CN104732494A - Tissue culturing monitoring method and system based on image mist elimination - Google Patents

Abstract

The invention relates to a tissue culturing monitoring method based on image mist elimination. The tissue culturing monitoring method includes the steps of obtaining real-time monitoring image information and real-time environment information of a tissue culturing monitoring point in a tissue culturing room; preprocessing the real-time monitoring image information to obtain a preprocessed image, and then obtaining a dark channel image of the preprocessed image; carrying out bilateral filtering on the dark channel image, calculating an atmospheric optical value A according to the bilaterally-filtered dark channel image, and calculating a transmittance predicted value according to the real-time monitoring image information; obtaining a mist-eliminated monitoring image according to a mist image forming model; sending the real-time monitoring image information, the real-time environment information and the mist-eliminated monitoring image to a user. The invention further discloses a tissue culturing monitoring system based on image mist elimination. By means of the tissue culturing monitoring method and system, automatic real-time collecting is achieved in the whole process, and the labor cost is greatly saved; preprocessing and mist elimination processing are carried out on obtained tissue culturing real-time monitoring information at the server side, and therefore the influences of mist on a tissue culturing monitoring picture are reduced as far as possible.

Description

Tissue culture monitoring method and system based on image defogging

Technical Field

The invention relates to the technical field of video monitoring image processing, in particular to a tissue culture monitoring method and system based on image defogging.

Background

At present, due to the scattering effect of a large number of suspended particles, atmospheric visibility is reduced, the color and contrast of an outdoor image are degraded, extraction of information in the image is influenced, outdoor definition is reduced, and traffic accidents are frequent. Therefore, image defogging has become an important subject of image processing and computer vision field research, and is one of the problems to be solved urgently, and removing fog in videos and improving image quality are key technologies for improving the value of video monitoring systems. With the continuous development of the technology, the defogging of the scenery images in foggy weather becomes possible, and the definition and the reality of the defogged images are greatly improved.

Image defogging is an important content in image processing and computer vision field research, and the main application fields of the image defogging are video monitoring, terrain surveying, automatic driving and target tracking. It can be seen that the application of the image defogging is mostly to the treatment of outdoor foggy images, while the tissue culture is in closed growth, and a layer of water fog is often attached to the wall of the tissue culture bottle, so that the real growth condition of plants in the closed tissue culture bottle can not be clearly seen when the plants are checked through remote monitoring. At present, rely on the manual work to organize tissue culture plant diseases and insect pests and inspect much, just so lead to unable in time discovering the plant diseases and insect pests to manual inspection needs to consume a large amount of manpowers, and group banks up the growth under sterile environment with earth, and people themselves can bring the bacterium into, can increase the possibility that the tissue culture seedling infects the germ, and the manual management of being not convenient for moreover can't organize the tissue culture growth condition in a long-range and look over in real time, can consume very big human cost like this.

Disclosure of Invention

The invention aims to provide a tissue culture monitoring system which can automatically acquire remote tissue culture monitoring information in real time and save labor cost; and carrying out image defogging treatment on the acquired tissue culture monitoring information, so that the influence of fog on a tissue culture monitoring picture is reduced as much as possible.

In order to achieve the purpose, the invention adopts the following technical scheme: a tissue culture monitoring method based on image defogging comprises the following steps:

(1) acquiring real-time monitoring image information and real-time environment information of tissue culture monitoring points in a tissue culture room;

(2) preprocessing the real-time monitoring image information to obtain a preprocessed image, and then acquiring a dark channel map of the preprocessed image;

(3) carrying out bilateral filtering on the dark channel map, calculating an atmospheric light value A according to the dark channel map subjected to bilateral filtering, and calculating a transmittance pre-estimated value according to real-time monitoring image information;

(4) after obtaining the atmospheric light value A and the transmittance estimated value, obtaining a monitoring image after defogging according to a fog image forming model;

(5) and sending the real-time monitoring image information, the real-time environment information and the defogged monitoring image to a user.

The preprocessing of the real-time monitoring image information refers to performing image denoising and smoothing on each frame of the real-time acquired monitoring image, namely performing mean filtering on each frame of image and then performing histogram equalization processing.

The step of obtaining the dark channel map of the preprocessed image is to decompose the preprocessed image in an RGB space into a plurality of square windows, firstly calculate the minimum value of RGB components of each pixel in each square window, use the minimum value as the gray value of the dark channel map, and store the minimum value of RGB components of each pixel in each square window into the gray map with the same size as the original real-time monitoring image, wherein the gray map is the dark channel map;

the dark channel expression is:

<math> <mrow> <msup> <mi>J</mi> <mi>dark</mi> </msup> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <mo>=</mo> <munder> <mi>min</mi> <mrow> <mi>y</mi> <mo>&Element;</mo> <mi>&Omega;</mi> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> </mrow> </munder> <mrow> <mo>(</mo> <munder> <mi>min</mi> <mrow> <mi>c</mi> <mo>&Element;</mo> <mo>{</mo> <mi>R</mi> <mo>,</mo> <mi>G</mi> <mo>,</mo> <mi>B</mi> <mo>}</mo> </mrow> </munder> <msup> <mi>J</mi> <mi>c</mi> </msup> <mrow> <mo>(</mo> <mi>y</mi> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>,</mo> </mrow> </math>

wherein J^dark(x) Gray value representing the dark channel map, J^cRepresents each channel of the real-time monitored image, Ω (x) represents a square window centered on pixel x; y represents each pixel in the omega (x) window.

The bilateral filtering is to calculate the dark channel image through a bilateral filtering template, and the bilateral filtering template is obtained by performing point multiplication operation on a Gaussian template and a Gaussian function template.

The calculation method of the atmospheric light value A comprises the following steps: firstly, taking the first 0.1% of pixel points from a dark channel map according to the gray value; and (3) the pixel points are corresponding to the original real-time monitoring image, the pixel point with the highest brightness is searched in the original real-time monitoring image, and the highest brightness is used as an atmospheric light value A.

The calculation formula of the transmittance estimated value is as follows:

<math> <mrow> <mover> <mi>t</mi> <mo>~</mo> </mover> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <mo>=</mo> <mn>1</mn> <mo>-</mo> <mi>&omega;</mi> <munder> <mi>min</mi> <mrow> <mi>y</mi> <mo>&Element;</mo> <mi>&Omega;</mi> </mrow> </munder> <mrow> <mo>(</mo> <munder> <mi>min</mi> <mi>c</mi> </munder> <mfrac> <mrow> <msup> <mi>I</mi> <mi>c</mi> </msup> <mrow> <mo>(</mo> <mi>y</mi> <mo>)</mo> </mrow> </mrow> <msup> <mi>A</mi> <mi>c</mi> </msup> </mfrac> <mo>)</mo> </mrow> </mrow> </math>

wherein,omega is a constant set by a user and is in a range of 0.6 to 0.7 for the transmittance estimated value,^cdenotes R, G, B three channels, I^cRepresenting each channel, A, in the original real-time monitoring image^cRepresenting R, G, B the atmospheric light values for three channels, Ω (x) represents a square window centered on pixel x, and y represents each pixel in the Ω (x) window.

The fog diagram forming model is as follows:

$I\left(x\right)=J\left(x\right)\tilde{t}\left(x\right)+A(1-\tilde{t}\left(x\right))$

wherein, I (x) is the original real-time monitoring image, J (x) is the monitoring image after defogging, A is the atmospheric light value,is a pre-estimated value of transmittance, which is a constant,the acquisition process is as follows:

first, i (x) ═ j (x) t (x) + a (1-t (x)) is transformed intoAs described above, superscript C denotes R, G, B three channels, and first assumes that the transmittance t (x) is constant within each window, defining it asAnd the value of A has been given, thenTwo minimum value operations are calculated on two sides to obtain <math> <mrow> <munder> <mi>min</mi> <mrow> <mi>y</mi> <mo>&Element;</mo> <mi>&Omega;</mi> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> </mrow> </munder> <mrow> <mo>(</mo> <munder> <mi>min</mi> <mi>c</mi> </munder> <mrow> <mo>(</mo> <mfrac> <mrow> <msup> <mi>I</mi> <mi>c</mi> </msup> <mrow> <mo>(</mo> <mi>y</mi> <mo>)</mo> </mrow> </mrow> <msup> <mi>A</mi> <mi>c</mi> </msup> </mfrac> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>=</mo> <mover> <mi>t</mi> <mo>~</mo> </mover> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <munder> <mi>min</mi> <mrow> <mi>y</mi> <mo>&Element;</mo> <mi>&Omega;</mi> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> </mrow> </munder> <mrow> <mo>(</mo> <munder> <mi>min</mi> <mi>c</mi> </munder> <mrow> <mo>(</mo> <mfrac> <mrow> <msup> <mi>J</mi> <mi>c</mi> </msup> <mrow> <mo>(</mo> <mi>y</mi> <mo>)</mo> </mrow> </mrow> <msup> <mi>A</mi> <mi>c</mi> </msup> </mfrac> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>+</mo> <mn>1</mn> <mo>-</mo> <mover> <mi>t</mi> <mo>~</mo> </mover> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <mo>,</mo> </mrow> </math> According to the dark channel prior theory <math> <mrow> <msup> <mi>J</mi> <mi>dark</mi> </msup> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <mo>=</mo> <munder> <mi>min</mi> <mrow> <mi>y</mi> <mo>&Element;</mo> <mi>&Omega;</mi> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> </mrow> </munder> <mrow> <mo>(</mo> <munder> <mi>min</mi> <mi>c</mi> </munder> <mrow> <mo>(</mo> <msup> <mi>J</mi> <mi>c</mi> </msup> <mrow> <mo>(</mo> <mi>y</mi> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>=</mo> <mn>0</mn> <mo>,</mo> </mrow> </math> It can be known that <math> <mrow> <munder> <mi>min</mi> <mrow> <mi>y</mi> <mo>&Element;</mo> <mi>&Omega;</mi> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> </mrow> </munder> <mrow> <mo>(</mo> <munder> <mi>min</mi> <mi>c</mi> </munder> <mrow> <mo>(</mo> <mfrac> <mrow> <msup> <mi>J</mi> <mi>c</mi> </msup> <mrow> <mo>(</mo> <mi>y</mi> <mo>)</mo> </mrow> </mrow> <msup> <mi>A</mi> <mi>c</mi> </msup> </mfrac> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>=</mo> <mn>0</mn> <mo>,</mo> </mrow> </math> Thereby obtaining a predicted value of transmissivityIntroduction of a group in [0,1 ]]By a correction factor to obtain a final estimated value of transmittanceWhere ω is used to adjust the degree of defogging, where C represents R, G, B three channels, I^cRepresenting each channel in the image to be dehazed, A^cRepresenting the atmospheric light value corresponding to the three channels, wherein omega (X) represents a window taking a pixel X as a center, and y represents each pixel point in the omega (X) window;

can obtain the monitor after defoggingA control image, represented as:

wherein, I (x) is the original real-time monitoring image, J (x) is the monitoring image after defogging, A is the atmospheric light value,for the transmittance estimate, it is constant, representing the transmittance in each omega (x) window, t₀To representIs measured.

And after obtaining the dark channel image, carrying out minimum value filtering on the dark channel image, wherein the filtering radius is determined by the size of a square window, and the calculation formula of the filtering radius is as follows:

WindowSize＝2*Radius+1。

wherein, WindowSize is the side length of the square window, and Radius is the filtering Radius.

The invention also provides a tissue culture monitoring system based on image defogging, which comprises:

the acquisition terminal acquires real-time monitoring image information of a tissue culture monitoring point through a network camera arranged in the tissue culture room, acquires real-time environment information through a temperature sensor and an illumination sensor, and transmits the acquired information to the server through a wired/wireless network;

the server receives the acquired real-time monitoring image information and real-time environment information, performs preprocessing and defogging processing on each frame of the image, and sends the real-time monitoring image information, the real-time environment information and the defogged monitoring image information to the client through a wired/wireless network;

and the client is used for receiving the real-time image and the defogged monitoring image sent by the server and displaying the real-time image and the defogged monitoring image to a user through the mobile phone client, the portable PC terminal and the PC terminal.

According to the technical scheme, the remote tissue culture image information and the real-time environment information are acquired in real time through the network camera, the information is transmitted to the server end through the network and then transmitted to the client end through the server end, the whole process is automatically acquired in real time, manual monitoring is not needed, and the labor cost is greatly saved; the server side carries out pretreatment and defogging treatment on the acquired tissue culture real-time monitoring information, so that the influence of fog on a tissue culture monitoring picture is reduced as much as possible, and a clear image about tissue culture real-time growth and pathogen infection conditions is provided for managers so as to make corresponding treatment measures. In a word, the invention carries out defogging and clearing treatment on the collected video data, so that managers can observe the tissue culture growth condition more clearly.

Drawings

FIG. 1 is a flow chart of a method of the present invention;

fig. 2 is a block diagram of the system architecture of the present invention.

Detailed Description

As shown in fig. 1, a tissue culture monitoring method based on image defogging comprises the following steps: (1) acquiring real-time monitoring image information and real-time environment information of tissue culture monitoring points in a tissue culture room; (2) preprocessing the real-time monitoring image information to obtain a preprocessed image, and then acquiring a dark channel map of the preprocessed image; (3) carrying out bilateral filtering on the dark channel map, calculating an atmospheric light value A according to the dark channel map subjected to bilateral filtering, and calculating a transmittance pre-estimated value according to real-time monitoring image information; (4) after obtaining the atmospheric light value A and the transmittance estimated value, obtaining a monitoring image after defogging according to a fog image forming model; (5) and sending the real-time monitoring image information, the real-time environment information and the defogged monitoring image to a user.

The preprocessing of the real-time monitoring image information refers to performing image denoising and smoothing on each frame of the real-time acquired monitoring image, namely performing mean filtering on each frame of image and then performing histogram equalization processing.

The dark channel map of the pre-processed image is obtained by decomposing the pre-processed image in an RGB space into a plurality of square windows, generally taking 3-3 windows, firstly calculating the minimum value of RGB components of each pixel in each square window, taking the minimum value as the gray value of the dark channel map, and storing the minimum value of RGB components of each pixel in each square window into the gray map with the same size as the original real-time monitoring image, wherein the gray map is the dark channel map;

the dark channel expression is:

<math> <mrow> <msup> <mi>J</mi> <mi>dark</mi> </msup> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <mo>=</mo> <munder> <mi>min</mi> <mrow> <mi>y</mi> <mo>&Element;</mo> <mi>&Omega;</mi> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> </mrow> </munder> <mrow> <mo>(</mo> <munder> <mi>min</mi> <mrow> <mi>c</mi> <mo>&Element;</mo> <mo>{</mo> <mi>R</mi> <mo>,</mo> <mi>G</mi> <mo>,</mo> <mi>B</mi> <mo>}</mo> </mrow> </munder> <msup> <mi>J</mi> <mi>c</mi> </msup> <mrow> <mo>(</mo> <mi>y</mi> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>,</mo> </mrow> </math>

wherein J^dark(x) Gray value representing the dark channel map, J^cEach channel representing a real-time monitored image, and Ω (x) represents one centered on pixel xA square window; y represents each pixel in the omega (x) window.

And after obtaining the dark channel image, carrying out minimum value filtering on the dark channel image, wherein the filtering radius is determined by the size of a square window, and the calculation formula of the filtering radius is as follows:

WindowSize＝2*Radius+1。

wherein, WindowSize is the side length of the square window, and Radius is the filtering Radius. And performing minimum value filtering to mainly filter discrete points at the edge of the dark channel map so as to obtain the dark channel map with smooth edge.

The bilateral filtering is to calculate the dark channel image through a bilateral filtering template, and the bilateral filtering template is obtained by performing point multiplication operation on a Gaussian template and a Gaussian function template. The bilateral filtering is used for changing a zigzag broken line edge of a dark channel graph into a smooth curve edge, the Gaussian function template is generated by taking a difference value of gray levels as a Gaussian function coefficient, and the Gaussian template is a global template, so that the generation is only needed once, and the Gaussian function template needs to be calculated once for each pixel.

The calculation method of the atmospheric light value A comprises the following steps: firstly, taking the first 0.1% of pixel points from a dark channel map according to the gray value; and (3) the pixel points are corresponding to the original real-time monitoring image, the pixel point with the highest brightness is searched in the original real-time monitoring image, and the highest brightness is used as an atmospheric light value A.

The calculation formula of the transmittance estimated value is as follows:

<math> <mrow> <mover> <mi>t</mi> <mo>~</mo> </mover> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <mo>=</mo> <mn>1</mn> <mo>-</mo> <mi>&omega;</mi> <munder> <mi>min</mi> <mrow> <mi>y</mi> <mo>&Element;</mo> <mi>&Omega;</mi> </mrow> </munder> <mrow> <mo>(</mo> <munder> <mi>min</mi> <mi>c</mi> </munder> <mfrac> <mrow> <msup> <mi>I</mi> <mi>c</mi> </msup> <mrow> <mo>(</mo> <mi>y</mi> <mo>)</mo> </mrow> </mrow> <msup> <mi>A</mi> <mi>c</mi> </msup> </mfrac> <mo>)</mo> </mrow> </mrow> </math>

wherein,omega is a constant set by a user and is in a range of 0.6 to 0.7 for the transmittance estimated value,^cdenotes R, G, B three channels, I^cRepresenting each channel, A, in the original real-time monitoring image^cRepresenting R, G, B the atmospheric light values for three channels, Ω (x) represents a square window centered on pixel x, and y represents each pixel in the Ω (x) window.

The fog diagram forming model is as follows:

$I\left(x\right)=J\left(x\right)\tilde{t}\left(x\right)+A(1-\tilde{t}\left(x\right))$

wherein, I (x) is the original real-time monitoring image, J (x) is the monitoring image after defogging, A is the atmospheric light value,is a pre-estimated value of transmittance, which is a constant,the acquisition process is as follows:

first, i (x) ═ j (x) t (x) + a (1-t (x)) is transformed intoAs described above, superscript C denotes R, G, B three channels, and first assumes that the transmittance t (x) is constant within each window, defining it asAnd the value of A has been given, thenTwo minimum value operations are calculated on two sides to obtain <math> <mrow> <munder> <mi>min</mi> <mrow> <mi>y</mi> <mo>&Element;</mo> <mi>&Omega;</mi> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> </mrow> </munder> <mrow> <mo>(</mo> <munder> <mi>min</mi> <mi>c</mi> </munder> <mrow> <mo>(</mo> <mfrac> <mrow> <msup> <mi>I</mi> <mi>c</mi> </msup> <mrow> <mo>(</mo> <mi>y</mi> <mo>)</mo> </mrow> </mrow> <msup> <mi>A</mi> <mi>c</mi> </msup> </mfrac> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>=</mo> <mover> <mi>t</mi> <mo>~</mo> </mover> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <munder> <mi>min</mi> <mrow> <mi>y</mi> <mo>&Element;</mo> <mi>&Omega;</mi> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> </mrow> </munder> <mrow> <mo>(</mo> <munder> <mi>min</mi> <mi>c</mi> </munder> <mrow> <mo>(</mo> <mfrac> <mrow> <msup> <mi>J</mi> <mi>c</mi> </msup> <mrow> <mo>(</mo> <mi>y</mi> <mo>)</mo> </mrow> </mrow> <msup> <mi>A</mi> <mi>c</mi> </msup> </mfrac> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>+</mo> <mn>1</mn> <mo>-</mo> <mover> <mi>t</mi> <mo>~</mo> </mover> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <mo>,</mo> </mrow> </math> According to the dark channel prior theory <math> <mrow> <msup> <mi>J</mi> <mi>dark</mi> </msup> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <mo>=</mo> <munder> <mi>min</mi> <mrow> <mi>y</mi> <mo>&Element;</mo> <mi>&Omega;</mi> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> </mrow> </munder> <mrow> <mo>(</mo> <munder> <mi>min</mi> <mi>c</mi> </munder> <mrow> <mo>(</mo> <msup> <mi>J</mi> <mi>c</mi> </msup> <mrow> <mo>(</mo> <mi>y</mi> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>=</mo> <mn>0</mn> <mo>,</mo> </mrow> </math> It can be known that <math> <mrow> <munder> <mi>min</mi> <mrow> <mi>y</mi> <mo>&Element;</mo> <mi>&Omega;</mi> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> </mrow> </munder> <mrow> <mo>(</mo> <munder> <mi>min</mi> <mi>c</mi> </munder> <mrow> <mo>(</mo> <mfrac> <mrow> <msup> <mi>J</mi> <mi>c</mi> </msup> <mrow> <mo>(</mo> <mi>y</mi> <mo>)</mo> </mrow> </mrow> <msup> <mi>A</mi> <mi>c</mi> </msup> </mfrac> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>=</mo> <mn>0</mn> <mo>,</mo> </mrow> </math> Thereby obtaining a predicted value of transmissivityIntroduction of a group in [0,1 ]]By a correction factor to obtain a final estimated value of transmittanceWhere ω is used to adjust the degree of defogging, where C represents R, G, B three channels, I^cRepresenting each channel in the image to be dehazed, A^cRepresenting the atmospheric light value corresponding to the three channels, wherein omega (X) represents a window taking a pixel X as a center, and y represents each pixel point in the omega (X) window;

the monitoring image after defogging can be obtained and is represented as:

wherein, I (x) is the original real-time monitoring image, J (x) is the monitoring image after defogging, A is the atmospheric light value,for the transmittance estimate, it is constant, representing the transmittance in each omega (x) window, t₀To representIs measured.

As shown in fig. 2, the present system includes: the acquisition terminal acquires real-time monitoring image information of a tissue culture monitoring point through a network camera arranged in the tissue culture room, acquires real-time environment information through a temperature sensor and an illumination sensor, and transmits the acquired information to the server through a wired/wireless network; the server receives the acquired real-time monitoring image information and real-time environment information, performs preprocessing and defogging processing on each frame of the image, and sends the real-time monitoring image information, the real-time environment information and the defogged monitoring image information to the client through a wired/wireless network; and the client is used for receiving the real-time image and the defogged monitoring image sent by the server and displaying the real-time image and the defogged monitoring image to a user through the mobile phone client, the portable PC terminal and the PC terminal.

In conclusion, the remote tissue culture image information and the real-time environment information are acquired in real time through the network camera, transmitted to the server end through the network and transmitted to the client end through the server end, automatically acquired in real time in the whole process, manual monitoring is not needed, and the labor cost is greatly saved; the server side carries out pretreatment and defogging treatment on the acquired tissue culture real-time monitoring information, so that the influence of fog on a tissue culture monitoring picture is reduced as much as possible, and a clear image about tissue culture real-time growth and pathogen infection conditions is provided for managers so as to make corresponding treatment measures.

Claims (9)

1. A tissue culture monitoring method based on image defogging comprises the following steps:

(1) acquiring real-time monitoring image information and real-time environment information of tissue culture monitoring points in a tissue culture room;

(2) preprocessing the real-time monitoring image information to obtain a preprocessed image, and then acquiring a dark channel map of the preprocessed image;

(3) carrying out bilateral filtering on the dark channel map, calculating an atmospheric light value A according to the dark channel map subjected to bilateral filtering, and calculating a transmittance pre-estimated value according to real-time monitoring image information;

(4) after obtaining the atmospheric light value A and the transmittance estimated value, obtaining a monitoring image after defogging according to a fog image forming model;

(5) and sending the real-time monitoring image information, the real-time environment information and the defogged monitoring image to a user.

2. The image defogging based tissue culture monitoring method according to claim 1, wherein: the preprocessing of the real-time monitoring image information refers to performing image denoising and smoothing on each frame of the real-time acquired monitoring image, namely performing mean filtering on each frame of image and then performing histogram equalization processing.

3. The image defogging based tissue culture monitoring method according to claim 1, wherein: the step of obtaining the dark channel map of the preprocessed image is to decompose the preprocessed image in an RGB space into a plurality of square windows, firstly calculate the minimum value of RGB components of each pixel in each square window, use the minimum value as the gray value of the dark channel map, and store the minimum value of RGB components of each pixel in each square window into the gray map with the same size as the original real-time monitoring image, wherein the gray map is the dark channel map;

the dark channel expression is:

<math> <mrow> <msup> <mi>J</mi> <mi>dark</mi> </msup> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <mo>=</mo> <munder> <mi>min</mi> <mrow> <mi>y</mi> <mo>&Element;</mo> <mi>&Omega;</mi> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> </mrow> </munder> <mrow> <mo>(</mo> <munder> <mi>min</mi> <mrow> <mi>c</mi> <mo>&Element;</mo> <mo>{</mo> <mi>R</mi> <mo>,</mo> <mi>G</mi> <mo>,</mo> <mi>G</mi> <mo>}</mo> </mrow> </munder> <msup> <mi>J</mi> <mi>c</mi> </msup> <mrow> <mo>(</mo> <mi>y</mi> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>,</mo> </mrow> </math>

wherein J^dark(x) Gray value representing the dark channel map, J^cRepresents each channel of the real-time monitored image, Ω (x) represents a square window centered on pixel x; y represents each pixel in the omega (x) window.

4. The image defogging based tissue culture monitoring method according to claim 1, wherein: the bilateral filtering is to calculate the dark channel image through a bilateral filtering template, and the bilateral filtering template is obtained by performing point multiplication operation on a Gaussian template and a Gaussian function template.

5. The image defogging based tissue culture monitoring method according to claim 1, wherein: the calculation method of the atmospheric light value A comprises the following steps: firstly, taking the first 0.1% of pixel points from a dark channel map according to the gray value; and (3) the pixel points are corresponding to the original real-time monitoring image, the pixel point with the highest brightness is searched in the original real-time monitoring image, and the highest brightness is used as an atmospheric light value A.

6. The image defogging based tissue culture monitoring method according to claim 1, wherein: the calculation formula of the transmittance estimated value is as follows:

<math> <mrow> <mover> <mi>t</mi> <mo>~</mo> </mover> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <mo>=</mo> <mn>1</mn> <mo>-</mo> <mi>&omega;</mi> <munder> <mi>min</mi> <mrow> <mi>y</mi> <mo>&Element;</mo> <mi>&Omega;</mi> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> </mrow> </munder> <mrow> <mo>(</mo> <munder> <mi>min</mi> <mi>c</mi> </munder> <mfrac> <mrow> <msup> <mi>I</mi> <mi>c</mi> </msup> <mrow> <mo>(</mo> <mi>y</mi> <mo>)</mo> </mrow> </mrow> <msup> <mi>A</mi> <mi>c</mi> </msup> </mfrac> <mo>)</mo> </mrow> </mrow> </math>

wherein,omega is a constant set by a user and is in a range of 0.6 to 0.7 for the transmittance estimated value,^cdenotes R, G, B three channels, I^cRepresenting each channel, A, in the original real-time monitoring image^cRepresenting R, G, B the atmospheric light values for three channels, Ω (x) represents a square window centered on pixel x, and y represents each pixel in the Ω (x) window.

7. The image defogging based tissue culture monitoring method according to claim 1, wherein: the fog diagram forming model is as follows:

$I\left(x\right)=J\left(x\right)\tilde{t}\left(x\right)+A(1-\tilde{t}\left(x\right))$

wherein, I (x) is the original real-time monitoring image, J (x) is the monitoring image after defogging, A is the atmospheric light value,is a pre-estimated value of transmittance, which is a constant,the acquisition process is as follows:

first, i (x) ═ j (x) t (x) + a (1-t (x)) is transformed intoAs described above, superscript C denotes R, G, B three channels, and first assumes that the transmittance t (x) is constant within each window, defining it asAnd the value of A has been given, thenTwo minimum value operations are calculated on two sides to obtain <math> <mrow> <munder> <mi>mni</mi> <mrow> <mi>y</mi> <mo>&Element;</mo> <mi>&Omega;</mi> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> </mrow> </munder> <mrow> <mo>(</mo> <munder> <mi>min</mi> <mi>c</mi> </munder> <mrow> <mo>(</mo> <mfrac> <mrow> <msup> <mi>I</mi> <mi>c</mi> </msup> <mrow> <mo>(</mo> <mi>y</mi> <mo>)</mo> </mrow> </mrow> <msup> <mi>A</mi> <mi>c</mi> </msup> </mfrac> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>=</mo> <mover> <mi>t</mi> <mo>~</mo> </mover> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <munder> <mi>min</mi> <mrow> <mi>y</mi> <mo>&Element;</mo> <mi>&Omega;</mi> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> </mrow> </munder> <mrow> <mo>(</mo> <munder> <mi>min</mi> <mi>c</mi> </munder> <mrow> <mo>(</mo> <mfrac> <mrow> <msup> <mi>J</mi> <mi>c</mi> </msup> <mrow> <mo>(</mo> <mi>y</mi> <mo>)</mo> </mrow> </mrow> <msup> <mi>A</mi> <mi>c</mi> </msup> </mfrac> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>+</mo> <mn>1</mn> <mo>-</mo> <mover> <mi>t</mi> <mo>~</mo> </mover> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <mo>,</mo> </mrow> </math> According to the dark channel prior theory <math> <mrow> <msup> <mi>J</mi> <mi>dark</mi> </msup> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <mo>=</mo> <munder> <mi>min</mi> <mrow> <mi>y</mi> <mo>&Element;</mo> <mi>&Omega;</mi> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> </mrow> </munder> <mrow> <mo>(</mo> <munder> <mi>min</mi> <mi>c</mi> </munder> <mrow> <mo>(</mo> <msup> <mi>J</mi> <mi>c</mi> </msup> <mrow> <mo>(</mo> <mi>y</mi> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>=</mo> <mn>0</mn> <mo>,</mo> </mrow> </math> It can be known that <math> <mrow> <munder> <mi>min</mi> <mrow> <mi>y</mi> <mo>&Element;</mo> <mi>&Omega;</mi> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> </mrow> </munder> <mrow> <mo>(</mo> <munder> <mi>min</mi> <mi>c</mi> </munder> <mrow> <mo>(</mo> <mfrac> <mrow> <msup> <mi>J</mi> <mi>c</mi> </msup> <mrow> <mo>(</mo> <mi>y</mi> <mo>)</mo> </mrow> </mrow> <msup> <mi>A</mi> <mi>c</mi> </msup> </mfrac> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>=</mo> <mn>0</mn> <mo>,</mo> </mrow> </math> Thereby obtaining a predicted value of transmissivityIntroduction of a group in [0,1 ]]By a correction factor to obtain a final estimated value of transmittanceWhere ω is used to adjust the degree of defogging, where C represents R, G, B three channels, I^cRepresenting each channel in the image to be dehazed, A^cRepresenting the atmospheric light value corresponding to the three channels, wherein omega (X) represents a window taking a pixel X as a center, and y represents each pixel point in the omega (X) window;

the monitoring image after defogging can be obtained and is represented as:

wherein, I (x) is the original real-time monitoring image, J (x) is the monitoring image after defogging, A is the atmospheric light value,for the transmittance estimate, it is constant, representing the transmittance in each omega (x) window, t₀To representIs measured.

8. The image defogging based tissue culture monitoring method according to claim 3, wherein: and after obtaining the dark channel image, carrying out minimum value filtering on the dark channel image, wherein the filtering radius is determined by the size of a square window, and the calculation formula of the filtering radius is as follows:

WindowSize＝2*Radius+1。

wherein, WindowSize is the side length of the square window, and Radius is the filtering Radius.

9. The utility model provides a group banks up monitored control system with earth based on image defogging which characterized in that: the method comprises the following steps:

the acquisition terminal acquires real-time monitoring image information of a tissue culture monitoring point through a network camera arranged in the tissue culture room, acquires real-time environment information through a temperature sensor and an illumination sensor, and transmits the acquired information to the server through a wired/wireless network;

the server receives the acquired real-time monitoring image information and real-time environment information, performs preprocessing and defogging processing on each frame of the image, and sends the real-time monitoring image information, the real-time environment information and the defogged monitoring image information to the client through a wired/wireless network;

and the client is used for receiving the real-time image and the defogged monitoring image sent by the server and displaying the real-time image and the defogged monitoring image to a user through the mobile phone client, the portable PC terminal and the PC terminal.