parameters and the optimal aperture parameters.
In a second aspect, an embodiment of the present invention provides an artificial intelligence-based camera aperture adjusting system for an air tightness detection process, including:
the system comprises an image acquisition unit, a data acquisition unit and a data processing unit, wherein the image acquisition unit is used for acquiring an initial image and a multi-frame water body image, and the initial image is the water body image acquired when the water body is stable;
the aperture determining unit comprises a difference processing module, a superposition processing module, a turbidity index obtaining module and a model establishing module, wherein:
the differential processing module is used for carrying out background modeling on the initial image to obtain a background image, and carrying out differential operation on the background image and each frame of water body image to obtain a differential image;
the superposition processing module is used for carrying out superposition processing on the difference image frame by frame, analyzing the bubble obvious degree of the difference image and determining the superposition frame number when the bubbles are obvious;
the turbidity index acquisition module is used for reflecting the absorbance of the water body according to the natural logarithm of the difference between the illumination intensity of the light source and the illumination intensity of the light after passing through the water body, and acquiring a turbidity index by taking the absorbance as the turbidity of the water body;
the model establishing module is used for adjusting the aperture size of the camera according to the turbidity index so as to minimize the number of superposed frames and further obtain the optimal aperture parameter;
and an aperture prediction unit for predicting an optimum aperture parameter for a future period from the optimum aperture parameter obtained for the history period.
Further, in the turbidity index obtaining module, the turbidity index refers to an absolute value of a difference value between the turbidity of the water body and the standard turbidity.
Further, in the model building module, the optimal aperture parameter is obtained according to the aperture adjustment model, and the size C of the optimal aperture parameter is as follows:
wherein, a and b are undetermined coefficients, and f is a turbidity index.
Further, the undetermined coefficient is obtained by fitting by using the aperture size and the turbidity index with obvious bubble characteristics as sample data.
Further, in the model building module, after the step of obtaining the optimal aperture parameter, the method further includes:
the image acquisition module is used for acquiring a plurality of frames of second water body images by utilizing the camera adjusted by the optimal aperture parameter;
and the error adjusting module is used for detecting whether the second water body image meets the condition that the bubble characteristics are obvious or not, linearly increasing the aperture size if the bubble characteristics are still not obvious under the optimal aperture parameters until the bubble characteristics are obvious, acquiring the second optimal aperture parameters, and adjusting and updating the optimal aperture parameters according to the difference value between the second optimal aperture parameters and the optimal aperture parameters.
The embodiment of the invention at least comprises the following beneficial effects:
according to the embodiment of the invention, the turbidity index is obtained according to the illumination intensity of the light source, the illumination intensity of the light after passing through the water body and the standard turbidity, the aperture adjusting model is established, the aperture size of the camera is adjusted according to the optimal aperture parameter obtained according to the turbidity of the water body in a targeted manner, and obvious bubble characteristics can be acquired;
according to the embodiment of the invention, a feedback regulation mechanism is introduced to optimize the diaphragm regulation model, so that the error of the model output data is reduced, and the accuracy of the result is improved;
according to the embodiment of the invention, the optimal aperture parameter in the future time period is predicted according to the optimal aperture parameter obtained in the historical time period, the camera aperture parameter is adjusted once by updating data every time period, and the ideal bubble characteristics are acquired from a changed water body on the premise of reducing hardware loss.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions and advantages of the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1 is a flowchart of a camera aperture adjustment method for an air tightness detection process based on artificial intelligence according to an embodiment of the present invention.
Fig. 2 is a schematic diagram of an initial image according to an embodiment of the present invention.
Fig. 3 is a block diagram of a camera aperture adjustment system for an artificial intelligence-based air tightness detection process according to an embodiment of the present invention.
Detailed Description
To further illustrate the technical means and effects of the present invention adopted to achieve the predetermined objects, the following detailed description will be given to a camera aperture adjusting method and system based on artificial intelligence air tightness detection process, and its specific implementation, structure, features and effects thereof according to the present invention with reference to the accompanying drawings and preferred embodiments.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The embodiment of the invention mainly aims at an air tightness detection scene, a device to be detected is placed in a water body, bubbles continuously overflow under fixed pressure, and bubble characteristics are obtained by shooting through a camera. The bubbles are sensitive to light, bright spots appear when the bubbles are irradiated by a light source, and the purpose of enhancing the characteristics of the bubbles is achieved by adjusting the aperture of the camera.
The invention provides a camera aperture adjusting method and a system based on artificial intelligence air tightness detection process.
Referring to fig. 1, a flow chart of a camera aperture adjustment method for an artificial intelligence based air tightness detection process is shown, the method comprising the following steps:
step S100: acquiring an initial image and a multi-frame water body image, wherein the initial image is the water body image acquired when the water body is stable.
The method comprises the steps of firstly deploying a camera, deploying a high-speed camera with a fixed pose at one side of a glass water tank for air tightness detection, deploying a light source under the high-speed camera, and acquiring an initial image after the water surface in the water tank is stable. Referring to fig. 2, which shows a schematic diagram of an initial image, a high-speed camera captures an image of a body of water 20 below a liquid level 10 at an orthographic angle to obtain an initial image 30.
The device to be measured is placed in a water body, the pressure is increased to a certain degree, continuous shooting is carried out by using a high-speed camera after the bubble escape rate is relatively stable, multi-frame water body images are obtained, the water body images contain bubble characteristics, and the high-speed camera can reduce bubble smear in the water body images. The initial image and the water body image do not contain a device to be tested and pressurizing equipment.
Step S200: and carrying out background modeling on the initial image to obtain a background image, and carrying out difference operation on the background image and each frame of water body image to obtain a difference image.
And carrying out background modeling on the initial image to obtain a background image.
And respectively placing each frame of water body image and the background image in an image coordinate system for background difference operation to obtain a difference image. The specific calculation formula is as follows:
wherein G isi(x, y) is the coordinates of the ith frame of water body image, Bi(x, y) is the coordinates of the background image of the i-th frame, DiAnd (x, y) is the coordinate of the differential image of the ith frame. T is a threshold value and is obtained by a maximum inter-class variance method.
Step S300: and (4) superposing the difference images frame by frame, analyzing the bubble obvious degree of the difference images, and determining the superposed frame number when the bubbles are obvious.
The obtained difference images are overlapped frame by frame, the foreground part of the obtained overlapped image is the bubble characteristic, when the overlapping operation or the aperture is adjusted to be large, the exposure of the overlapped image is influenced, along with the continuous increase of the exposure, the over-exposure phenomenon gradually appears in the overlapped image, the refraction and reflection of the bubbles to light are stronger than that of a water body, the brightness of the bubbles in the over-exposure area is larger than that of the water body, and the over-exposure area can also become the foreground area at the moment, so that the judgment of the bubble characteristic is influenced. Therefore, it is necessary to process the image after each superimposition to distinguish the overexposed region from the bubble region.
Obtaining a binary image of the overlay by using a maximum inter-class variance method; carrying out Gaussian filtering on the binary image to remove noise in the image; carrying out contour detection on the denoised image to obtain contour information of each connected domain and further obtain the minimum circumscribed rectangle of the contour; according to the priori knowledge, the bubble features are approximate to an elliptical shape, and the rising direction is vertical rising, so the length-width ratio of the minimum circumscribed rectangle of the bubble area is larger than that of the minimum circumscribed rectangle of the overexposure area; setting an experience threshold K, if h/w is larger than K, determining the area as a bubble area, and otherwise, determining the area as an overexposure area; wherein h is the length of the minimum circumscribed rectangle, and w is the width of the minimum circumscribed rectangle.
And after the influence of the overexposure area is eliminated, analyzing the bubble area of the binary image, and judging the obvious degree of the bubble characteristics.
Counting the number of white pixel points in the minimum circumscribed rectangle of the bubble region, and taking the ratio of the bubble region in the overlay map as the degree Pro of the obvious bubble characteristics, wherein the calculation formula is as follows:
wherein, W is the width of the superposition map, H is the length of the superposition map, namely the superposition map has W rows and H columns of pixels;
sum is the total number of white pixels in all bubble regions.
Setting an experience threshold D according to the use requirement of an actual scene, if Pro is less than D, overlapping a next frame of difference graph to obtain a new overlapping graph, and performing processing analysis again; otherwise, judging that the bubble characteristics are obvious, wherein at the moment, the corresponding superposition frame number N reflects the obvious degree of the bubbles in the water body image, and the larger N is, the less obvious the bubble characteristics in the water body image are.
Step S400: and reflecting the absorbance of the water body according to the natural logarithm of the difference between the illumination intensity of the light source and the illumination intensity of the light after passing through the water body, and acquiring a turbidity index by taking the absorbance as the turbidity of the water body.
And (5) sensing the turbidity of the water body by using the light intensity sensor deployed in the step S100. The illumination intensity of the light source is constant, the light of the light source reaches the light intensity sensor on the other side through the water body and can be attenuated to a certain extent, the absorbance of the water body is reflected according to the illumination intensity of the light source and the illumination intensity of the light collected by the light intensity sensor after passing through the water body, the turbidity degree F of the water body is reflected according to the absorbance, and the calculation formula is as follows:
F=ln(I0-I)
wherein, I0The light intensity of the light source is shown, and I is the number indicated by the light intensity sensor, namely the light intensity after the light passes through the water body.
In order to eliminate the influence of the water tank glass on the turbidity perception, the embodiment of the invention takes the turbidity of the water body during clarification as the standard turbidity F0Then, the final water turbidity index f is calculated as follows:
f=|F-F0|
the larger the f value, the more turbid the water body.
Step S500: and adjusting the aperture size of the camera according to the turbidity index so as to minimize the number of superposed frames and further obtain the optimal aperture parameter.
And constructing an aperture adjusting model according to the obtained turbidity index, and adjusting the aperture parameter of the camera through the turbidity information sensed by the sensor so as to minimize the number N of the superposed frames. According to the priori knowledge, the larger the turbidity of the water is, the larger the aperture should be adjusted to obtain obvious bubble characteristics, and the size C of the optimal aperture parameter at the moment is as follows:
wherein a and b are undetermined coefficients.
In the embodiment of the invention, the value of the coefficient to be determined is determined by a data fitting mode, and the specific method comprises the following steps:
setting the aperture parameters to linearly change within a certain range under the same turbidity index, wherein the specific change process is that the aperture parameters are linearly increased from the minimum value to the maximum value and then decreased from the maximum value to the minimum value at the same speed, and the above steps are circularly changed; and (5) acquiring a background image and a water body image according to the method in the step S100, and recording the aperture size of the camera when the bubble characteristics of the water body image are obvious and the corresponding number of superposed frames N is less than or equal to N according to the methods in the steps S200 and S300. And selecting the aperture parameter with the minimum N value and the corresponding turbidity index as a group of sample data.
N is an empirical threshold, and the maximum aperture value is set so as to avoid the phenomenon of overexposure distortion of the overlay image after the overlay of n frames. In the embodiment of the present invention, n is 10.
Corresponding to different water turbidity indexes, multiple experiments are carried out according to the fitting method to obtain different information sequences { f1,f2,f3,……fmAnd the corresponding { C }1,C2,C3,……CmAnd m is the data acquisition times, namely m groups of sample data are acquired, and the ith group of sample data is { f }i→Ci}。
Fitting the aperture adjusting model by using m groups of collected sample data, evaluating the fitting degree of the model by using a mean square error loss function until the loss function reaches the minimum, confirming each coefficient, and completing the fitting of the aperture adjusting model.
Therefore, an aperture adjusting model is preliminarily determined, and the model can calculate the size of the optimal aperture parameter according to the turbidity information.
Preferably, in order to make the model more accurate, the embodiment of the present invention introduces a feedback adjustment mechanism to optimize the aperture adjustment model. In an actual detection scene, after camera aperture parameters are adjusted to C according to turbidity indexes, verification is continued, bubble obvious degree of a water body image is obtained according to steps S100 to S300 in the same way, if bubble characteristics are still not obvious, aperture size is linearly increased until the bubble characteristics are obvious, the difference value delta C between the aperture size C' and C at the moment is recorded, multiple groups of data are collected to confirm the difference value delta C until model errors are relatively stable, a feedback mechanism is removed, and the difference value delta C is used as error adjustment of an aperture adjustment model and the optimal aperture parameters are updated.
After the feedback regulation mechanism is optimized, the accuracy of the aperture regulation model is improved, and the image which can be acquired by the camera under the optimal aperture parameter contains obvious bubble characteristics.
Step S600: and predicting the optimal aperture parameter of the future time period according to the optimal aperture parameter obtained in the historical time period.
It should be noted that the aperture adjustment model obtained in step S500 is a real-time change model, the turbidity is obtained according to the readings of the sensor, and the aperture size required at the current time is obtained through the aperture adjustment model. Specifically, in the actual situation, the change of the turbidity of the water body is slow, and the camera parameters do not need to be changed all the time, the time sequence prediction model is established, the optimal aperture parameters in the future time period t are accurately predicted according to the historical data of the time period S, the camera aperture parameters are adjusted by updating data every time period t, and the time sequence prediction model can acquire ideal bubble characteristics from the changed water body on the premise of reducing hardware loss.
Preferably, in the embodiment of the present invention, the TCN network is taken as an example to specifically describe the construction of the time sequence prediction model, and the required data set is the optimal aperture parameter output by the aperture adjustment model. Timing prediction networks such as LSTM networks, GRU networks, TCN networks, etc. may also be employed in other embodiments.
Specifically, the training process of the TCN network is as follows:
for convenience in training, a data set and a data label are normalized, the shape of an input TCN network is [ B, N, 1], B is the size of batch input by the network, N is the quantity of data acquired in an S time period, in the embodiment of the invention, a group of data is acquired in each t time period, the size of a camera aperture after the future t time period is predicted based on historical data in the S time period, and then N is S/t; and after the TCN extracts the features of the data, obtaining a feature descriptor of the data, connecting the full connection layer, and finally outputting the shape of [ B, 1 ]. And the TCN network is trained by adopting a cross entropy loss function, so that the result is more accurate.
And (3) putting the trained TCN into use, wherein the TCN network is not output in the stage of data acquisition in the initial S time period, so that the real-time aperture value is obtained according to the output of the aperture adjustment model in the S time period after the air tightness detection is started, and then the TCN network is called to predict the optimal aperture parameter at the future time t, so as to ensure that obvious bubble characteristics are continuously obtained.
The following example illustrates the case where S is 1 hour and t is 0.2 hour, that is, a method for predicting the aperture size required by the camera in the future 12 minutes from the 1 hour history data.
TCN network input was recorded every 12 minutes for the past 1 hourThe recorded sequence form of the optimal aperture parameters output by the aperture adjustment model is as follows: { Ca,Cb,Cc,Cd,CeAfter TCN network processing, output data { CfAs predicted optimal aperture parameters in the future 12 minutes for adjusting the camera aperture size. After 12 minutes, the TCN network continues to input new historical sequence data Cb,Cc,Cd,Ce,CfAnd, again predicting the best aperture parameter.
In summary, the embodiment of the invention provides an artificial intelligence-based camera aperture adjusting method in an air tightness detection process, the method comprises the steps of obtaining a turbidity index according to the illumination intensity of a light source, the illumination intensity of light after passing through a water body and standard turbidity, establishing an aperture adjusting model, obtaining the optimal aperture parameter according to the turbidity of the water body in a targeted manner, adjusting the aperture size of a camera, and acquiring obvious bubble characteristics; a feedback regulation mechanism is introduced to optimize the diaphragm regulation model, so that the error of the model output data is reduced, and the accuracy of the result is improved; the historical optimal aperture parameters output by the aperture adjusting model are used as a data set, and the aperture parameters in the future time period are obtained by utilizing a time sequence prediction network, so that the aperture parameters of the camera can be adjusted in time under the condition of protecting hardware facilities.
Based on the same inventive concept as the method embodiment, another embodiment of the present invention provides an air tightness detection process camera aperture adjustment system based on artificial intelligence.
Referring to fig. 3, it shows a block diagram of a camera aperture adjustment system 100 based on artificial intelligence air tightness detection process according to another embodiment of the present invention, which includes an image acquisition unit 40, an aperture determination unit 50 and an aperture prediction unit 60.
Specifically, the image acquiring unit 40 is configured to acquire an initial image and a multi-frame water body image, where the initial image is a water body image acquired when the water body is stable;
the aperture determination unit 50 includes a difference processing module 51, a superposition processing module 52, a turbidity index obtaining module 53, and a model establishing module 54, wherein:
the difference processing module 51 is configured to perform background modeling on the initial image to obtain a background image, and perform difference operation with each frame of water body image to obtain a difference image;
the superposition processing module 52 is configured to perform superposition processing on the difference image frame by frame, analyze the bubble significance degree of the difference image, and determine the number of superposition frames when bubbles are significant;
the turbidity index acquisition module 53 is configured to reflect the absorbance of the water body according to the natural logarithm of the difference between the illumination intensity of the light source and the illumination intensity of the light after passing through the water body, and acquire a turbidity index by using the absorbance as the turbidity of the water body;
the model establishing module 54 is used for adjusting the aperture size of the camera according to the turbidity index so as to minimize the number of superimposed frames and further obtain the optimal aperture parameter;
an aperture prediction unit 60 for predicting an optimum aperture parameter for a future period from the optimum aperture parameter obtained for the history period.
Further, in the turbidity index obtaining module 53, the turbidity index refers to an absolute value of a difference between the water turbidity and the standard turbidity.
Further, in the model building module 54, the optimal aperture parameter is obtained according to the aperture adjustment model, and the size C of the optimal aperture parameter is:
wherein, a and b are undetermined coefficients, and f is a turbidity index.
Further, the undetermined coefficient is obtained by fitting by using the aperture size and the turbidity index with obvious bubble characteristics as sample data.
Further, in the model building module 54, the step of obtaining the optimal aperture parameter further includes:
the image acquisition module is used for acquiring a plurality of frames of second water body images by utilizing the camera adjusted by the optimal aperture parameter;
and the error adjusting module is used for detecting whether the second water body image meets the condition that the bubble characteristics are obvious or not, linearly increasing the aperture size if the bubble characteristics are still not obvious under the optimal aperture parameters until the bubble characteristics are obvious, acquiring the second optimal aperture parameters, and adjusting and updating the optimal aperture parameters according to the difference value between the second optimal aperture parameters and the optimal aperture parameters.
In conclusion, the invention provides an artificial intelligence-based camera aperture adjusting system for an air tightness detection process, which comprises a turbidity index acquisition module and a model establishment module, wherein the turbidity index acquisition module and the model establishment module are used for acquiring optimal aperture parameters according to the turbidity of a water body in a targeted manner to adjust the aperture size of a camera and acquire obvious bubble characteristics; an error adjusting module is used for optimizing the diaphragm adjusting model, so that the error of the output data of the model is reduced, and the accuracy of the result is improved; the aperture prediction unit is used for predicting the aperture parameters in the future time period, so that the aperture parameters of the camera can be adjusted in time under the condition of protecting hardware facilities.
It should be noted that: the precedence order of the above embodiments of the present invention is only for description, and does not represent the merits of the embodiments. And specific embodiments thereof have been described above. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments.
The present invention is not limited to the above preferred embodiments, and any modifications, equivalent replacements, improvements, etc. within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A camera aperture adjusting method based on artificial intelligence in an air tightness detection process is characterized by comprising the following steps:
acquiring an initial image and a multi-frame water body image, wherein the initial image is the water body image acquired when the water body is stable;
carrying out background modeling on the initial image to obtain a background image, and carrying out difference operation on the background image and each frame of water body image to obtain a difference image;
superposing the difference images frame by frame, analyzing the bubble obvious degree of the difference images, and determining the superposed frame number when bubbles are obvious;
reflecting the absorbance of the water body according to the natural logarithm of the difference between the illumination intensity of the light source and the illumination intensity of the light after passing through the water body, and obtaining a turbidity index by taking the absorbance as the turbidity of the water body;
adjusting the aperture size of the camera according to the turbidity index to minimize the number of the superimposed frames so as to obtain an optimal aperture parameter;
and predicting the optimal aperture parameter of the future time period according to the optimal aperture parameter obtained in the historical time period.
2. The method for adjusting the aperture of the camera in the air tightness detection process based on the artificial intelligence as claimed in claim 1, wherein the turbidity index is an absolute value of a difference value between the turbidity of the water body and a standard turbidity.
3. The method for adjusting an aperture of a camera in an airtightness detection process based on artificial intelligence, according to claim 1, wherein the optimal aperture parameter is obtained according to an aperture adjustment model, and the size C of the optimal aperture parameter is:
wherein, a and b are undetermined coefficients, and f is a turbidity index.
4. The method for adjusting the aperture of the camera in the air tightness detection process based on the artificial intelligence as claimed in claim 3, wherein the undetermined coefficient is obtained by fitting using the aperture size under the condition that the characteristics of the bubbles are obvious and the turbidity index as sample data.
5. The method for adjusting an aperture of a camera during airtightness detection based on artificial intelligence, according to claim 1, wherein the step of obtaining the optimal aperture parameter is followed by:
acquiring a plurality of frames of second water body images by using the camera adjusted by the optimal aperture parameter;
and detecting whether the second water body image meets the condition of obvious bubble characteristics, if the bubble characteristics are still not obvious under the optimal aperture parameters, linearly increasing the aperture size until the bubble characteristics are obvious, obtaining the second optimal aperture parameters, and adjusting and updating the optimal aperture parameters according to the difference value between the second optimal aperture parameters and the optimal aperture parameters.
6. The utility model provides an air tightness testing process camera light ring governing system based on artificial intelligence which characterized in that includes:
the system comprises an image acquisition unit, a data acquisition unit and a data processing unit, wherein the image acquisition unit is used for acquiring an initial image and a multi-frame water body image, and the initial image is the water body image acquired when the water body is stable;
the aperture determining unit comprises a difference processing module, a superposition processing module, a turbidity index obtaining module and a model establishing module, wherein:
the difference processing module is used for carrying out background modeling on the initial image to obtain a background image, and carrying out difference operation on the background image and each frame of water body image to obtain a difference image;
the superposition processing module is used for carrying out superposition processing on the difference image frame by frame, analyzing the bubble obvious degree of the difference image and determining the superposition frame number when the bubbles are obvious;
the turbidity index acquisition module is used for reflecting the absorbance of the water body according to the natural logarithm of the difference between the illumination intensity of the light source and the illumination intensity of the light after the light passes through the water body, and acquiring a turbidity index by taking the absorbance as the turbidity of the water body;
the model establishing module is used for adjusting the aperture size of the camera according to the turbidity index so as to minimize the superposed frame number and further obtain the optimal aperture parameter;
an aperture prediction unit for predicting an optimum aperture parameter for a future period from the optimum aperture parameter obtained for a history period.
7. The system for adjusting an aperture of a camera in an airtight detection process based on artificial intelligence as claimed in claim 6, wherein in the turbidity index obtaining module, the turbidity index is an absolute value of a difference between a turbidity of a water body and a standard turbidity.
9. The system of claim 8, wherein the undetermined coefficient is fit by using the aperture size with obvious bubble characteristics and the turbidity index as sample data.
10. The system of claim 6, wherein the step of obtaining optimal aperture parameters in the model building module further comprises:
the image acquisition module is used for acquiring a plurality of frames of second water body images by utilizing the camera adjusted by the optimal aperture parameter;
and the error adjusting module is used for detecting whether the second water body image meets the condition that the bubble characteristics are obvious or not, linearly increasing the aperture size if the bubble characteristics are still not obvious under the optimal aperture parameters until the bubble characteristics are obvious, acquiring second optimal aperture parameters, and adjusting and updating the optimal aperture parameters according to the difference value between the second optimal aperture parameters and the optimal aperture parameters.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011467133.1A CN112683459A (en) | 2020-12-14 | 2020-12-14 | Camera aperture adjusting method and system based on artificial intelligence in air tightness detection process |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011467133.1A CN112683459A (en) | 2020-12-14 | 2020-12-14 | Camera aperture adjusting method and system based on artificial intelligence in air tightness detection process |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112683459A true CN112683459A (en) | 2021-04-20 |
Family
ID=75447596
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202011467133.1A Withdrawn CN112683459A (en) | 2020-12-14 | 2020-12-14 | Camera aperture adjusting method and system based on artificial intelligence in air tightness detection process |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112683459A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113567058A (en) * | 2021-09-22 | 2021-10-29 | 南通中煌工具有限公司 | Light source parameter adjusting method based on artificial intelligence and visual perception |
-
2020
- 2020-12-14 CN CN202011467133.1A patent/CN112683459A/en not_active Withdrawn
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113567058A (en) * | 2021-09-22 | 2021-10-29 | 南通中煌工具有限公司 | Light source parameter adjusting method based on artificial intelligence and visual perception |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108918539B (en) | Apparent disease detection device and method for tunnel structure | |
CN113469177B (en) | Deep learning-based drainage pipeline defect detection method and system | |
CN111598156B (en) | PM based on multi-source heterogeneous data fusion2.5Prediction method | |
CN112785560A (en) | Air tightness detection water body updating method and system based on artificial intelligence | |
CN109167928A (en) | Fast automatic exposure method and system based on defects of display panel detection | |
CN101819024B (en) | Machine vision-based two-dimensional displacement detection method | |
CN116310845B (en) | Intelligent monitoring system for sewage treatment | |
CN114792314B (en) | Laser beam-based metal mesh defect detection method and artificial intelligence system | |
CN116030060A (en) | Plastic particle quality detection method | |
CN112258544A (en) | Camera filter self-adaptive switching method and system based on artificial intelligence | |
CN112329620A (en) | Water change amount prediction system of tightness detection device based on visual perception | |
CN116416252B (en) | Method for detecting sedimentation image of wastewater in boehmite production process | |
KR20090047657A (en) | System for meserment of the snowfall and method for meserment of the snowfall | |
CN111487225A (en) | Rapid and high-precision water quality transparency measuring device and measuring method thereof | |
CN112683459A (en) | Camera aperture adjusting method and system based on artificial intelligence in air tightness detection process | |
KR20230132686A (en) | A method for damage identification and volume quantification of concrete pipes based on PointNet++ neural network | |
CN112330695A (en) | Automatic water adding and changing decision method and system based on artificial intelligence | |
CN110533626B (en) | All-weather water quality identification method | |
CN109949725B (en) | Image gray level standardization method and system for AOI system | |
CN114639064A (en) | Water level identification method and device | |
CN113554632A (en) | Artificial intelligence-based workpiece detection optimization method in turbid water body | |
CN117115610A (en) | Metal surface defect detection method and system based on multi-angle image exposure fusion | |
CN109856133B (en) | Test paper detection method utilizing multiple illumination intensities and multiple color illumination | |
CN113436120B (en) | Image fuzzy value identification method and device | |
CN109544481B (en) | Aviation image dodging method and system based on machine learning |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20210420 |
|
WW01 | Invention patent application withdrawn after publication |