CN105319382A - Blood cell counting method and application - Google Patents

Abstract

The invention discloses a blood cell counting method and application. The blood cell counting method comprises an automatic sample conveying step, a synchronous control step, a blood cell recognition step and a result display step, wherein the blood cell recognition step is used for recognizing blood cells in a blood sample; the synchronous control step is used for controlling the synchronization of the automatic sample conveying step and the blood cell recognition step; the automatic sample conveying step is used for conveying blood cell samples; the result display step is used for outputting blood cell sample and/or data results, and displaying blood cell counting results. By adoption of the blood cell counting method, the impact of manual operation on detection results can be eliminated. The blood cell counting method and the application not only have important academical values, but also is wide in prospect, and can create considerable social and economic benefits.

Description

Blood cell counting method and application

Technical Field

The invention relates to the technical field of life science, in particular to a blood cell counting method and application thereof.

Background

A blood cell counter is used in a blood station or a hospital for detecting various parameters of red blood cells, platelets and white blood cells in whole blood of a blood donor or a patient, and a detection result is used for judging whether the platelets can be collected or not in the blood station; in hospitals, the test results are an important basis for the clinician to decide on the treatment regimen. Whether the blood cell counter sucks enough blood samples without air or air bubbles is an important basis for determining the accuracy of the detection result. In the operation process, especially when the sample is more, when new personnel of going on duty use, can lead to the cytometer sampling needle not in place, inhale the appearance not enough because sample test tube locating place is improper, or inhale the air, influence the accuracy of blood testing result.

Along with the annual promotion of national information construction and the increasing maturity of computer technology, two different subjects of computer and medicine are also permeated, which indicates that the trend of the biomedicine towards diversification, intellectualization and automation under the background of a new era. The prosperity of a country is strong and the support of medicine can not be opened, which is an important guarantee for the life quality of people and is an embodiment of the comprehensive strength of a country.

The development of medicine has always been accompanied by the development of basic disciplines, and particularly in the 19 th century, with the rapid rise of natural science disciplines at that time, medicine has also shifted from traditional medicine to the first to represent generations of medicine. Cytology arose since the beginning of the 19 th century, cytopathology was suggested by the German pathologist, the Fei-shi, who advocated the possible pathological phenomena by angular analysis of cellular abnormalities, morphology, etc. The basic principles of his teachings include: the cell is from a cell; the body is the sum of the cells; diseases can be described by cytopathology, etc. Through development and innovation for many years and progress of various visualization technologies, medical images become important carriers in clinical medicine, and medical images of cells are widely regarded and agreed in clinical analysis and play an important role in vast medical research institutions or hospitals.

A technique of processing an image using a computer image technique is called a digital image processing technique. The digital image processing technology can well solve the problems of low speed, low efficiency and difficult transmission in the execution of a plurality of traditional image processing methods, originates from the digital transmission experiment used in the early period, develops for decades and becomes a comprehensive subject at present. Compared with the general digital image digital technology, the difficulty and the requirement of medical image processing are higher compared with the common image processing.

First, medical images are often human tissues, so that the components are complex and the requirement on the accuracy of image processing is high.

Secondly, the medical image has non-uniformity of gray scale, and the trend of gray scale change does not have a fixed change pattern in the same tissue, which further increases the difficulty and potential of medical image processing.

Because the digital image has high use value in medical cross application, especially in medical image slice application with large cell number and extremely high requirement on diagnosis error, such as red blood cell identification in blood, cell identification of canceration in each organ and the like. The method has a very wide prospect, if the method can be realized by means of technologies such as graphic image processing, mode recognition and the like, the medical staff can be liberated from complicated slice analysis, the operation efficiency of hospital institutions can be greatly improved, the past method of simply depending on manual recognition is converted into the realization of algorithm intelligence, and errors in human eye judgment are reduced.

Disclosure of Invention

The invention aims to solve the technical problem of providing a blood cell counting method, which can automatically feed samples, fix the height and fix the distance according to the position of equipment, automatically start a blood cell counting instrument to suck and detect samples after reaching a standard sample sucking position and does not need excessive manual intervention operation. The blood cell counting method of the invention aims at the defect of manual blood cell counting identification in the prior art, and utilizes the image identification technology to combine with the characteristics of microscopic images to judge and count the number of cells in blood. Through a series of pretreatment to the sample picture, further then adopt methods such as mark processing and statistics correction, start from two aspects of preliminary treatment and judgement overlapping cell, realize finally utilizing the quick count of erythrocyte in the blood.

In order to solve the above technical problems, the present invention provides a blood cell counting method, comprising:

automatically feeding a sample;

a synchronous control step;

a blood cell recognition step;

and displaying the result.

The blood cell identification step of identifying blood cells in the blood sample;

the synchronous control step is used for controlling the automatic sample feeding device and the blood cell recognition device to be synchronous;

the automatic sample feeding step is used for conveying the blood cell sample;

and the result display step is used for outputting the blood cell sample and/or data result and displaying the blood cell counting result.

The blood cell identification step adopts the following identification method:

(1) reading a cell image, and converting the image into an HIS space;

(2) histogram equalization;

(3) smoothing, segmenting and edge correcting the image;

(4) extracting characteristics;

(5) cells are identified.

Reading in a cell image, and converting the image into an HIS space in the step (1) as follows:

for any RGB value in the interval [0, 1], the conversion formula of H, S, I component in HSI model corresponding to the value is:

$I=\frac{1}{3}(R+G+B)$

$S=1-\frac{3}{(R+G+B)}\[min(R,G,B)\]$

$H=\arccos \left\{\frac{\frac{1}{2}\[(R-G)+(R-B)\]}{{\[{\left(R-G\right)}^2+\left(R-G\right)\left(G-B\right)\]}^{\frac{1}{2}}}\right\}.$

the synchronization control step: the PID control method adopts the PID control of the permanent magnet synchronous motor, and comprises the following steps:

(1) tracking differentiator

$\left\{\begin{array}{c}{e}_0={\mathrm{\ω}}_i-\widehat{\mathrm{\ω}}\\ \stackrel{\·}{\mathrm{\ω}}=-rfal(e_0,a_0,{\mathrm{\δ}}_0)\end{array}\right.$

In the above formula, the first and second carbon atoms are,is the system given speed, ω isThe tracking speed of (2);

(2) extended state observer

$\left\{\begin{array}{c}{e}_1=z_{21}-{\mathrm{\ω}}^{*}\\ {\stackrel{\·}{z}}_{21}=z_{22}-{\mathrm{\β}}_{1}fal(e_1,a_1,{\mathrm{\δ}}_1)+b_{0}u\left(t\right)\\ {\stackrel{\·}{z}}_{22}=z_{33}-{\mathrm{\β}}_{2}fal(e_1,a_1,{\mathrm{\δ}}_1)\\ {\stackrel{\·}{z}}_{33}=-{\mathrm{\β}}_{3}fal(e_1,a_2,{\mathrm{\δ}}_2)\end{array}\right.$

In the above formula, omega^*Is a feedback signal of the system, z₂₁Is omega^*Of the tracking signal z₂₂Is an observed value of an unknown disturbance of the system;

(3) nonlinear state error feedback control law

$\left\{\begin{array}{c}{e}_2=\widehat{\mathrm{\ω}}-z_{21}\\ u_1\left(t\right)={\mathrm{\λ}}_{1}fal(e_2,a_2,{\mathrm{\δ}}_2)+{\mathrm{\λ}}_{2}fal(e_3,a_3,{\mathrm{\δ}}_3)\end{array}\right.$

The control quantity of disturbance compensation is:

$u\left(t\right)=u_1\left(t\right)-\frac{z_{22}}{b}$

u (t) is the control quantity input to the current loop after disturbance compensation.

The synchronization control step further includes: the blood sample test tube is placed in a sample feeder test tube support of the blood cell counter, the sample feeder is pushed forward and synchronously lifted and pushed forward according to the design and the lifting distance by pressing a start key, the sample feeder reaches a specified position, an arm touches a detection start key, the sample is sucked, and after 15 seconds, the sample feeder automatically returns.

In order to solve the technical problem, the invention also provides the application of the blood cell counting method in any one of the preceding items in preparing a blood cell counter.

In order to solve the above technical problem, the present invention further provides a blood cell counter operating by the blood cell counting method according to any one of the preceding claims, comprising: blood cell recognition device, synchronous control device, automatic sample feeding device and result display device.

The blood cell recognition device is respectively connected with the result display device and the synchronous control device and is used for recognizing blood cells in the blood sample;

the synchronous control device is respectively connected with the automatic sample feeding device and the blood cell recognition device and is used for controlling the automatic sample feeding device and the blood cell recognition device to be synchronous;

the automatic sample feeding device is respectively connected with the automatic control device and the blood cell recognition device and is used for conveying blood cell samples;

and the result display device is used for displaying the result of blood cell counting and outputting the data result.

The blood cell recognition device adopts the following recognition method:

(1) reading a cell image, and converting the image into an HIS space;

(2) histogram equalization;

(3) smoothing, segmenting and edge correcting the image;

(4) extracting characteristics;

(5) identifying the cell;

the synchronization control device: the PID control method adopts the PID control of the permanent magnet synchronous motor, and comprises the following steps:

(1) tracking differentiator

$\left\{\begin{array}{c}{e}_0={\mathrm{\ω}}_i-\widehat{\mathrm{\ω}}\\ \stackrel{\·}{\mathrm{\ω}}=-rfal(e_0,a_0,{\mathrm{\δ}}_0)\end{array}\right.$

In the above formula, the first and second carbon atoms are,is the system given speed, ω isThe tracking speed of (2);

(2) extended state observer

$\left\{\begin{array}{c}{e}_1=z_{21}-{\mathrm{\ω}}^{*}\\ {\stackrel{\·}{z}}_{21}=z_{22}-{\mathrm{\β}}_{1}fal(e_1,a_1,{\mathrm{\δ}}_1)+b_{0}u\left(t\right)\\ {\stackrel{\·}{z}}_{22}=z_{33}-{\mathrm{\β}}_{2}fal(e_1,a_1,{\mathrm{\δ}}_1)\\ {\stackrel{\·}{z}}_{33}=-{\mathrm{\β}}_{3}fal(e_1,a_2,{\mathrm{\δ}}_2)\end{array}\right.$

In the above formula, omega^*Is a feedback signal of the system, z₂₁Is omega^*Of the tracking signal z₂₂Is an observed value of an unknown disturbance of the system;

(3) nonlinear state error feedback control law

$\left\{\begin{array}{c}{e}_2=\widehat{\mathrm{\ω}}-z_{21}\\ u_1\left(t\right)={\mathrm{\λ}}_{1}fal(e_2,a_2,{\mathrm{\δ}}_2)+{\mathrm{\λ}}_{2}fal(e_3,a_3,{\mathrm{\δ}}_3)\end{array}\right.$

The control quantity of disturbance compensation is:

$u\left(t\right)=u_1\left(t\right)-\frac{z_{22}}{b}$

u (t) is the control quantity input to the current loop after disturbance compensation.

The beneficial technical effects of the invention are as follows: the blood cell counting method can automatically send the blood sample to be detected to the accurate sample sucking position of the blood cell counting instrument according to the design requirement, ensure the sufficient sample blood volume to be sucked, ensure the accuracy of the detection result and eliminate the influence of manual operation on the detection result. The cell counting method has important academic value and wide prospect, and creates considerable social and economic benefits.

Drawings

FIG. 1 is a block diagram of a PID controller according to an embodiment of the invention;

FIG. 2 is a diagram of a basic structure of a fuzzy controller according to an embodiment of the present invention;

fig. 3 is a schematic diagram of specific tasks performed by the fuzzy PID controller according to the embodiment of the present invention.

Detailed Description

The following embodiments of the present invention will be described in detail with reference to the accompanying examples, so that how to apply the technical means to solve the technical problems and achieve the technical effects can be fully understood and implemented.

It should be noted that, in order to save the written space of the specification and avoid unnecessary repetition and waste, the embodiments and features of the embodiments in the present application may be combined with each other without conflict.

The full-automatic blood cell counter of the invention comprises: blood cell recognition device, synchronous control device, automatic sample feeding device, input/output device and result display device.

The blood cell recognition system device of the present invention adopts the following recognition method:

1. reading a cell image, and converting the image into an HIS space;

2. histogram equalization;

3. smoothing, segmenting and edge correcting the image;

4. extracting characteristics;

5. cells are identified.

The HSI model, proposed by munsell, is more in line with the human eye's habit and way of observing the colors of objects than the RGB model, which makes it intuitive and natural to manipulate colors. In the HIS model, H represents Hue (Hue), S represents Saturation (Saturation), and I represents brightness (Intensity). The basis for this model establishment is: (1) the I component of a pixel point in the HSI model is not related to the color information of the image at the point. (2) The pattern of H and S components and the perception of color by the human eye is similar to convention. Since the three components in the HSI model conform to the habit of people to perceive colors in nature, the color image algorithm based on human visual perception has better effect than other color models. In the HSI color model, the value of the H component is expressed as radian, and the variation range is between [0, 180] degrees; the S component is expressed as the radius length r of a circle, the smaller r is, the larger the value transformation of the H component is, namely, the worse the stability of the H component is; the I component, which reflects the grey scale of the color, is the height h of the cylinder from a more intuitive perspective. The H, S component at the floor and ceiling of the cylinder is meaningless when all dots are black in color at the bottom plane of the cylinder and white at the top.

Conversion of RGB space to HSI space.

Because the process of identifying and sensing the blood cells by human eyes has obvious color sensing components, the process of processing the image selects to map the image from the RGB space to the HSI space, which is more in line with the habit and the characteristic of human beings when identifying colored targets, and in the invention, the conversion process is as follows:

for any RGB value in the interval [0, 1], the conversion formula of H, S, I component in HSI model corresponding to the value is:

$I=\frac{1}{3}(R+G+B)$

$S=1-\frac{3}{(R+G+B)}\[min(R,G,B)\]$

$H=\arccos \left\{\frac{\frac{1}{2}\[(R-G)+(R-B)\]}{{\[{\left(R-G\right)}^2+\left(R-G\right)\left(G-B\right)\]}^{\frac{1}{2}}}\right\}$

and (5) histogram equalization.

The image is grayed first. The distribution of the gray values of the image after graying can be reflected by a histogram, which can be regarded as a gray level function of the image, and the abscissa and the ordinate of the histogram represent the gray values of the image and the occurrence frequency of the gray values respectively. The essence of the method is that the transformation function of histogram equalization of a histogram correction method based on an accumulative distribution function transformation method is as follows:

in the formula: ω is an integral variable, andis the cumulative integral function of r.

Here, the cumulative integration function is considered to be a function of r, and the function monotonically increases from 0 to 1, so this transformation function satisfies T (r) monotonically increasing in value at 0 ≦ r ≦ 1. In the range of r is more than or equal to 0 and less than or equal to 1, there are two conditions of T is more than or equal to 0 and less than or equal to 1 (r).

By taking the derivative of r in the formula, then

$\begin{array}{c}\frac{ds}{dr}=p_r\left(r\right)\\ p_r\left(s\right)={\[p_r\left(r\right)\·\frac{dr}{ds}\]}_{r=T^{-1}\left(s\right)}={\[p_r\left(r\right)\·\frac{1}{ds/dr}\]}_{r=T{\left(s\right)}^{-1}}\\ =\[p_r\left(r\right)\·\frac{1}{p_r\left(r\right)}\]=1\end{array}$

From the above results, it is understood that the probability densities within the domain of the transformed variable s are uniformly distributed. Therefore, the original image can be transformed into a new image having a gray value distribution corresponding to a uniform probability density by using the cumulative distribution function of r as a transformation function, and the dynamic range of the pixel value can be expanded. For discrete images, frequency is used instead of probability.

Image smoothing

The image smoothing is a processing method for improving the image quality by analyzing low-frequency components, a main body part or suppressing image noise and interfering high-frequency components, so that the image brightness is gradually changed, abrupt gradients are reduced. For example, gaussian noise, i.e., an n-dimensional distribution of the noise amplitude of each point of an image follows a gaussian distribution and is therefore also referred to as a normal distribution. For a random variable X, its probability density function is denoted as N (μ, σ 2), where μ, σ 2 are parameters of the distribution, the expectation and variance of the gaussian distribution, respectively. The formula of the probability density function is:

the method for smoothing an image used in the present invention includes: interpolation method, linear smoothing method, convolution method. In a specific case, the processing method for selecting the image smoothing is to be treated according to the difference of the image noise categories, so that an ideal effect can be achieved.

Image segmentation:

the invention adopts the maximum class difference method for segmentation. The maximum inter-class difference method, sometimes referred to as the Otsu algorithm, is usually the best algorithm for obtaining the threshold value in the melon score of the image of the person, and the algorithm is characterized in that: (1) easy calculation, high algorithm efficiency (2) and no influence of image brightness and comparability. (3) The performance is the most stable. Because of these advantages of the maximum class difference method, it is commonly utilized in digital image processing. The algorithm principle of the method is based on threshold segmentation of an image, each gray value characteristic which can divide the image into two parts of a background and a foreground in the image is searched, and when the inter-class variance between the background and the foreground is larger, the difference between the two parts forming the image is larger, and when the local foreground is mistaken for the background or the local background is mistaken for the foreground, the difference between the two parts of the image is reduced. The greatest benefit of using inter-class variance is that the probability of choosing the wrong threshold value at image segmentation is always minimized.

If the gray level of the gray image is L, the gray range is [0, L-1], and the optimal threshold of the image is calculated by using the maximum inter-class difference algorithm as follows:

t＝Max[w₀(t)·(u₀(t)-u)²+w₁(t)·(u₁(t)-u)²]

wherein the variables are: when the threshold value of the segmentation isAt t, w₀As background proportion, μ₀As background mean value, w₁In foreground proportion, μ₁Is the foreground mean value, and mu is the gray mean value of the whole image. And filling the interior of the image after the image is segmented.

Cell boundary correction

The picture after pretreatment needs to be corrected as necessary, and the elimination of the micro-pores in the cells needs to be considered. The main way of correction is by choosing the appropriate opening and closing operations, since the characteristics of the closing operations themselves mentioned in the preprocessing are suitable for making corrections to the edges of the image.

If the size and the times of selecting the reasonable template are selected, the interference of the blurring of the cell boundary to the later steps can be well reduced. The selection times of the opening operation and the closing operation and the selection of the template parameters are crucial to the cell boundary correction process, and the expected effect of correcting the boundary can be well achieved by selecting the proper times and the size of the template through analysis of the effect of the result obtained by experiments.

Feature extraction of cells

(1) Contour extraction

In identifying cells, it is important to extract the contour of the object. Since different regions in the binary image have different pixel values but the pixel values are the same for the same region. The contour algorithm is thus implemented as: sequentially scanning each point in the image, if 8 neighbors of the point in the image are black pixels, indicating that the point is an internal point, deleting the point, then continuing to scan the next point, and obtaining the extraction of the image contour when all the points in the image are scanned completely, wherein the obtained contour comprises a closed-connection area formed by independent cells and overlapped cells.

(2) Obtaining characteristics of cells

After the contour of the target is obtained, the main morphological features of the cells required by some subsequent algorithms can be calculated. These include circle center, perimeter, area, shape, etc. where we calculate these features to prepare the data in the previous stage for the final counting statistics, and the algorithm can actually solve some simple identification problems of independent cells by approximately calculating the number of connected regions. However, in the case of large-area adhesion, overlapped or overlapped cells cannot be judged, the counted numerical value is greatly smaller than the actual numerical value, but the judgment on the condition is not needed, and only errors of obtaining characteristic data are reduced as much as possible, so that the calculation in the future is facilitated.

Extracting features of connected regions

Here, the perimeter, area and particle of each connected component need to be extracted by query, and the obtained data can be submitted to the cell identification algorithm.

The area algorithm flow of the connected region is as follows:

(1) if the total number of the connected areas is 1-n connected areas are traversed;

(2) traversing each pair of connected regions row by row and column by column, and recording two pixels P (x) with pixel value not 0 in the same row₁，y₁)，P(x₂，y₂) The number of pixels in between is added, and the sum at this time is N.

(3) If there is no pixel in the ith row which is not 0, the process is terminated, and the value of N at this time is the required area.

After Fn communicating regions are scanned completely, counting the number of communicating regions with the shape factor larger than 0.85 and the area reaching the minimum area standard T of the set standard, and calculating the total area of the communicating regions, so that the standard cell area can be expected, and errors caused by bringing the areas of the closed communicating regions which generate the overlapping phenomenon and are communicated with the interference points into the calculated average cell area are avoided. After the area of the standard cell area S is obtained, further feedback verification is needed, namely, the connected region which is expected to be S and participates in the standard cell area is compared again, the area of the connected region which is 1.5 times or more than the area of S is subjected to sub division, and then the standard area S is calculated again. The formula for the standard cell area expectation S is as follows:

$S=\frac{\underset{1}{\overset{x}{\Σ}}{A}_{Fx}}{X}$

wherein X is the number of the communication areas with the shape factor larger than 0.85 and excluding the possibility of being an interference point, and is the area of the communication areas with the shape factors larger than 0.85 from Fi to Fx.

The area calculated by the method as the standard cell area has the following advantages:

(1) due to the limitation of the shape factor, the cells with the increased overlapped area are excluded from being involved in the calculation value of the average area S, and the accuracy of the standard cell area is ensured.

(2) Because the shape factor is irrelevant to the size of the independent cells in the image and only relevant to the shape of the blood cells, the method is well suitable for the problem of different sizes of the blood cells in pictures with different scale sizes. The self-adaptive capacity of the system to the image is greatly improved.

(3) Can help to distinguish overlapping cells more accurately.

(4) A connected region area calculation feedback verification algorithm is introduced, and the condition that multiple cells are overlapped but the shape factors are not changed in the actual situation is avoided.

The shape factor of the overlapped cells is greatly changed, and the cells in the connected region with the shape factor less than 0.85 can be judged to be overlapped to different degrees. The degree of overlap is estimated by determining the ratio of the area of the overlapping cells to the area of the standard cells. They are considered to overlap by two for ratios between 1 and 1.5, by 3 cells for ratios between 1.5 and 2.5, and so on. By the method for treating independent cells and overlapped cells differently, the closed-connected region in the image is scanned for multiple times, and the number of blood cells in the image can be well counted.

In yet another embodiment, the present invention provides a fully automatic blood cell counter comprising: blood cell recognition device, synchronous control device and automatic sample feeding device.

The blood cell recognition device is respectively connected with the input device and the synchronous control device and is used for recognizing blood cells in the blood sample;

the synchronous control device is respectively connected with the automatic sample feeding device and the blood cell recognition device and is used for controlling the automatic sample feeding device and the blood cell recognition device to be synchronous;

the automatic sample feeding device is respectively connected with the automatic control device and the blood cell recognition device and is used for conveying blood cell samples;

the blood cell recognition device adopts the following recognition method:

(1) reading a cell image, and converting the image into an HIS space;

(2) histogram equalization;

(3) smoothing, segmenting and edge correcting the image;

(4) extracting characteristics;

(5) cells are identified.

The synchronous control device adopts fuzzy PID control, and a fuzzy rule table of the fuzzy PID control is as follows:

Δk_pfuzzy rule table of (1):

Δk_ifuzzy rule table of (1):

Δk_dfuzzy rule table of (1):

the synchronous control device adopts fuzzy PID control, and a fuzzy control quantity lookup table of the fuzzy PID control is as follows:

Δk_pfuzzy control amount look-up table of (1):

Δk_ifuzzy control amount look-up table of (1):

Δk_dfuzzy control amount look-up table of (1):

the automatic sample feeding device comprises: the test tube rack comprises a base, an up-down lifting track, a depth propelling track and a starting arm with a test tube rack;

the base is an anti-static bakelite plate;

the touch arm with the test tube rack is made of hard and firm materials.

In another embodiment, the present invention further provides a method for synchronizing a fully automatic blood cell counter, comprising:

automatically feeding a sample;

a synchronous control step;

and (3) identifying blood cells.

The blood cell identification step of identifying blood cells in the blood sample;

the synchronous control step is used for controlling the automatic sample feeding device and the blood cell recognition device to be synchronous;

the automatic sample feeding step is used for conveying the blood cell sample;

the blood cell identification step adopts the following identification method:

(1) reading a cell image, and converting the image into an HIS space;

(2) histogram equalization;

(3) smoothing, segmenting and edge correcting the image;

(4) extracting characteristics;

(5) cells are identified.

The synchronous control step uses fuzzy PID control, the fuzzy rule table of the fuzzy PID control is as follows:

Δk_pfuzzy rule table of (1):

Δk_ifuzzy rule table of (1):

Δk_dfuzzy rule table of (1):

the step of synchronous control adopts fuzzy PID control, and a fuzzy control quantity lookup table of the fuzzy PID control is as follows:

Δk_pfuzzy control amount look-up table of (1):

Δk_ifuzzy control amount look-up table of (1):

Δk_dfuzzy control amount look-up table of (1):

a method of synchronizing the fully automatic cytometer using the fully automatic blood cell count of any of the preceding claims;

when the device is used, a blood sample test tube is placed in the sample feeder test tube support, the start key is pressed, the sample feeder is pushed forward according to the design and synchronously lifted and pushed with the lifting distance, the specified position is reached, the arm touches the detection start key, the sample is sucked, and after 15 seconds, the sample feeder automatically returns to prepare for next sample detection.

A synchronization control device of the present invention:

the method adopts a permanent magnet synchronous motor, and a synchronous control device of the permanent magnet synchronous motor adopts PID control. The PID control strategy is a control method applied in many industrial fields, which combines three basic control elements, i.e. past, present and future control, and fig. 1 accurately describes the conventional PID control principle.

The main components of the control system are a controller and a control object. The PID controller is based on a linear combination control principle, compares a given value r (t) with an output value y (t) to obtain a control deviation, and finally linearly combines the proportion, the integral and the derivative of the deviation to obtain an adjustment quantity required by a control system to realize the control of a controlled object, so the PID controller is called as the PID controller. The control algorithm is as follows:

(formula 1)

Wherein e (t) ═ r (t) y (t)

K_p-scale factor

K_i-integration time constant

In practical application of this embodiment, the control system is discontinuous, so that the conventional PID control cannot be directly used, and it needs to discretize the control process to obtain a non-continuous digital quantity, which is referred to as digital PID control. With the development of computer control technology, the capability of the controller is greatly improved, and particularly, good-performance digital PID control can be realized in the aspects of computing capability and realizing speed. Before the digital PID control, the control system is subjected to high-dispersion calculation by the practical method, and the steps are as follows:

u(t)＝u(kT)

e(t)＝e(kT)

(formula 2)

In equation 2, T is the sampling period and K is the number of samples, and in order to ensure sufficient accuracy, the sampling period T must be short enough to perform approximate interval integration on a series of kT sampling time points, and the difference is replaced by the increment between adjacent sampling points.

Substituting formula 2 into formula 1 can obtain the digital PID expression after discrete processing:

(formula 3)

Or

(formula 4)

Wherein:

u (k) -the computed candidate value for the k-th sample;

e (k-1) -deviation from the k-1 th sample;

ki-integral coefficient of the light beam,

kd-the differential coefficient of the signal,

the execution mechanism directly controls the output value u (k) of the computer, namely a correspondence between the value of u (k) and the target position. Therefore, equations 2 and 3 are position-based PID processing modes. According to equation 2 or equation 3, each output of u (k) is associated with a past state, so that all the previous e (k) are overlapped in calculation, and the calculation amount is very large. In addition, since the output value u (k) corresponds to the field output of the controlled object, once the monitor shows an abnormality, u (k) will be affected to cause the controlled object to fluctuate obviously. In some cases, this phenomenon may cause a serious accident. Therefore, the invention also provides an incremental PID control algorithm.

The output obtained by the incremental PID control algorithm is an increment delta u (k), and when the increment of the controlled object is a control parameter, the incremental PID control method is suitable for being adopted. According to a recursion principle:

(formula 5)

Obtaining the PID expression form of the increment form by subtracting the formula 4 and the formula 5

u(k)＝K_p[e(k)e(k1)]+K_ie(k)+K_d[e(k)2e(k1)+e(k2)]

＝K_pe(k)+K_ie(k)+K_d[e(k)e(k1)](formula 6)

Where e (k) ═ e (k) e (k 1).

Expressed by formula 6, the control quantity Δ u (k) is the increment needed by the controlled object at this time, and is calculated and output by the computer. In the present embodiment, the expression used for the control amount output is u (k) ═ u (k-1) + Δ u (k).

The original position type PID is improved on the basis of the algorithm, and the obtained advantages are very obvious: because the output is the increment of the controlled quantity, the negative influence caused by misoperation is reduced. If necessary, the influence can also be eliminated by a logical judgment method. The impact of the improved control mode is small when the control mode is switched, and the non-moving switching can be realized. Once the computer has false action, the incremental control mode can store the information of the control system, so that the anti-interference capability of the system is ensured, and the algorithm does not need to be accumulated, thereby being easy to realize. The value of the control increment au (k) depends on the last sampling of information, and therefore, performing the correlation in the form of partial weighting improves the control characteristics of the system.

In order to ensure that the whole system has good operation effect, the response proportionality coefficient, the integral coefficient and the differential coefficient need to be properly selected, because these parameters determine the multiple performances of the control system, and the specific influence is shown as follows:

1) coefficient of proportionality K_pThe response speed and the control precision of the controlled object are influenced. K_pThe value is increased, the reaction speed of the control system is increased, the control precision is improved, but the overshoot phenomenon is easy to generate, so that the system is unstable. When k is_pIf the size is too small, the adjustment accuracy of the whole control is small, the response process of the system is prolonged, the setting time is further prolonged, and the control performance of the whole system is seriously hindered.

2) Integral parameter K_iThe method is used for reducing errors generated by a static control system, and particularly influences the speed of eliminating the errors, when the coefficient is adjusted to be large, the adjusting process is greatly shortened, and if K is adopted_iIf the value of (b) is too large, integral saturation may occur in the early response of the process, resulting in overshoot in the response process. When K is_iWhen the value is too low, the error generated statically cannot be completely eliminated, and the system cannot be ensured to achieve a good control effect.

3) Differential coefficient K_dIs the influence on the dynamic characteristic of the system, and deviation change can be restrained and predicted in the operation process. K_dThe too large value leads to the suppression being advanced, the response process is delayed, the setting time of the system is increased, and the external interference resisting capability of the controller is reduced.

A modal phase controller. The fuzzy set and the corresponding control concept are proposed in order to research the problems of fixed expression form and fuzziness, and the problems are controlled in a fuzzy mode to be clear and have more orderliness and predictable functions. After continuing intensive research of scientific work, the fuzzy controller with the self-adaptive function is produced, and self-trimming of the control rule can be carried out according to the real-time feedback effect of the control system, so that a good control state is finally achieved. The self-organizing fuzzy controller of the invention has the autonomous learning ability on the basis of the self-adapting fuzzy controller, and overcomes the defect of a pure model lake controller.

The core technology of fuzzy control is to simulate subjective thinking judgment of a human, the intelligent technology is evolved into a specific fuzzy reasoning system, and the concept of a fuzzy set is introduced, and linguistic variables are created, so that a fuzzy control mode becomes a complete solution. The first step in the fuzzy controller design process is to determine the specific function of the system that is used to establish the input output based on actual conditions. Then establishing fuzzy control rules, fuzzifying the specific actual input information, and mapping the fuzzified actual input information to a fuzzy linguistic variable area. And establishing a fuzzy control lookup table, obtaining fuzzy output according to the input situation, and converting the fuzzy output into a specific numerical value.

The fuzzy controller is used as a core part of a fuzzy control system, and the constitutional elements of the fuzzy controller are shown in fig. 2, which comprises three parts, specifically:

1. fuzzification

Fuzzification, that is, processing the clear data to make it no longer concrete, and the actual operation is to map the concrete system input data set to the corresponding fuzzy domain. The fuzzy sets after fuzzy processing are composed of linguistic variables with certain defined meanings, the fuzzy sets corresponding to the input and the output of the system have membership functions with own characteristics, and the selection of the types of the membership functions is based on the experience accumulation of field debugging.

The expression form of the language variable is taken from the language form understood by people, and the specific data discourse domain is divided into areas with different levels according to the visual feeling: for example, feedback much less than expected is shown as negative and large, letters are shown as NB, small comparisons are mostly Negative and Medium (NM), small and small are negative and small (Ns), near the target is zero (z), when output is larger is shown as Ps (positive and small), PM (positive and small), PB (positive and large). The number of the divided areas is determined according to the past control experience, the accuracy is improved due to the fact that the number of the intervals is too large, huge design engineering quantity is caused, programming is complicated, and the control effect is adversely affected due to the fact that the number is too small and the control effect is concise to apply. As to how many gears are most suitable for division, designers need to have rich experience and timely adjust the control process. The selection of membership functions in fuzzy control is various, such as statistical methods, expert methods, etc., and although there are various membership functions for the same control system, the final goal is the same, i.e., stable and accurate control performance is obtained.

2. Knowledge base and fuzzy inference

The database is a part of a knowledge base and consists of fuzzy values after the fuzzy control system variables are fuzzified and specific information elements of the membership functions corresponding to the fuzzy values. The fuzzification of the input is to quantize the actual input quantity according to the system and map the quantized input quantity into the domain range meeting the requirements. The linguistic variables obtained by converting the actual input quantities constitute the input space of the fuzzy controller, and similarly, the linguistic variables obtained by converting the output quantities constitute the output space of the controller. The fuzzy control divides the system into different areas according to the variable range of the system, the range characteristic of each area is corresponding to a language name, and the name clearly defines the area and indicates the bundle of the area division. The number of fuzzy languages corresponding to a single fuzzy set depends on the complexity of the control. The names of these languages usually have a certain meaning, such as negative big, negative small, positive big, positive small, etc. In general, the distribution of fuzzy sets is artificially defined, there is no requirement for symmetry, and the number of sets represents the complexity of control.

3. Defuzzification

The output of the fuzzy controller is a fuzzy variable, namely a language name representing a certain domain, so that the fuzzy controller cannot be directly used for parameter adjustment of field control. The process of converting the fuzzy output quantity into a specific numerical value is fuzzy resolving, and the fuzzy output is summarized into a parameter numerical value applied to field control through a certain reasoning method, which represents the end of the regulation task of the fuzzy controller.

Fuzzy PID controller structure:

the fuzzy PID controller is a complementary strategy of a PID strategy and a fuzzy idea. And acquiring the deviation E and the deviation change rate EC of the control variable according to the acquisition of the field feedback signal, wherein the two parameters are used as the input of the fuzzy PID controller. As shown in fig. 3, the specific tasks completed by the controller are to input specific numerical values and perform fuzzification processing, so as to change the numerical values into language names in different ranges. The analog-to-digital reasoning is performed empirically to obtain fuzzy outputs, which are then de-fuzzy to convert them into corresponding specific values, which are then used to control the system in real time.

The design method of the fuzzy PID controller comprises the following steps:

1) the variable composition of the controller is determined according to the characteristics of system control and the final realization target, in the control of multiple inputs and multiple outputs, the error E and the error change EC of the actual value and the target value are commonly used as the controller input, and the integral coefficient, the mark coefficient and the proportionality coefficient of the PID controller are used as the output.

2) Setting a quantization grade, a quantization factor and a scale factor, and quantizing the actual input signal of the system;

3) setting the quantitative fuzzy subsets of input and output variables according to the field operation experience, selecting proper subset number and linguistic variables, selecting the membership function matched with the subset number and the linguistic variables,

4) the determination of the fuzzy control rule is the core established by the fuzzy PID controller, the fuzzy control rule depends on the professional knowledge and the control experience of field operators, a fuzzy statement which accords with the field process is obtained according to the accumulation of the daily operation experience, and the writing principle of the control rule is on the premise of ensuring the good control performance of the system;

5) making a fuzzy lookup table: according to the fuzzy control rule, the input and output variables are in one-to-one correspondence in the fuzzy control table according to the fuzzy control rule, namely, the output quantity corresponding to the input variable is inquired according to the input variable.

6) Output quantity deblurring processing: the process of converting the fuzzy output interval into a specific numerical value is called deblurring.

7) After the design of the fuzzy PID controller is finished, the validity and the reliability of the fuzzy PID controller still need to be verified and adjusted, and the fuzzy PID controller can be observed on line or can be tested whether to accord with an expected target by using an off-line simulation experiment or computer simulation.

The traditional PID controller has good control performance and is widely applied to the control field, but cannot cope with the influence caused by various interferences and load changes generated in the actual traffic process due to the self limitation, so that certain limitation exists in the application occasions with higher control requirements. The invention combines the traditional PID control and fuzzy control technology, so that the traditional PID controller has self-adaptive capacity and improves the application in the industrial field. According to the embodiment, an alternating-phase and combined-phase control mode is selected according to the system control requirement, and the model lake PID controller is adopted to perform speed compensation on the second motor, so that the synchronous operation mode of the two motors is finally realized.

Using MATLAB, fuzzy is input in the main window, a input output variable is selected based on the system properties, and the motor synchronous control system of the present invention is a two-input three-output system, and the fuzzy input space is a deviation e and a deviation change rate ec, and the proportional, integral, and differential increments Δ kp, Δ ki, and Δ kd are used as a fuzzy output space.

Determining a membership function: and setting the membership function of the input and output variables to be a triangular membership function according to experience, wherein the boundary value adopts a Gaussian membership function.

The fuzzy language value is defined as { NB, NM, NS, ZE, PS, PM, PB }. According to the synchronous precision required by the synchronous control of the motor, the deviation e and the deviation variable ec are converted into domains of { -3, -2, -1, 0, 1, 2 and 3 }; the quantization domain of the fuzzy output delta kp is { -0.3, -0.2, -0.1, 0, 0.1, 0.2, 0.3}, the quantization domain of the delta ki is { -0.06, -0.04, -0.02, 0, 0.02, 0.04, 0.06}, and the quantization domain of the delta Kd is { -0.3, -0.2, -0.1, 0, 0.1, 0.2, 0.3).

The fuzzy rule is summarized by operation experience and expert knowledge, so that according to the action of PID parameters, and the practical years of operation experience of operators and debuggers when the operators and the debuggers change load and generate disturbance in the synchronous control of the motor, the following fuzzy control rule table is established:

tables 1-1. delta. k_pFuzzy rule table of

Tables 1-2. delta. k_iFuzzy rule table of

Tables 1 to 3. delta. k_dFuzzy rule table of

The fuzzy control quantity lookup table has the functions of: knowing the fuzzy quantity input, the fuzzy output of the controller is obtained through the query of the table. Establishing a fuzzy inference rule according to tables 1-1 to 1-3, and resolving the fuzzy by adopting a gravity center method to obtain a fuzzy control quantity query table for PLC offline query.

TABLE 2-1. DELTA.k_pFuzzy control quantity look-up table

Tables 2-2. delta. k_iFuzzy control quantity look-up table

Tables 2 to 3. delta. k_dFuzzy control quantity look-up table

Selection of quantization factor and scale factor:

1. determination of a quantization factor

The fuzzy domains of the deviation e and the deviation change ec in the control system are { a, b }, the change range of the actual error e is [ -30, 30], the change range of the error change ec is [ -60, 60],

Ke＝(ba)/[30(30)]＝6/60＝0.1

Kec＝(ba)/[60(60)]＝6/120＝0.05

where Ke is the quantization factor for error e and Kec is the quantization factor for error change ec.

x＝Keey＝Kecec

When x is an integer, the fuzzy quantity E is x, and when x is a decimal, the fuzzy quantity E is rounded to E, and y can be treated as the fuzzy input quantity Ec in the same way.

During the synchronous test, the magnitude of the quantization factor affects the dynamic characteristics of the control system. When Ke is large, the system can peak quickly, but rapid adjustments bring about a severe overshoot illusion and the overshoot time is quite long. Kec is more favorable for system stability and reduces overshoot time, but results in slow system response. Therefore, the quantization factor needs to be adjusted according to the actual situation, so that the synchronous control of the system reaches the optimal state.

2. Determination of a scaling factor

After the fuzzy controller inputs the fuzzy quantity E and EC, the fuzzy output quantity U after table look-up in the PLC is changed into a specific output value U through the action of a scale factor.

Let the ambiguity field of U be [ a, b]The actual output range is [ u ]_min，u_max]Then the scale factor K_uComprises the following steps:

Ku＝(u_maxu_min)/(ba)

Kk_P＝0.4/0.6＝0.67

Kk_i＝50/0.12＝426.67

Kk_d＝0.2/0.6＝0.33

the finally obtained control quantity u is KuU, wherein the size of the scaling factor also influences the dynamic response effect of the control system, and when the scaling factor is too large, the adjustment speed is too rapid, so that large fluctuation is brought; the response process of the system is reduced by the partial reduction of the scale factors, and the adjustment precision of the system in a stable state is influenced. The magnitude of the scaling factor affects the tuning effect of the fuzzy controller.

In another embodiment of the present invention, there is provided another synchronization control apparatus:

the method adopts a permanent magnet synchronous motor, and a synchronous control device of the permanent magnet synchronous motor adopts PID control. PID control is a method based on error feedback control, and heretofore, PID controllers are widely used in industrial control systems.

The speed output equation of the permanent magnet synchronous motor of the invention is expressed as follows:

$\stackrel{\·}{\mathrm{\ω}}=\frac{n_p{\mathrm{\ψ}}_r}{J}{i}_q-\frac{B}{J}\mathrm{\ω}-\frac{T_L}{J}=f(\mathrm{\ω},T_L)+{\mathrm{bi}}_q$

wherein, $f(\mathrm{\ω},T_L)=-\frac{B}{J}\mathrm{\ω}-\frac{T_L}{J},b=\frac{n_p{\mathrm{\ψ}}_r}{J}$

then, the speed active disturbance rejection controller algorithm of the permanent magnet synchronous motor can be deduced as follows:

(1) tracking differentiator

$\left\{\begin{array}{c}{e}_0={\mathrm{\ω}}_i-\widehat{\mathrm{\ω}}\\ \stackrel{\·}{\mathrm{\ω}}=-rfal(e_0,a_0,{\mathrm{\δ}}_0)\end{array}\right.$

In the above formula, the first and second carbon atoms are,is the system given speed, ω isThe tracking speed of (2).

(2) Extended state observer

$\left\{\begin{array}{c}{e}_1=z_{21}-{\mathrm{\ω}}^{*}\\ {\stackrel{\·}{z}}_{21}=z_{22}-{\mathrm{\β}}_{1}fal(e_1,a_1,{\mathrm{\δ}}_1)+b_{0}u\left(t\right)\\ {\stackrel{\·}{z}}_{22}=z_{33}-{\mathrm{\β}}_{2}fal(e_1,a_1,{\mathrm{\δ}}_1)\\ {\stackrel{\·}{z}}_{33}=-{\mathrm{\β}}_{3}fal(e_1,a_2,{\mathrm{\δ}}_2)\end{array}\right.$

In the formula of omega^*Is a feedback signal of the system, z₂₁Is omega^*Of the tracking signal z₂₂Is an observed value of an unknown disturbance of the system.

(3) Nonlinear state error feedback control law

$\left\{\begin{array}{c}{e}_2=\widehat{\mathrm{\ω}}-z_{21}\\ u_1\left(t\right)={\mathrm{\λ}}_{1}fal(e_2,a_2,{\mathrm{\δ}}_2)+{\mathrm{\λ}}_{2}fal(e_3,a_3,{\mathrm{\δ}}_3)\end{array}\right.$

The control quantity of disturbance compensation is:

$u\left(t\right)=u_1\left(t\right)-\frac{z_{22}}{b}$

u (t) is the control quantity input to the current loop after disturbance compensation.

The automatic sample feeding device of the full-automatic blood cell counter consists of a base, an up-down lifting track, a depth propelling track and a starting arm with a test tube rack. The length of the base is 165mm, the width is 85mm, and the height is 30 mm; the length of the depth propulsion track is 120mm, the width of the depth propulsion track is 90mm, and the maximum depth propulsion distance is 80 mm; the height of the lifting shaft is 160mm, and the maximum lifting distance is 80 mm; take test tube holder's start arm length 80mm, the free end is semi-circular, and width 20mm, support set up 4 holes, every hole diameter 15mm, degree of depth 45 mm. The base is an anti-static bakelite plate, the touch arm with the test tube rack is made of hard firm materials, and the lifting shaft is made of stainless steel materials. When the device is used, a blood sample test tube is placed in the sample feeder test tube support, the start key is pressed, the sample feeder is pushed forward according to the design and synchronously lifted and pushed with the lifting distance, the specified position is reached, the arm touches the detection start key, the sample is sucked, and after 15 seconds, the sample feeder automatically returns to prepare for next sample detection.

All of the above mentioned intellectual property rights are not intended to be restrictive to other forms of implementing the new and/or new products. Those skilled in the art will take advantage of this important information, and the foregoing will be modified to achieve similar performance. However, all modifications or alterations are based on the new products of the invention and belong to the reserved rights.

Claims (10)

1. A method of blood cell counting comprising:

automatically feeding a sample;

a synchronous control step;

a blood cell recognition step;

and displaying the result.

2. The fully automatic blood cell counting method according to claim 1,

the blood cell identification step of identifying blood cells in the blood sample;

the synchronous control step is used for controlling the automatic sample feeding device and the blood cell recognition device to be synchronous;

the automatic sample feeding step is used for conveying the blood cell sample;

and the result display step is used for outputting the blood cell sample and/or data result and displaying the blood cell counting result.

3. The blood cell counting method according to claim 1 or 2, wherein the blood cell identification step uses an identification method as follows:

(1) reading a cell image, and converting the image into an HIS space;

(2) histogram equalization;

(3) smoothing, segmenting and edge correcting the image;

(4) extracting characteristics;

(5) cells are identified.

4. A method according to any one of claims 1 to 3, wherein the step (1) of reading in the cell image and converting the image into HIS space comprises the following steps:

for any RGB value in the interval [0, 1], the conversion formula of H, S, I component in HSI model corresponding to the value is:

$I=\frac{1}{3}(R+G+B)$

$S=1-\frac{3}{(R+G+B)}\[min(R,G,B)\]$

$H=\arccos \left\{\frac{\frac{1}{2}\[(R-G)+(R-B)\]}{{\[{(R-G)}^2+(R-G)(G-B)\]}^{\frac{1}{2}}}\right\}.$

5. a method according to any one of claims 1 to 4, wherein the synchronization control step: the PID control method adopts the PID control of the permanent magnet synchronous motor, and comprises the following steps:

(1) tracking differentiator

$\left\{\begin{array}{c}{e}_0={\mathrm{\ω}}_i-\widehat{\mathrm{\ω}}\\ \stackrel{\·}{\mathrm{\ω}}=-rfal(e_0,a_0,{\mathrm{\δ}}_0)\end{array}\right.$

In the above formula, the first and second carbon atoms are,is the system given speed, ω isThe tracking speed of (2);

(2) extended state observer

$\left\{\begin{array}{c}{e}_1=z_{21}-{\mathrm{\ω}}^{*}\\ {\stackrel{\·}{z}}_{21}=z_{22}-{\mathrm{\β}}_{1}fal(e_1,a_1,{\mathrm{\δ}}_1)+b_{0}u\left(t\right)\\ {\stackrel{\·}{z}}_{22}=z_{33}-{\mathrm{\β}}_{2}fal(e_1,a_1,{\mathrm{\δ}}_1)\\ {\stackrel{\·}{z}}_{33}=-{\mathrm{\β}}_{3}fal(e_1,a_2,{\mathrm{\δ}}_2)\end{array}\right.$

In the above formula, omega^*Is a feedback signal of the system, z₂₁Is omega^*Of the tracking signal z₂₂Is an observed value of an unknown disturbance of the system;

(3) nonlinear state error feedback control law

$\left\{\begin{array}{c}{e}_2=\stackrel{\·}{\mathrm{\ω}}-z_{21}\\ u_1\left(t\right)={\mathrm{\λ}}_{1}fal(e_2,a_2,{\mathrm{\δ}}_2)+{\mathrm{\λ}}_{2}fal(e_3,a_3,{\mathrm{\δ}}_3)\end{array}\right.$

The control quantity of disturbance compensation is:

$u\left(t\right)=u_1\left(t\right)-\frac{z_{22}}{b}$

u (t) is the control quantity input to the current loop after disturbance compensation.

6. The fully automatic blood cell counting method according to any one of claims 1 to 5, wherein the synchronization control step further comprises: the blood sample test tube is placed in a sample feeder test tube support of the blood cell counter, the sample feeder is pushed forward and synchronously lifted and pushed forward according to the design and the lifting distance by pressing a start key, the sample feeder reaches a specified position, an arm touches a detection start key, the sample is sucked, and after 15 seconds, the sample feeder automatically returns.

7. Use of a method according to any one of claims 1 to 6 in the preparation of a cytometer.

8. A cytometer operating according to the method of any of claims 1 to 6, comprising: blood cell recognition device, synchronous control device, automatic sample feeding device and result display device.

9. The cytometer of claim 8 wherein,

the blood cell recognition device is respectively connected with the result display device and the synchronous control device and is used for recognizing blood cells in the blood sample;

the synchronous control device is respectively connected with the automatic sample feeding device and the blood cell recognition device and is used for controlling the automatic sample feeding device and the blood cell recognition device to be synchronous;

the automatic sample feeding device is respectively connected with the automatic control device and the blood cell recognition device and is used for conveying blood cell samples;

and the result display device is used for displaying the result of blood cell counting and outputting the data result.

10. The cytometer according to claim 8 or 9, wherein the blood cell identification device adopts the following identification method:

(1) reading a cell image, and converting the image into an HIS space;

(2) histogram equalization;

(3) smoothing, segmenting and edge correcting the image;

(4) extracting characteristics;

(5) identifying the cell;

the synchronization control device: the PID control method adopts the PID control of the permanent magnet synchronous motor, and comprises the following steps:

(1) tracking differentiator

$\left\{\begin{array}{c}{e}_0={\mathrm{\ω}}_i-\widehat{\mathrm{\ω}}\\ \stackrel{\·}{\mathrm{\ω}}=-rfal(e_0,a_0,{\mathrm{\δ}}_0)\end{array}\right.$

In the above formula, the first and second carbon atoms are,is the system given speed, ω isThe tracking speed of (2);

(2) extended state observer

$\left\{\begin{array}{c}{e}_1=z_{21}-{\mathrm{\ω}}^{*}\\ {\stackrel{\·}{z}}_{21}=z_{22}-{\mathrm{\β}}_{1}fal(e_1,a_1,{\mathrm{\δ}}_1)+b_{0}u\left(t\right)\\ {\stackrel{\·}{z}}_{22}=z_{33}-{\mathrm{\β}}_{2}fal(e_1,a_1,{\mathrm{\δ}}_1)\\ {\stackrel{\·}{z}}_{33}=-{\mathrm{\β}}_{3}fal(e_1,a_2,{\mathrm{\δ}}_2)\end{array}\right.$

In the above formula, omega^*Is a feedback signal of the system, z₂₁Is omega^*Of the tracking signal z₂₂Is an observed value of an unknown disturbance of the system;

(3) nonlinear state error feedback control law

$\left\{\begin{array}{c}{e}_2=\stackrel{\·}{\mathrm{\ω}}-z_{21}\\ u_1\left(t\right)={\mathrm{\λ}}_{1}fal(e_2,a_2,{\mathrm{\δ}}_2)+{\mathrm{\λ}}_{2}fal(e_3,a_3,{\mathrm{\δ}}_3)\end{array}\right.$

The control quantity of disturbance compensation is:

$u\left(t\right)=u_1\left(t\right)-\frac{z_{22}}{b}$

u (t) is the control quantity input to the current loop after disturbance compensation.