CN103543277A - Blood type result recognition algorithm based on grey level analysis and type recognition - Google Patents

Blood type result recognition algorithm based on grey level analysis and type recognition Download PDF

Info

Publication number
CN103543277A
CN103543277A CN201310416614.3A CN201310416614A CN103543277A CN 103543277 A CN103543277 A CN 103543277A CN 201310416614 A CN201310416614 A CN 201310416614A CN 103543277 A CN103543277 A CN 103543277A
Authority
CN
China
Prior art keywords
blood type
image
reagent card
gray
type reagent
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201310416614.3A
Other languages
Chinese (zh)
Other versions
CN103543277B (en
Inventor
罗刚银
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Suzhou Institute of Biomedical Engineering and Technology of CAS
Original Assignee
Suzhou Institute of Biomedical Engineering and Technology of CAS
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Suzhou Institute of Biomedical Engineering and Technology of CAS filed Critical Suzhou Institute of Biomedical Engineering and Technology of CAS
Priority to CN201310416614.3A priority Critical patent/CN103543277B/en
Publication of CN103543277A publication Critical patent/CN103543277A/en
Application granted granted Critical
Publication of CN103543277B publication Critical patent/CN103543277B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
    • G01N33/48—Biological material, e.g. blood, urine; Haemocytometers
    • G01N33/50—Chemical analysis of biological material, e.g. blood, urine; Testing involving biospecific ligand binding methods; Immunological testing
    • G01N33/80—Chemical analysis of biological material, e.g. blood, urine; Testing involving biospecific ligand binding methods; Immunological testing involving blood groups or blood types or red blood cells

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Hematology (AREA)
  • Engineering & Computer Science (AREA)
  • Molecular Biology (AREA)
  • Biomedical Technology (AREA)
  • Chemical & Material Sciences (AREA)
  • Immunology (AREA)
  • Urology & Nephrology (AREA)
  • Biotechnology (AREA)
  • Microbiology (AREA)
  • Cell Biology (AREA)
  • Food Science & Technology (AREA)
  • Medicinal Chemistry (AREA)
  • Physics & Mathematics (AREA)
  • Analytical Chemistry (AREA)
  • Biochemistry (AREA)
  • General Health & Medical Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • Pathology (AREA)
  • Investigating Or Analysing Biological Materials (AREA)

Abstract

本发明公开了一种基于灰度分析与种类识别的血型结果识别算法，它包括以下步骤：灰度变换、平滑滤波、倾斜校正、微柱小管分割、阈值处理、灰度分析、种类识别、对比检测结果，本发明采用灰度分析与种类识别相结合的方法对血型检测结果进行自动识别，提高了血型检测结果识别的效率，具有快速准确的优点，同时通过抖动消除、倾斜校正、微柱小管分割、种类识别等操作，提高了结果识别的有效性。

Figure 201310416614

The invention discloses a blood type result recognition algorithm based on grayscale analysis and category recognition, which comprises the following steps: grayscale transformation, smoothing filter, inclination correction, segmentation of microcolumns and tubules, threshold value processing, grayscale analysis, category recognition, and comparison Detection results, the present invention adopts the method of combining grayscale analysis and type recognition to automatically identify the blood type detection results, which improves the efficiency of blood type detection result recognition and has the advantages of fast and accurate Segmentation, category identification and other operations improve the effectiveness of result identification.

Figure 201310416614

Description

一种基于灰度分析与种类识别的血型结果识别算法A Blood Type Result Recognition Algorithm Based on Gray Scale Analysis and Type Recognition

技术领域 technical field

本发明涉及血型检测结果识别领域，具体涉及一种基于灰度分析与种类识别的血型结果识别算法。  The invention relates to the field of blood type detection result recognition, in particular to a blood type result recognition algorithm based on grayscale analysis and type recognition. the

背景技术 Background technique

 血型检测技术可以分为玻片法、试管法、微板法和血型试剂卡法等。血型试剂卡法是目前国际卫生组织推荐的最先进的血型检测技术。玻片法和试管法是出现较早的血型分析技术，具有检测条件简单的优点，但是其加样过程往往通过手动实现，反应结果的判断通常是通过人眼来判断，其自动化程度不高，人为因素干扰比较严重，且结果的判断受人的主观意识影响较强。微板法是后来发展起来的血型分析技术，但是采用微板法进行血型分析时反应后的微型小板保存不方便，不符合输血要求的可溯源性。  Blood type detection techniques can be divided into slide method, test tube method, microplate method and blood type reagent card method. The blood type reagent card method is currently the most advanced blood type detection technology recommended by the International Health Organization. The glass slide method and the test tube method are earlier blood type analysis techniques, which have the advantage of simple detection conditions, but the process of adding samples is often done manually, and the judgment of the reaction results is usually judged by human eyes, and the degree of automation is not high. The interference of human factors is relatively serious, and the judgment of the results is strongly influenced by people's subjective consciousness. The microplate method is a blood type analysis technology developed later, but it is inconvenient to store the reacted microplatelets when the microplate method is used for blood type analysis, and it does not meet the traceability requirements of blood transfusion. the

血型试剂卡由微柱小管和铭牌组成。微柱小管是血型检测时的反应容器，检测时将红细胞、血清或红细胞试剂加在微柱小管的上部，通过分析反应后微柱小管中红细胞的分布情况来判断检测结果。铭牌是血型试剂卡的种类和信息标识，它上面印有血型试剂卡名称、一维条码、微柱小管各管信息等。  The blood type reagent card consists of a microcolumn tube and a nameplate. The micro-column tube is the reaction container for blood type testing. Red blood cells, serum or red blood cell reagents are added to the upper part of the micro-column tube during the test, and the test result can be judged by analyzing the distribution of red blood cells in the micro-column tube after the reaction. The nameplate is the type and information identification of the blood type reagent card. It is printed with the name of the blood type reagent card, one-dimensional barcode, and the information of each tube of the microcolumn and small tube. the

国内对血型试剂卡法的应用主要还停留在手工阶段，即用人眼判读试剂卡的检测结果，其检测自动化程度还很低，采用人眼判读试剂卡的检测结果，极有可能出现检测结果的人为误判  The domestic application of the blood type reagent card method is still mainly in the manual stage, that is, the detection results of the reagent cards are interpreted by human eyes, and the degree of automation of the detection is still very low. Human misjudgment

国外对血型试剂卡法的应用已经进入自动化阶段，即用仪器判读试剂卡的检测结果，用条码枪扫描血型试剂卡的种类。但是，国外仪器在判读血型试剂卡的检测结果时，往往默认所采集的血型试剂卡的图像是端正清晰的，而未对血型试剂卡的抖动情况进行消除和倾斜情况进行校正；同时，血型图像分析时也只对血型试剂卡上的固定区域进行扫描，而没有智能化地分割出用于识别血型结果的有效区域（即微柱小管）。这两个问题都会导致所分析目标区域与有效区域的偏离，从而影响检测结果识别的准确性。同时，对试剂卡的种类采用附加的条码枪进行识别，而未能通过铭牌自有的颜色和字符信息进行种类识别，其缺陷是增加了仪器的复杂性和检测的操作步骤。 The application of the blood type reagent card method in foreign countries has entered the automation stage, that is, the test results of the reagent card are interpreted by instruments, and the types of the blood type reagent card are scanned with a barcode gun. However, when foreign instruments interpret the test results of blood type reagent cards, they often assume that the images of the collected blood type reagent cards are correct and clear, and do not eliminate the vibration of the blood type reagent cards and correct the tilt; at the same time, the blood type images During the analysis, only the fixed area on the blood type reagent card is scanned, and the effective area (ie, the microcolumn tubule) for identifying the blood type result is not intelligently segmented. These two problems will lead to the deviation between the analyzed target area and the effective area, thus affecting the accuracy of detection results. At the same time, an additional barcode gun is used to identify the type of the reagent card, but the type cannot be identified through the color and character information of the nameplate. The defect is that the complexity of the instrument and the operation steps of the detection are increased.

发明内容 Contents of the invention

本发明的目的在于克服现有血型检测结果识别方法的不足，提供一种基于灰度分析与种类识别的血型结果识别算法。  The purpose of the present invention is to overcome the shortcomings of the existing blood type test result recognition methods, and provide a blood type result recognition algorithm based on grayscale analysis and type recognition. the

为实现上述技术目的，达到上述技术效果，本发明通过以下技术方案实现：  In order to achieve the above-mentioned technical purpose and achieve the above-mentioned technical effect, the present invention is realized through the following technical solutions:

一种基于灰度分析与种类识别的血型结果识别算法，包括以下步骤： A blood type result recognition algorithm based on grayscale analysis and type recognition, comprising the following steps:

 1、一种基于灰度分析与种类识别的血型结果识别算法，其特征在于，该算法包括以下步骤： 1. A blood type result recognition algorithm based on grayscale analysis and type recognition, characterized in that the algorithm comprises the following steps:

步骤1）灰度变换 Step 1) Grayscale transformation

采用加权平均值方法对血型试剂卡的24位真彩色图像进行灰度变换，设坐标（ 

Figure 830693DEST_PATH_IMAGE002
）处各分量的亮度值分别为
Figure 396803DEST_PATH_IMAGE004
、
Figure 636155DEST_PATH_IMAGE006
、
Figure 996729DEST_PATH_IMAGE008
，则灰度变换以后该点的灰度值
Figure 383848DEST_PATH_IMAGE010
， The 24-bit true-color image of the blood type reagent card is grayscale transformed using the weighted average method, and the coordinates (
Figure 830693DEST_PATH_IMAGE002
) at the brightness values of each component are
Figure 396803DEST_PATH_IMAGE004
,
Figure 636155DEST_PATH_IMAGE006
,
Figure 996729DEST_PATH_IMAGE008
, then the gray value of the point after the gray scale transformation
Figure 383848DEST_PATH_IMAGE010
,

       

Figure 439135DEST_PATH_IMAGE012
 ，其中，
Figure 329731DEST_PATH_IMAGE014
、
Figure 126786DEST_PATH_IMAGE016
、
Figure 673305DEST_PATH_IMAGE018
分别为各颜色分量的权重，即得到转换后的灰度图像；
Figure 439135DEST_PATH_IMAGE012
,in,
Figure 329731DEST_PATH_IMAGE014
,
Figure 126786DEST_PATH_IMAGE016
,
Figure 673305DEST_PATH_IMAGE018
are the weights of each color component, that is, the converted grayscale image is obtained;

步骤2）平滑滤波 Step 2) smooth filtering

采用中值滤波方法对血型试剂卡灰度图像进行噪声滤波，设

Figure 846797DEST_PATH_IMAGE020
为选择局部区域中灰度值的集合，
Figure 44429DEST_PATH_IMAGE022
运算为求序列的中值，
Figure 12385DEST_PATH_IMAGE024
为想要更换灰度值的指定点，则： The grayscale image of the blood type reagent card was filtered by the median filter method.
Figure 846797DEST_PATH_IMAGE020
To select a set of gray values in a local area,
Figure 44429DEST_PATH_IMAGE022
The operation is to find the median value of the sequence,
Figure 12385DEST_PATH_IMAGE024
For the specified point where you want to replace the gray value, then:

                        

Figure 108517DEST_PATH_IMAGE026
                        
Figure 108517DEST_PATH_IMAGE026

即得到更加清晰的灰度图像； That is, a clearer grayscale image is obtained;

步骤3）抖动消除 Step 3) Jitter Elimination

采用膨胀腐蚀方法对血型试剂卡灰度图像进行抖动消除，得到更加清晰的灰度图像； The dilation and corrosion method is used to eliminate the jitter of the gray-scale image of the blood type reagent card to obtain a clearer gray-scale image;

步骤4）倾斜校正 Step 4) Tilt Correction

采用斜率计算方法得到血型试剂卡的倾斜角度，通过反向旋转相应的角度校正其倾斜状态，血型试剂卡的铭牌中有四个空白平行四边形区域，找出其中的平行四边形区域，取其左上角坐标（）和左下角坐标（

Figure 622992DEST_PATH_IMAGE030
），通过公式：
Figure 761849DEST_PATH_IMAGE032
=
Figure 33693DEST_PATH_IMAGE034
计算出图像倾斜的角度； The inclination angle of the blood type reagent card is obtained by using the slope calculation method, and the inclination state is corrected by reversely rotating the corresponding angle. There are four blank parallelogram areas in the nameplate of the blood type reagent card, find out the parallelogram area, and take its upper left corner coordinate( ) and the coordinates of the lower left corner (
Figure 622992DEST_PATH_IMAGE030
), through the formula:
Figure 761849DEST_PATH_IMAGE032
=
Figure 33693DEST_PATH_IMAGE034
Calculate the angle at which the image is tilted;

Figure DEST_PATH_IMAGE035
步骤5）微柱小管分割
Figure DEST_PATH_IMAGE035
Step 5) Microcolumn tubule segmentation

采用模板匹配方法，即通过相关性公式：分割出血型试剂卡中用于图像识别的有效区域，即微柱小管； Using the template matching method, that is, through the correlation formula: segment the effective area for image recognition in the blood type reagent card, that is, the microcolumn tubule;

步骤6）阈值处理 Step 6) Thresholding

采用二值化方法对分割出的微柱小管图像进行阈值处理，把微柱小管图像由灰度图像转换成为只有黑白两种颜色的二值化图像，设定某一阈值

Figure DEST_PATH_IMAGE037
，将原始灰度图像各像素的像素值和阈值作比较，大于
Figure 516124DEST_PATH_IMAGE038
的重新取值为255，反之则为0，设阈值处理后的像素值为，即： Use the binarization method to perform threshold value processing on the segmented microcolumn tubule image, convert the microcolumn tubule image from a grayscale image to a binary image with only black and white colors, and set a certain threshold
Figure DEST_PATH_IMAGE037
, the pixel value of each pixel of the original grayscale image and threshold for comparison, greater than
Figure 516124DEST_PATH_IMAGE038
The re-value of is 255, otherwise it is 0, and the pixel value after thresholding is set to be ,Right now:

         

Figure 666985DEST_PATH_IMAGE044
Figure 85328DEST_PATH_IMAGE046
           
Figure 666985DEST_PATH_IMAGE044
Figure 85328DEST_PATH_IMAGE046
 

步骤7）灰度分析 Step 7) Grayscale Analysis

采用像素求和方法把每个微柱小管的下端分成上、中、下三个小区域，分别计算上、中、下三个小区域的像素之和，即用像素的分布情况代表微柱小管下端的红细胞分布情况； The lower end of each microcolumn tubule is divided into upper, middle, and lower small areas by pixel summation method, and the sum of pixels in the upper, middle, and lower three small areas is calculated respectively, that is, the distribution of pixels is used to represent the microcolumn tubule The distribution of red blood cells at the lower end;

步骤8）种类识别 Step 8) Type Identification

采用颜色识别和字符识别结合的方法，通过分析血型试剂卡铭牌中不同颜色区域中的字符差异来识别其种类的不同。 The method of combining color recognition and character recognition is adopted to identify the difference of the blood type reagent card by analyzing the character difference in different color areas on the nameplate of the blood type reagent card.

步骤9）对比检测结果  Step 9) Compare the test results

据微柱小管中红细胞的分布情况，同时结合当前血型试剂卡的种类，得到该血型试剂卡的血型检测结果。 According to the distribution of red blood cells in the microcolumn tubules, combined with the type of the current blood type reagent card, the blood type detection result of the blood type reagent card is obtained.

本发明的有益效果是:  The beneficial effects of the present invention are:

采用本发明技术方案，倾斜校正和抖动消除提高了血型检测结果识别的准确性，采用模板匹配分割出微柱小管提高了目标区域识别的准确性，采用颜色识别和字符识别相结合的试剂卡种类识别方法减少了仪器的复杂性和检测的操作步骤。 By adopting the technical solution of the present invention, the tilt correction and shake elimination improve the accuracy of blood type test result recognition, the use of template matching to segment microcolumn tubules improves the accuracy of target area recognition, and the combination of color recognition and character recognition is used to identify the types of reagent cards The identification method reduces the complexity of the instrument and the operational steps of the detection.

上述说明仅是本发明技术方案的概述，为了能够更清楚了解本发明的技术手段，并可依照说明书的内容予以实施，以下以本发明的较佳实施例并配合附图详细说明如后。本发明的具体实施方式由以下实施例及其附图详细给出。  The above description is only an overview of the technical solutions of the present invention. In order to understand the technical means of the present invention more clearly and implement them according to the contents of the description, the preferred embodiments of the present invention and accompanying drawings are described in detail below. The specific embodiment of the present invention is given in detail by the following examples and accompanying drawings. the

附图说明 Description of drawings

此处所说明的附图用来提供对本发明的进一步理解，构成本申请的一部分，本发明的示意性实施例及其说明用于解释本发明，并不构成对本发明的不当限定。在附图中：  The accompanying drawings described here are used to provide a further understanding of the present invention and constitute a part of the application. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations to the present invention. In the attached picture:

图1为本发明的实施流程图； Fig. 1 is the implementation flowchart of the present invention;

图2为本发明中进行倾斜校正的过程图； Fig. 2 is a process diagram of tilt correction in the present invention;

图3为本发明中进行微柱小管分割的三种匹配模板； Fig. 3 is three kinds of matching templates that carry out microcolumn tubule segmentation in the present invention;

图4为本发明中微柱小管灰度图像进行阈值处理后的二值化图像； Fig. 4 is the binarized image after the grayscale image of the microcolumn tubule in the present invention is subjected to threshold value processing;

图5为本发明中常见七种血型试剂卡的铭牌； Fig. 5 is the nameplate of common seven kinds of blood type reagent cards among the present invention;

图6为本发明中进行种类识别时从铭牌中提取的字符； Fig. 6 is the character extracted from the nameplate when carrying out type identification in the present invention;

图7为本发明中字符识别时采用的11特征点的截取方式。 Fig. 7 is the interception method of 11 feature points adopted in character recognition in the present invention.

具体实施方式 Detailed ways

下面将参考附图并结合实施例，来详细说明本发明。  The present invention will be described in detail below with reference to the accompanying drawings and in combination with embodiments. the

参照图1所示，一种基于灰度分析与种类识别的血型结果识别算法，包括以下步骤：  Referring to Figure 1, a blood type result recognition algorithm based on grayscale analysis and type recognition includes the following steps:

步骤1）灰度变换 Step 1) Grayscale transformation

采用加权平均值方法对血型试剂卡的24位真彩色图像进行灰度变换，设坐标（

Figure 377770DEST_PATH_IMAGE002
）处各分量的亮度值分别为
Figure 234867DEST_PATH_IMAGE004
、
Figure 620499DEST_PATH_IMAGE006
、
Figure 272060DEST_PATH_IMAGE008
，则灰度变换以后该点的灰度值
Figure 317376DEST_PATH_IMAGE010
， The 24-bit true-color image of the blood type reagent card is grayscale transformed using the weighted average method, and the coordinates (
Figure 377770DEST_PATH_IMAGE002
) at the brightness values of each component are
Figure 234867DEST_PATH_IMAGE004
,
Figure 620499DEST_PATH_IMAGE006
,
Figure 272060DEST_PATH_IMAGE008
, then the gray value of the point after the gray scale transformation
Figure 317376DEST_PATH_IMAGE010
,

       

Figure 650269DEST_PATH_IMAGE012
 ，其中，
Figure 199062DEST_PATH_IMAGE014
、
Figure 287104DEST_PATH_IMAGE016
、分别为各颜色分量的权重，即得到转换后的灰度图像，本图例取
Figure 939988DEST_PATH_IMAGE048
=0.299、=0.587、
Figure 280970DEST_PATH_IMAGE018
=0.114分对血型试剂卡的24位真彩色图像进行灰度变换，得到转换后的灰度图像；
Figure 650269DEST_PATH_IMAGE012
,in,
Figure 199062DEST_PATH_IMAGE014
,
Figure 287104DEST_PATH_IMAGE016
, are the weights of each color component, that is, the converted grayscale image is obtained. In this example,
Figure 939988DEST_PATH_IMAGE048
=0.299, =0.587,
Figure 280970DEST_PATH_IMAGE018
=0.114 points The 24-bit true-color image of the blood type reagent card is converted into grayscale to obtain the converted grayscale image;

步骤2）平滑滤波 Step 2) smooth filtering

采用中值滤波方法对血型试剂卡灰度图像进行噪声滤波，设

Figure 539913DEST_PATH_IMAGE020
为选择局部区域中灰度值的集合，
Figure 294243DEST_PATH_IMAGE022
运算为求序列的中值，
Figure 719670DEST_PATH_IMAGE024
为想要更换灰度值的指定点，则： The grayscale image of the blood type reagent card was filtered by the median filter method.
Figure 539913DEST_PATH_IMAGE020
To select a set of gray values in a local area,
Figure 294243DEST_PATH_IMAGE022
The operation is to find the median value of the sequence,
Figure 719670DEST_PATH_IMAGE024
For the specified point where you want to replace the gray value, then:

                        

Figure 243055DEST_PATH_IMAGE026
                        
Figure 243055DEST_PATH_IMAGE026

即得到更加清晰的灰度图像，本图例取滤波单元为3*3的正方形区域进行滤波； That is, a clearer grayscale image is obtained. In this example, a square area with a filter unit of 3*3 is used for filtering;

步骤3）抖动消除 Step 3) Jitter Elimination

采用膨胀腐蚀方法对血型试剂卡灰度图像进行抖动消除，得到铭牌中清晰的平行四边形区域，得到校正后更加清晰的灰度图像； The dilation and corrosion method is used to eliminate the jitter of the gray-scale image of the blood type reagent card, and obtain a clear parallelogram area in the nameplate, and obtain a clearer gray-scale image after correction;

步骤4）倾斜校正 Step 4) Tilt Correction

参照图2所示，采用斜率计算方法得到血型试剂卡的倾斜角度，通过反向旋转相应的角度校正其倾斜状态，血型试剂卡的铭牌中有四个空白平行四边形区域，找出其中的平行四边形区域，取其左上角坐标（

Figure 672900DEST_PATH_IMAGE050
）和左下角坐标（），通过公式：
Figure DEST_PATH_IMAGE051
=
Figure 127332DEST_PATH_IMAGE052
计算出图像倾斜的角度； Referring to Figure 2, use the slope calculation method to obtain the inclination angle of the blood type reagent card, and correct the inclination state by reversely rotating the corresponding angle. There are four blank parallelogram areas on the nameplate of the blood type reagent card, find out the parallelogram in it area, take the coordinates of its upper left corner (
Figure 672900DEST_PATH_IMAGE050
) and the coordinates of the lower left corner ( ), through the formula:
Figure DEST_PATH_IMAGE051
=
Figure 127332DEST_PATH_IMAGE052
Calculate the angle at which the image is tilted;

步骤5）微柱小管分割 Step 5) Microcolumn tubule segmentation

采用模板匹配方法，即通过相关性公式：分割出血型试剂卡中用于图像识别的有效区域，即微柱小管，具体操作过程为：把图像的二维分布简化为二维数组存储数据的形式。假设模板图像的像素信息存储在数组

Figure 957753DEST_PATH_IMAGE054
中，其中数组各行中的数据依次对应的是模板图像各行的像素值。同理，假设原始图像的像素信息存储在数组
Figure 292920DEST_PATH_IMAGE056
中。其中，
Figure 21841DEST_PATH_IMAGE058
，，即模板图像的尺寸一定要小于或者等于原始图像的尺寸，首先，求模板图像的自相关值，其计算公式为： The template matching method is used, that is, through the correlation formula: segment the effective area for image recognition in the blood type reagent card, that is, the microcolumn tubule, and the specific operation process is: simplify the two-dimensional distribution of the image into the form of two-dimensional array storage data . Suppose the pixel information of the template image is stored in the array
Figure 957753DEST_PATH_IMAGE054
, where the data in each row of the array corresponds to the pixel values in each row of the template image. Similarly, suppose the pixel information of the original image is stored in the array
Figure 292920DEST_PATH_IMAGE056
middle. in,
Figure 21841DEST_PATH_IMAGE058
, , that is, the size of the template image must be smaller than or equal to the size of the original image. First, find the autocorrelation value of the template image , whose calculation formula is:

              

Figure 42384DEST_PATH_IMAGE064
                                  
Figure 42384DEST_PATH_IMAGE064
                   

式中，为数组

Figure 826593DEST_PATH_IMAGE054
中行为
Figure 179077DEST_PATH_IMAGE068
、列为
Figure 59308DEST_PATH_IMAGE070
上的像素值，因为是二值化图像，所以
Figure 497243DEST_PATH_IMAGE066
、的值都只能为0或1，接着，从左到右、从上到下依次扫描原始图像中左上角坐标为（
Figure 577380DEST_PATH_IMAGE074
）
Figure 425250DEST_PATH_IMAGE076
，尺寸大小为
Figure 350481DEST_PATH_IMAGE078
，即与模板图像大小相同的局部图像区域
Figure 512472DEST_PATH_IMAGE080
，计算该局部区域的自相关值
Figure 839548DEST_PATH_IMAGE082
和与模板图像的互相关值
Figure 123899DEST_PATH_IMAGE084
。 In the formula, for the array
Figure 826593DEST_PATH_IMAGE054
Bank of China
Figure 179077DEST_PATH_IMAGE068
, listed as
Figure 59308DEST_PATH_IMAGE070
The pixel value on , because it is a binarized image, so
Figure 497243DEST_PATH_IMAGE066
, The value of can only be 0 or 1, and then, scan from left to right, from top to bottom in order to scan the coordinates of the upper left corner of the original image (
Figure 577380DEST_PATH_IMAGE074
)
Figure 425250DEST_PATH_IMAGE076
, the size of which is
Figure 350481DEST_PATH_IMAGE078
, that is, the local image region with the same size as the template image
Figure 512472DEST_PATH_IMAGE080
, to calculate the autocorrelation value of the local area
Figure 839548DEST_PATH_IMAGE082
and cross-correlation values with the template image
Figure 123899DEST_PATH_IMAGE084
.

                

Figure 224841DEST_PATH_IMAGE086
                                     
Figure 224841DEST_PATH_IMAGE086
                    

               

Figure 987261DEST_PATH_IMAGE088
                             
Figure 987261DEST_PATH_IMAGE088
             

式中，为数组中行为

Figure 461602DEST_PATH_IMAGE068
、列为
Figure 27712DEST_PATH_IMAGE070
上的像素值，因此，本次扫描的互相关系数为： In the formula, for the array Bank of China
Figure 461602DEST_PATH_IMAGE068
, listed as
Figure 27712DEST_PATH_IMAGE070
The pixel value on , therefore, the cross-correlation coefficient of this scan for:

         

Figure 876906DEST_PATH_IMAGE096
 ，再比较每次得到的互相关系数
Figure DEST_PATH_IMAGE097
，求出最大互相关系数所对应的当前扫描坐标（
Figure 201708DEST_PATH_IMAGE074
）即为与模板图像相匹配的局部图像左上角坐标，在原始图像中与模板图像相匹配的局部图像区域的右上角坐标为（
Figure DEST_PATH_IMAGE099
），左下角坐标为（），右下角坐标为（
Figure DEST_PATH_IMAGE103
），参照图3所示，采用三种匹配模板中的第二种对灰度图像进行模板匹配，找到微柱小管的有效区域进行分割。其中，原始灰度图像的大小为768*576，第二种匹配模板的大小为118*268，进行一次模板匹配运算即能找到血型试剂卡中的六个微柱小管，并把它们分割出来，匹配运算所需时间为3. 167s；
Figure 876906DEST_PATH_IMAGE096
, and then compare the cross-correlation coefficient obtained each time
Figure DEST_PATH_IMAGE097
, find the current scanning coordinate corresponding to the maximum cross-correlation coefficient (
Figure 201708DEST_PATH_IMAGE074
) is the coordinates of the upper left corner of the local image matching the template image, and the coordinates of the upper right corner of the local image area matching the template image in the original image are (
Figure DEST_PATH_IMAGE099
), the coordinates of the lower left corner are ( ), the coordinates of the lower right corner are (
Figure DEST_PATH_IMAGE103
), as shown in Figure 3, the gray-scale image is template-matched using the second of the three matching templates, and the effective area of the microcolumn tubule is found for segmentation. Among them, the size of the original grayscale image is 768*576, the size of the second matching template is 118*268, and the six microcolumn tubes in the blood type reagent card can be found and separated out by performing a template matching operation. The time required for the matching operation is 3.167s;

步骤6）阈值处理 Step 6) Thresholding

采用二值化方法对分割出的微柱小管图像进行阈值处理，把微柱小管图像由灰度图像转换成为只有黑白两种颜色的二值化图像，设定某一阈值

Figure 821494DEST_PATH_IMAGE037
，将原始灰度图像各像素的像素值
Figure 977669DEST_PATH_IMAGE010
和阈值作比较，大于
Figure 321242DEST_PATH_IMAGE038
的重新取值为255，反之则为0，设阈值处理后的像素值为，即： Use the binarization method to perform threshold value processing on the segmented microcolumn tubule image, convert the microcolumn tubule image from a grayscale image to a binary image with only black and white colors, and set a certain threshold
Figure 821494DEST_PATH_IMAGE037
, the pixel value of each pixel of the original grayscale image
Figure 977669DEST_PATH_IMAGE010
and threshold for comparison, greater than
Figure 321242DEST_PATH_IMAGE038
The re-value of is 255, otherwise it is 0, and the pixel value after thresholding is set to be ,Right now:

           

Figure 660323DEST_PATH_IMAGE044
 参照图4所示取阈值为80，对分割出的微柱小管图像进行二值化处理，得到只有黑白两种颜色的二值化图像；
Figure 660323DEST_PATH_IMAGE044
With reference to Figure 4, the threshold value is 80, and the segmented microcolumn tubule image is binarized to obtain a binarized image with only two colors, black and white;

步骤7）灰度分析 Step 7) Grayscale Analysis

采用像素求和方法把每个微柱小管的下端分成上、中、下三个小区域，分别计算上、中、下三个小区域的像素之和，即用像素的分布情况代表微柱小管下端的红细胞分布情况； The lower end of each microcolumn tubule is divided into upper, middle, and lower small areas by pixel summation method, and the sum of pixels in the upper, middle, and lower three small areas is calculated respectively, that is, the distribution of pixels is used to represent the microcolumn tubule The distribution of red blood cells at the lower end;

步骤8）种类识别 Step 8) Type Identification

采用颜色识别和字符识别结合的方法，通过分析血型试剂卡铭牌中不同颜色区域中的字符差异来识别其种类的不同，参照图5所示，常见的血型试剂卡的铭牌有七种，其颜色和字符有很大的差异，首先，采用灰度变换方法，依次取

Figure 140163DEST_PATH_IMAGE048
、
Figure 270930DEST_PATH_IMAGE049
、
Figure 409787DEST_PATH_IMAGE018
为100、010、001对血型试剂卡的铭牌进行单色处理，得到不同种类铭牌的单色分布情况，其次，参照图6所示，采用字符区域分割的方法，提取出各个不同单色区域中的字符，然后参照图7所示，采用11特征点法截取单个字符的11个区域T1-T11，计算每个区域中的像素之和Tsum，构成一个像素和向量（T1sum，T2sum，…，T11sum），根据像素和向量的差异识别不同的字符，进而得出所对应血型试剂卡的种类； The method of combining color recognition and character recognition is used to identify the difference of the type by analyzing the character difference in the different color areas of the nameplate of the blood type reagent card. Referring to Figure 5, there are seven common nameplates of the blood type reagent card, and their colors It is very different from characters. First, adopt the grayscale transformation method, and take
Figure 140163DEST_PATH_IMAGE048
,
Figure 270930DEST_PATH_IMAGE049
,
Figure 409787DEST_PATH_IMAGE018
Perform monochrome processing on the nameplates of blood type reagent cards for 100, 010, and 001 to obtain the monochrome distribution of different types of nameplates. Secondly, as shown in Figure 6, use the method of character area segmentation to extract the different monochrome areas. character, then as shown in Figure 7, using the 11 feature point method to intercept 11 regions T1-T11 of a single character, calculate the sum Tsum of pixels in each region, and form a pixel and vector (T1sum, T2sum, ..., T11sum ), identify different characters according to the difference between pixels and vectors, and then obtain the type of the corresponding blood type reagent card;

步骤9）对比检测结果 Step 9) Compare the detection results

据微柱小管中红细胞的分布情况，同时结合当前血型试剂卡的种类，得到该血型试剂卡的血型检测结果。  According to the distribution of red blood cells in the microcolumn tubules, combined with the type of the current blood type reagent card, the blood type detection result of the blood type reagent card is obtained. the

以上所述仅为本发明的优选实施例而已，并不用于限制本发明，对于本领域的技术人员来说，本发明可以有各种更改和变化。凡在本发明的精神和原则之内，所作的任何修改、等同替换、改进等，均应包含在本发明的保护范围之内。  The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention. the

Claims (1)

1. A blood type result recognition algorithm based on gray level analysis and type recognition is characterized by comprising the following steps:
step 1) Gray level conversion
Carrying out gray scale transformation on 24-bit true color images of the blood type reagent card by adopting a weighted average method, and setting coordinates (A and B)) The brightness values of the respective components are respectively
Figure 866161DEST_PATH_IMAGE002
、
Figure DEST_PATH_IMAGE003
、
Figure 500054DEST_PATH_IMAGE004
The gray value of the point after the gray conversion
Figure DEST_PATH_IMAGE005
，
Figure 875672DEST_PATH_IMAGE006
Wherein
Figure DEST_PATH_IMAGE007
、
Figure 296289DEST_PATH_IMAGE008
、
Figure DEST_PATH_IMAGE009
respectively weighting each color component to obtain a converted gray level image;
step 2) smoothing filtering
Noise filtering is carried out on the gray level image of the blood type reagent card by adopting a median filtering method
Figure 955810DEST_PATH_IMAGE010
To select a set of gray values in a local region,
Figure DEST_PATH_IMAGE011
the operation is to find the median value of the sequence,
Figure 6942DEST_PATH_IMAGE012
to want to change the gray scaleA specified point of value, then:
Figure DEST_PATH_IMAGE013
a clearer gray level image is obtained;
step 3) jitter elimination
The gray image of the blood type reagent card is shaken and eliminated by adopting an expansion corrosion method, so that a clearer gray image is obtained;
step 4) Tilt correction
Obtaining the inclination angle of the blood type reagent card by adopting a slope calculation method, correcting the inclination state of the blood type reagent card by reversely rotating the corresponding angle, finding out a parallelogram area in four blank parallelogram areas in a nameplate of the blood type reagent card, and taking the upper left corner coordinate of the parallelogram area (the step (c) () And coordinates of lower left corner: (
Figure DEST_PATH_IMAGE015
) By the formula:
Figure 844896DEST_PATH_IMAGE016
=
Figure DEST_PATH_IMAGE017
calculating the inclination angle of the image;
Figure 844076DEST_PATH_IMAGE018
step 5) micro-column tubule segmentation
Adopting a template matching method, namely, through a correlation formula: segmenting an effective area for image identification in the blood type reagent card, namely a microcolumn tubule;
step 6) threshold processing
Performing threshold processing on the segmented microcolumn tubule image by using a binarization methodConverting the image of the microcolumn tubule from a gray image to a binary image with only black and white colors, and setting a certain threshold
Figure DEST_PATH_IMAGE019
The pixel value of each pixel of the original gray imageAnd a threshold value
Figure 727905DEST_PATH_IMAGE020
By comparison, is greater than
Figure 755903DEST_PATH_IMAGE020
Is 255, otherwise is 0, and the pixel value after the threshold processing is set as
Figure DEST_PATH_IMAGE021
Namely:
Figure DEST_PATH_IMAGE023
Figure 64842DEST_PATH_IMAGE024
step 7) Gray level analysis
Dividing the lower end of each micro-column tubule into an upper small area, a middle small area and a lower small area by adopting a pixel summation method, respectively calculating the sum of pixels of the upper small area, the middle small area and the lower small area, namely representing the distribution condition of red blood cells at the lower end of the micro-column tubule by the distribution condition of the pixels;
step 8) species identification
Identifying the difference of the types of the blood type reagent card by analyzing the character difference in different color areas in the nameplate of the blood type reagent card by adopting a method combining color identification and character identification, and 9) comparing the detection results
And according to the distribution condition of the red blood cells in the microcolumn tubules, simultaneously combining the types of the current blood type reagent cards to obtain the blood type detection result of the blood type reagent card.
CN201310416614.3A 2013-09-13 2013-09-13 A kind of blood group result recognizer based on gray analysis and category identification Active CN103543277B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201310416614.3A CN103543277B (en) 2013-09-13 2013-09-13 A kind of blood group result recognizer based on gray analysis and category identification

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201310416614.3A CN103543277B (en) 2013-09-13 2013-09-13 A kind of blood group result recognizer based on gray analysis and category identification

Publications (2)

Publication Number Publication Date
CN103543277A true CN103543277A (en) 2014-01-29
CN103543277B CN103543277B (en) 2016-03-16

Family

ID=49966933

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201310416614.3A Active CN103543277B (en) 2013-09-13 2013-09-13 A kind of blood group result recognizer based on gray analysis and category identification

Country Status (1)

Country Link
CN (1) CN103543277B (en)

Cited By (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104198355A (en) * 2014-07-16 2014-12-10 电子科技大学 Automatic detection method for red cells in feces
CN105403714A (en) * 2015-10-21 2016-03-16 中山市生科试剂仪器有限公司 A software control method for plotter
CN105550507A (en) * 2015-12-09 2016-05-04 上海斐讯数据通信技术有限公司 Intelligent terminal for measuring blood type and measuring method
CN105678752A (en) * 2015-12-30 2016-06-15 合肥天一生物技术研究所 Blood type identification method through segmentation of microcolumn vasculum based on fixed parameters
CN106199014A (en) * 2016-06-29 2016-12-07 中山生物工程有限公司 A Blood Type Identification Method Based on Blood Type Card
CN106372596A (en) * 2016-08-30 2017-02-01 孟玲 Biological information collection device
CN107167616A (en) * 2017-03-13 2017-09-15 中国科学院苏州生物医学工程技术研究所 Blood type testing methods and device
CN107356774A (en) * 2017-06-28 2017-11-17 苏州长光华医生物医学工程有限公司 Micro-column gel card aggegation testing result recognition methods
CN107389957A (en) * 2017-06-28 2017-11-24 苏州长光华医生物医学工程有限公司 Micro-column gel card aggegation testing result identifying system and blood type analytical instrument
CN109117764A (en) * 2018-07-29 2019-01-01 国网上海市电力公司 Using the method for color threshold method identification target object region electrical symbol in power monitoring
CN109409395A (en) * 2018-07-29 2019-03-01 国网上海市电力公司 Using the method for template matching method identification target object region electrical symbol in power monitoring
CN109919054A (en) * 2019-02-25 2019-06-21 电子科技大学 A detection method for automatic classification of reagent cards based on machine vision
CN110794151A (en) * 2019-10-30 2020-02-14 中山生物工程有限公司 Method, device, terminal, medium and equipment for detecting quality of blood type card
CN112233082A (en) * 2020-10-13 2021-01-15 深圳市瑞沃德生命科技有限公司 Automatic exposure method and device for cell image
CN112485457A (en) * 2020-11-18 2021-03-12 天津德祥生物技术有限公司 Detection result identification method
CN112907665A (en) * 2021-02-07 2021-06-04 吉林大学 Micro-fluidic blood type detection card micro-cavity reaction tank precise positioning method based on RGB color space

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0864858A2 (en) * 1997-03-13 1998-09-16 Ortho-Clinical Diagnostics, Inc. Method and apparatus for calibrating imaging systems for analyzing agglutination reactions
CN2821572Y (en) * 2003-09-28 2006-09-27 张传国 Blood type analyzer
CN101765773A (en) * 2007-06-29 2010-06-30 贝克曼考尔特公司 Agglutination image automatic judging method by MT system, device, program, and recording medium
CN102446353A (en) * 2010-09-30 2012-05-09 广州阳普医疗科技股份有限公司 Machine vision interpretation method and device for blood type analysis

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0864858A2 (en) * 1997-03-13 1998-09-16 Ortho-Clinical Diagnostics, Inc. Method and apparatus for calibrating imaging systems for analyzing agglutination reactions
CN2821572Y (en) * 2003-09-28 2006-09-27 张传国 Blood type analyzer
CN101765773A (en) * 2007-06-29 2010-06-30 贝克曼考尔特公司 Agglutination image automatic judging method by MT system, device, program, and recording medium
CN102446353A (en) * 2010-09-30 2012-05-09 广州阳普医疗科技股份有限公司 Machine vision interpretation method and device for blood type analysis

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
罗刚银: "全自动血型分析系统关键技术的研究", 《中国博士学位论文全文数据库 医药卫生科技辑》, no. 9, 15 September 2012 (2012-09-15), pages 053 - 19 *

Cited By (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104198355A (en) * 2014-07-16 2014-12-10 电子科技大学 Automatic detection method for red cells in feces
CN105403714A (en) * 2015-10-21 2016-03-16 中山市生科试剂仪器有限公司 A software control method for plotter
CN105550507A (en) * 2015-12-09 2016-05-04 上海斐讯数据通信技术有限公司 Intelligent terminal for measuring blood type and measuring method
CN105678752A (en) * 2015-12-30 2016-06-15 合肥天一生物技术研究所 Blood type identification method through segmentation of microcolumn vasculum based on fixed parameters
CN106199014A (en) * 2016-06-29 2016-12-07 中山生物工程有限公司 A Blood Type Identification Method Based on Blood Type Card
CN106372596A (en) * 2016-08-30 2017-02-01 孟玲 Biological information collection device
CN107167616A (en) * 2017-03-13 2017-09-15 中国科学院苏州生物医学工程技术研究所 Blood type testing methods and device
CN107167616B (en) * 2017-03-13 2019-06-07 中国科学院苏州生物医学工程技术研究所 Blood type testing methods and device
WO2019000740A1 (en) * 2017-06-28 2019-01-03 苏州长光华医生物医学工程有限公司 Micro-column gel card agglutination test result identification system and blood grouping analyzer
CN107389957A (en) * 2017-06-28 2017-11-24 苏州长光华医生物医学工程有限公司 Micro-column gel card aggegation testing result identifying system and blood type analytical instrument
WO2019000741A1 (en) * 2017-06-28 2019-01-03 苏州长光华医生物医学工程有限公司 Method of identifying micro-column gel card agglutination test result
CN107356774A (en) * 2017-06-28 2017-11-17 苏州长光华医生物医学工程有限公司 Micro-column gel card aggegation testing result recognition methods
CN109117764A (en) * 2018-07-29 2019-01-01 国网上海市电力公司 Using the method for color threshold method identification target object region electrical symbol in power monitoring
CN109409395A (en) * 2018-07-29 2019-03-01 国网上海市电力公司 Using the method for template matching method identification target object region electrical symbol in power monitoring
CN109919054A (en) * 2019-02-25 2019-06-21 电子科技大学 A detection method for automatic classification of reagent cards based on machine vision
CN109919054B (en) * 2019-02-25 2023-04-07 电子科技大学 Machine vision-based reagent card automatic classification detection method
CN110794151A (en) * 2019-10-30 2020-02-14 中山生物工程有限公司 Method, device, terminal, medium and equipment for detecting quality of blood type card
CN112233082A (en) * 2020-10-13 2021-01-15 深圳市瑞沃德生命科技有限公司 Automatic exposure method and device for cell image
CN112485457A (en) * 2020-11-18 2021-03-12 天津德祥生物技术有限公司 Detection result identification method
CN112907665A (en) * 2021-02-07 2021-06-04 吉林大学 Micro-fluidic blood type detection card micro-cavity reaction tank precise positioning method based on RGB color space

Also Published As

Publication number Publication date
CN103543277B (en) 2016-03-16

Similar Documents

Publication Publication Date Title
CN103543277B (en) A kind of blood group result recognizer based on gray analysis and category identification
JP6710135B2 (en) Cell image automatic analysis method and system
JP4071186B2 (en) Method and system for identifying an object of interest in a biological specimen
WO2016062159A1 (en) Image matching method and platform for testing of mobile phone applications
CN110490836B (en) dPCR microarray image information processing method
WO2021030952A1 (en) Base recognition method and system, computer program product, and sequencing system
CN108961301B (en) A method for image segmentation of Chaetoceros sp. based on unsupervised pixel-by-pixel classification
JP6733983B2 (en) Image analysis device
CN110020692A (en) A kind of handwritten form separation and localization method based on block letter template
CN110866932A (en) Multi-channel tongue edge detection device and method and storage medium
WO2019181072A1 (en) Image processing method, computer program, and recording medium
WO2020037573A1 (en) Method and device for detecting bright spots on image, and computer program product
CN112289377B (en) Method, apparatus and computer program product for detecting bright spots on an image
CN117576121A (en) Automatic segmentation method, system, equipment and medium for microscope scanning area
CN112215865A (en) Automatic detection method for micro-droplets under fluorescence microscopic image
Marcuzzo et al. Automated Arabidopsis plant root cell segmentation based on SVM classification and region merging
CN111210447B (en) A method and terminal for hierarchical segmentation of hematoxylin-eosin stained pathological images
CN112289381B (en) Method, device and computer product for constructing sequencing template based on image
CN108877030B (en) Image processing method, device, terminal and computer readable storage medium
WO2020037570A1 (en) Method and device for image registration, and computer program product
CN110298347B (en) Method for identifying automobile exhaust analyzer screen based on GrayWorld and PCA-CNN
CN112288781B (en) Image registration method, apparatus and computer program product
WO2020037574A1 (en) Method for constructing sequencing template based on image, and base recognition method and device
WO2020037571A1 (en) Method and apparatus for building sequencing template on basis of images, and computer program product
CN115917594A (en) Entire slide annotation transfer using geometric features

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
C53 Correction of patent of invention or patent application
CB03 Change of inventor or designer information

Inventor after: Luo Gangyin

Inventor after: Wang Bidou

Inventor after: Chen Yueyan

Inventor after: Zhang Yunping

Inventor before: Luo Gangyin

COR Change of bibliographic data

Free format text: CORRECT: INVENTOR; FROM: LUO GANGYIN TO: LUO GANGYIN WANG BIDOU CHEN YUEYAN ZHANG YUNPING

C14 Grant of patent or utility model
GR01 Patent grant