CN103426155A - Column diagram boundary dividing method based on solving column diagram changing rate - Google Patents
Column diagram boundary dividing method based on solving column diagram changing rate Download PDFInfo
- Publication number
- CN103426155A CN103426155A CN2012101511556A CN201210151155A CN103426155A CN 103426155 A CN103426155 A CN 103426155A CN 2012101511556 A CN2012101511556 A CN 2012101511556A CN 201210151155 A CN201210151155 A CN 201210151155A CN 103426155 A CN103426155 A CN 103426155A
- Authority
- CN
- China
- Prior art keywords
- peak
- histogram
- change
- rate
- point
- 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.)
- Pending
Links
Images
Abstract
A column diagram boundary dividing method based on solving column diagram changing rate comprises the steps of starting from the right/left side of a column diagram, searching a first point whose changing rate jumps to be larger than or equal to zero from the changing rate smaller than zero towards the left/right of the column diagram to be determined as the peak value of a right peak/left peak, and continuously searching a first point whose changing rate jumps to be larger than or equal to zero from the changing rate smaller than zero towards the left/right of the column diagram to be determined as the peak value of a left peak/right peak; searching a first temporary zero crossing point whose changing rate jumps to be smaller than or equal to zero from the changing rate larger than zero from the peak value of the right peak/left peak towards the left/right of the column diagram; searching a first zero crossing point whose changing rate jumps to be smaller than zero from the changing rate larger than or equal to zero from the temporary zero crossing point towards the left/right of the column diagram, and searching a second zero crossing point whose changing rater jumps to be larger than zero from the changing rate smaller than or equal to zero between the first zero passing point and the temporary zero crossing point; averaging the first zero crossing point and the second zero crossing point, and determining the average value as the boundary dividing point of the left peak and the right peak. The column diagram boundary dividing method based on solving the column diagram changing rate can enable a column diagram to be provided with a stable and accurate grouping result, and is simple and easy to implement.
Description
Technical field
The present invention relates to the histogrammic treatment technology of a kind of blood cell analyzer, especially relate to a kind of histogram boundary method based on asking the histogram rate of change.
Background technology
In blood medical science is detected, need statistics blood cell quantity, usually can utilize histogram as auxiliary.Histogram is a kind of statistical report figure, and the longitudinal stripe do not waited by a series of height or line segment mean the situation that data distribute.Generally with transverse axis, mean data characteristics, the longitudinal axis means distribution situation.
Usually, in order to obtain required result, need to the information of needs be extracted by histogram, typical example is exactly by define boundaries in histogram (being generally " valley point "), to divide different group (shape that usually presents peak).
The white blood cell count(WBC) of blood cell analyzer of take is example, while by Ku Ertefa, measuring the quantity of leucocyte in blood cell, and can be by obtained pulse statistics on histogram, as shown in Figure 1.The size that histogrammic transverse axis is pulse (characterizing the size of particle), the quantity that the longitudinal axis is pulse.These pulses will mainly present two peaks on histogram: leucocyte and blood shadow (chip that red blood cell produces after by the hemolytic agent effect is less than leucocyte usually on volume).By determining histogrammic valley point (being the decomposition point at Zuo Feng and right peak), can distinguish above-mentioned two particles and hive off.
In fact, because histogrammic statistical information is subject to many factors, disturb, the histogram presented is varied, often interferes with the searching of valley.The Major Difficulties of below demarcating for histogram:
(1) the histogram fluctuation that, undesired signal (as noise signal) causes.Undesired signal makes tiny " the pseudo-peak " and " pseudo-paddy " of appearance on histogram, thereby affect histogrammic boundary, calculates.As shown in Figure 2, fluctuation has also appearred in the position of valley point, and true valley point disappears, and local minimum point is not two peak-to-peak valley points.
(2), peak height is lower, even there is no peak.Sample difference with statistics, there will be the wherein lower situation of peak value at a peak (not consider the bimodal situation that is low value sometimes on histogram.Because now add up peak, can't separate with interference range, these type of data need to be considered to process by the method beyond histogram).In the time of this, the peak value at this peak and two peak-to-peak valleies differ less, and the histogram wave zone that is difficult to cause with other interference separates.When two peaks approach, due to the synergistic effect at two peaks, the situation that a peak disappears even appears.Now by searching of local minimum, also can't obtain two peak-to-peak separations, as shown in Figure 3.
(3), for situation about mentioning in (2), usually can be by every bit on compute histograms and wait thereafter the difference (characterizing histogrammic rate of change on this aspect) between the point of step-length to obtain histogrammic change information, thus two peak-to-peak separations found.This concept is equivalent to a continuous function curve differentiate.
The derivative of continuous function is:
As enough hour of Δ x, can think that f ' (x) is approximately equal to Δ y/ Δ x.This concept is introduced in discrete histogram, can be obtained histogrammic change rate curve.In a desirable continuous function curve, the derivative of peak and paddy is 0.Although histogram is comprised of discrete point, the rate of change of valley point and peak point can be thought close to 0.Therefore generally can pass through mistake 0 point (from negative value to the point on the occasion of saltus step) of histogram rate of change, determine the position of valley point, see that the single line bar that in Fig. 4 A(Fig. 4 A, go up position is histogram curve, position is histogrammic rate of change than next root lines).But in fact, very responsive for histogrammic subtle change to histogram changes persuing rate, see that the single line bar that in Fig. 4 B(Fig. 4 B, go up position is histogram curve, position is histogrammic rate of change than next root lines).Because the interference of mentioning in (1) exists, by the method for changes persuing rate, determine that there is larger difficulty valley point.
(4), another difficult point by histogram rate of change calculating valley point is: in the face of " paddy " is located to put down, be that the paddy place is while existing the histogram of continuous equivalent point, be difficult to obtain valley point accurately, (the single line bar that in Fig. 5, go up position is histogram, and following single line bar is histogrammic rate of change) as shown in Figure 5.
Summary of the invention
In order to solve the above-mentioned interference that may run in histogram boundary, the present invention proposes a kind of histogram boundary method based on asking the histogram rate of change to cellanalyzer, and stable, grouping result accurately can be provided.
The present invention adopts following technical scheme to realize: a kind of histogram boundary method based on asking the histogram rate of change, and it comprises step:
Suppose the peak that in histogram, the right/left peak is stable existence, from the right/left side of stable peaks, search first rate of change from being less than 0 saltus step to the point that is more than or equal to 0 to left/right, determine that this point is for the peak at /Zuo peak, the right peak of histogram, continuation is searched first rate of change from being less than 0 saltus step to the point that is more than or equal to 0 to left/right, this point is defined as to the peak value at /You peak, histogrammic left peak;
From the peak value at You Feng/left peak, on histogram to left/right search rate of change first from being greater than 0 saltus step to the interim zero crossing that is less than or equal to 0;
From interim zero crossing, on histogram, continue to left/right search rate of change first from being more than or equal to 0 saltus step to the first zero crossing that is less than 0; Since the first zero crossing, to right/left, between the first zero crossing and interim zero crossing, search first from being less than or equal to 0 saltus step to the second zero crossing that is greater than 0;
The mean value of getting the first zero crossing and the second zero crossing is defined as the separation at He You peak, the left peak of histogram.
Wherein, described method also comprises step: the change rate curve of compute histograms.
Wherein, described method also comprises step: change rate curve is carried out to smoothing processing.
Wherein, described method also comprises step: before the change rate curve of compute histograms, and the smoothing processing histogram.
Wherein, if continue, to left/right, search first rate of change when being less than 0 saltus step to the point that is more than or equal to 0, can't find this point, the point of getting histogram high order end/low order end is defined as the peak value at histogrammic left peak.
Compared with prior art, the present invention has following beneficial effect:
The present invention proposes to determine histogrammic boundary (determining the left peak of histogram and right peak-to-peak " valley point " position) by the rate of change of compute histograms, makes that histogram can provide stable, the method for grouping result, and boundary accurately is simple, is easy to realization.
The accompanying drawing explanation
Fig. 1 is the histogram of the pulse statistics obtained while by Ku Ertefa, measuring the quantity of leucocyte in blood cell;
Fig. 2 the histogram of fluctuation occurs because undesired signal causes the position, valley point;
Fig. 3 is the low histogram that causes obtaining two peak-to-peak separations of a crest height;
Fig. 4 A and Fig. 4 B are all schematic diagram of histogram and rate of change thereof;
Fig. 5 is the histogram that there is continuous equivalent point in the paddy place;
Fig. 6 is schematic flow sheet of the present invention;
Fig. 7 A and Fig. 7 B be the histogram of smoothing processing front and back respectively;
Fig. 8 is the change rate curve schematic diagram of Fig. 7 B;
Fig. 9 is the schematic diagram of determining histogrammic right peak-to-peak value position;
Figure 10 A and Figure 10 B determine the schematic diagram of histogrammic left peak-to-peak value position in two kinds of situations;
Figure 11 determines the position view of interim zero crossing on histogram;
Figure 12 determines the position view of the first zero crossing on histogram;
Figure 13 determines the position view of the second zero crossing on histogram;
Figure 14 determines the position view of separation on histogram.
Embodiment
In histogram in view of cellanalyzer, different peaks represents different populations of cells, therefore, only have accurately demarcated between histogrammic cutting edge of a knife or a sword and peak (determining adjacent two peak-to-peak " valley points "), stable, grouping result accurately can be provided.
Shown in Fig. 6, the present invention proposes a kind of histogram boundary method based on asking the histogram rate of change, and stable, grouping result accurately can be provided.Specifically, the present invention includes following performing step:
Can selected frequency field smothing filtering, the multiple smoothing processing method such as medium filtering.Level and smooth degree can be chosen according to histogrammic feature and the needed result of required processing.As shown in Fig. 7 A and Fig. 7 B, the histogram before and after the difference smoothing processing.
Histogrammic change rate curve after step S2, calculating smoothly, and change rate curve is carried out to smoothing processing.
Such as, Fig. 8 is the change rate curve schematic diagram of Fig. 7 B.The computing method of histogrammic change rate curve as previously mentioned, can be introduced the Method of Seeking Derivative of continuous function in discrete histogram.Such as a definite Δ x, the rate of change of each point of histogram only and histogrammic y direction of principal axis increment Delta y=f (x0+ Δ x)-f (x0) is relevant.The H for value (i) that arbitrfary point on histogram is set means, histogrammic change rate curve H ' (i)=H (i+k) – H (i).Wherein H (i) means the value at i some place on the histogram transverse axis; K means changes persuing rate step-length, and it is higher that k is equivalent to more greatly level and smooth degree.
Be more than a kind of method that can be used for asking histogrammic change rate curve, but be not limited to this method.
The peak that in step S3, supposition histogram, the right/left peak is stable existence, from the right side of histogram stable peaks, the rightmost side of desirable histogrammic change rate curve, search first left from being less than 0 saltus step to the point that is more than or equal to 0, search first rate of change left from being less than 0 saltus step to the point that is more than or equal to 0, think that this puts the peak at the right peak of corresponding histogram.
In histogrammic two peaks of supposing to need to demarcate, there is a peak always to exist, and there is obvious peak type (if bimodal all without the stable peaks type, this sample is not suitable for being calculated by histogram, not in the discussion scope of the method), and suppose that ，Ze Cong right side, Zong Zai right side, this peak starts, on histogrammic change rate curve, search first left from being less than 0 saltus step to the point that is more than or equal to 0, think that this point is for the peak at right peak, as shown in Figure 9.
In fact, as described in background technology, searching peak point by the rate of change chart also may have certain deviation.But just as the pilot process that calculates valley point, do not need accurately to obtain the position of peak point due to this peak point.
Step S4, on histogrammic change rate curve, continue to search next from being less than 0 saltus step to the point that is more than or equal to 0 left, continue to search a rate of change left from being less than 0 saltus step to the point that is more than or equal to 0, this point is defined as to the peak value at histogrammic left peak, as shown in Figure 10 A.Even left peak does not exist, this point is got histogrammic high order end, as shown in Figure 10 B.Wherein, in Figure 10 A and Figure 10 B, the single line bar that go up position is histogram curve, and position is histogrammic rate of change than next root lines, and the vertical line on the left side means the peak at left peak, and the vertical line on the right means the peak at right peak.
Step S5, from the peak value at right peak, on histogrammic change rate curve, search rate of change left from being greater than 0 saltus step to the point that is less than or equal to 0, be designated as interim zero crossing, as the vertical line position of Figure 11.The single line bar that in Figure 11, go up position is histogram curve, and position is histogrammic rate of change than next root lines.
Step S6, from interim zero crossing, on histogrammic change rate curve, continue to search first left from being more than or equal to 0 saltus step to the point that is less than 0, be designated as the first zero crossing, vertical line position as shown in figure 12.The single line bar that in Figure 12, go up position is histogram curve, and the single line bar of position under is histogrammic rate of change.
Step S7, since the first zero crossing, on the histogrammic change rate curve between the first zero crossing and interim zero crossing, search first from being less than or equal to 0 saltus step to the point that is greater than 0, be designated as the second zero crossing, vertical line position as shown in figure 13.The single line bar that in Figure 13, go up position is histogram curve, and the single line bar of position under is histogrammic rate of change.
Step S8, get the mean value of the first zero crossing and the second zero crossing, on histogram curve to point (the horizontal ordinate of this point equals described mean value) that should mean value the separation as Zuo Feng and right peak, as shown in figure 14.Can see, this method has accurately obtained Zuo Feng and right peak-to-peak " valley point " position.Wherein, the single line bar that in Figure 14, go up position is histogram, and position is histogrammic rate of change than next root lines.
Therefore, the present invention proposes to determine histogrammic boundary (determining the left peak of histogram and peak-to-peak " valley point " position, the right side) by the rate of change of compute histograms, makes histogram that stable, grouping result accurately can be provided.
Need supplementary notes, above description is the histogram with 2 peaks, and starts to be processed from the histogrammic rightmost side.Certainly, this case also can start to be processed from the histogrammic leftmost side, first finds left peak-to-peak value position.Utilize the inventive method, can be to thering is the processing of demarcating of 2 histograms with the superiors.
The foregoing is only preferred embodiment of the present invention, not in order to limit the present invention, all any modifications of doing within the spirit and principles in the present invention, be equal to and replace and improvement etc., within all should being included in protection scope of the present invention.
Claims (5)
1. the histogram boundary method based on asking the histogram rate of change, is characterized in that, described method comprises step:
From histogrammic right/left side, search first rate of change from being less than 0 saltus step to the point that is more than or equal to 0 to left/right, determine that this point is for the peak at /Zuo peak, the right peak of histogram, continuation is searched first rate of change from being less than 0 saltus step to the point that is more than or equal to 0 to left/right, this point is defined as to the peak value at /You peak, histogrammic left peak;
From the peak value at You Feng/left peak, on histogram left peak /You search at peak rate of change first from being greater than 0 saltus step to the interim zero crossing that is less than or equal to 0;
From interim zero crossing, on histogram, continue to left/right search rate of change first from being more than or equal to 0 saltus step to the first zero crossing that is less than 0; Since the first zero crossing, to right/left, between the first zero crossing and interim zero crossing, search first from being less than or equal to 0 saltus step to the second zero crossing that is greater than 0;
The mean value of getting the first zero crossing and the second zero crossing is defined as the separation at He You peak, the left peak of histogram.
2. the histogram boundary method based on asking the histogram rate of change according to claim 1, is characterized in that, described method also comprises step: the change rate curve of compute histograms.
3. the histogram boundary method based on asking the histogram rate of change according to claim 2, is characterized in that, described method also comprises step: change rate curve is carried out to smoothing processing.
4. the histogram boundary method based on asking the histogram rate of change according to claim 2, is characterized in that, described method also comprises step: before the change rate curve of compute histograms, and the smoothing processing histogram.
5. according to any one described histogram boundary method based on asking the histogram rate of change of claim 1-4, it is characterized in that, search first rate of change when being less than 0 saltus step to the point that is more than or equal to 0 if continue to left/right, can't find this point, the point of getting histogram high order end/low order end is defined as the peak value at histogrammic left peak.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012101511556A CN103426155A (en) | 2012-05-16 | 2012-05-16 | Column diagram boundary dividing method based on solving column diagram changing rate |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012101511556A CN103426155A (en) | 2012-05-16 | 2012-05-16 | Column diagram boundary dividing method based on solving column diagram changing rate |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103426155A true CN103426155A (en) | 2013-12-04 |
Family
ID=49650849
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2012101511556A Pending CN103426155A (en) | 2012-05-16 | 2012-05-16 | Column diagram boundary dividing method based on solving column diagram changing rate |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103426155A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104730181A (en) * | 2013-12-18 | 2015-06-24 | 北京普源精电科技有限公司 | Chromatographic peak end point adjusting method and chromatographic work station having chromatographic peak end point adjusting function |
| CN105654501A (en) * | 2016-02-22 | 2016-06-08 | 北方工业大学 | Self-adaptive image segmentation method based on fuzzy threshold |
| CN106501160A (en) * | 2016-09-08 | 2017-03-15 | 长春迪瑞医疗科技股份有限公司 | A kind of method for classifying particles and particle classifying device |
| CN106537416A (en) * | 2014-10-31 | 2017-03-22 | 欧姆龙株式会社 | Image processing device, character recognition device, image processing method, and program |
| CN107798681A (en) * | 2016-09-02 | 2018-03-13 | 天津工业大学 | Small object image Fast Threshold dividing method based on mathematical morphology |
| CN109059789A (en) * | 2018-08-21 | 2018-12-21 | 成都天衡智造科技有限公司 | Cable pitch online test method based on machine vision |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030016864A1 (en) * | 2001-07-20 | 2003-01-23 | Mcgee Tom | Methods of and system for detecting a cartoon in a video data stream |
| CN101059870A (en) * | 2007-05-23 | 2007-10-24 | 华中科技大学 | Image cutting method based on attribute histogram |
| CN101710425A (en) * | 2009-12-25 | 2010-05-19 | 南京航空航天大学 | Self-adaptive pre-segmentation method based on gray scale and gradient of image and gray scale statistic histogram |
| CN102184389A (en) * | 2011-05-17 | 2011-09-14 | 浙江大学 | Automatic annual ring identification method used for tree species with high chromatic aberration of heartwood and sapwood |
-
2012
- 2012-05-16 CN CN2012101511556A patent/CN103426155A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030016864A1 (en) * | 2001-07-20 | 2003-01-23 | Mcgee Tom | Methods of and system for detecting a cartoon in a video data stream |
| CN101059870A (en) * | 2007-05-23 | 2007-10-24 | 华中科技大学 | Image cutting method based on attribute histogram |
| CN101710425A (en) * | 2009-12-25 | 2010-05-19 | 南京航空航天大学 | Self-adaptive pre-segmentation method based on gray scale and gradient of image and gray scale statistic histogram |
| CN102184389A (en) * | 2011-05-17 | 2011-09-14 | 浙江大学 | Automatic annual ring identification method used for tree species with high chromatic aberration of heartwood and sapwood |
Non-Patent Citations (1)
| Title |
|---|
| 王海星: "《图像背景自适应分割技术研究》", 《昆明冶金高等专科学校学报》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104730181A (en) * | 2013-12-18 | 2015-06-24 | 北京普源精电科技有限公司 | Chromatographic peak end point adjusting method and chromatographic work station having chromatographic peak end point adjusting function |
| CN104730181B (en) * | 2013-12-18 | 2017-10-24 | 北京普源精电科技有限公司 | Spectral peak terminal method of adjustment and the chromatographic work station that function is adjusted with spectral peak terminal |
| CN106537416A (en) * | 2014-10-31 | 2017-03-22 | 欧姆龙株式会社 | Image processing device, character recognition device, image processing method, and program |
| CN105654501A (en) * | 2016-02-22 | 2016-06-08 | 北方工业大学 | Self-adaptive image segmentation method based on fuzzy threshold |
| CN105654501B (en) * | 2016-02-22 | 2019-07-09 | 北方工业大学 | Self-adaptive image segmentation method based on fuzzy threshold |
| CN107798681A (en) * | 2016-09-02 | 2018-03-13 | 天津工业大学 | Small object image Fast Threshold dividing method based on mathematical morphology |
| CN107798681B (en) * | 2016-09-02 | 2021-01-15 | 天津工业大学 | Small target image fast threshold segmentation method based on mathematical morphology |
| CN106501160A (en) * | 2016-09-08 | 2017-03-15 | 长春迪瑞医疗科技股份有限公司 | A kind of method for classifying particles and particle classifying device |
| CN109059789A (en) * | 2018-08-21 | 2018-12-21 | 成都天衡智造科技有限公司 | Cable pitch online test method based on machine vision |
| CN109059789B (en) * | 2018-08-21 | 2020-04-07 | 成都天衡智造科技有限公司 | Cable pitch online detection method based on machine vision |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103426155A (en) | Column diagram boundary dividing method based on solving column diagram changing rate | |
| CN102629305B (en) | Feature selection method facing to SNP (Single Nucleotide Polymorphism) data | |
| CN109088869B (en) | APT attack detection method and device | |
| CN102999886A (en) | Image edge detector and ruler raster grid line precision detection system | |
| CN105518455A (en) | Peak detection method | |
| CN104573050A (en) | Continuous attribute discretization method based on Canopy clustering and BIRCH hierarchical clustering | |
| CN103020626A (en) | Method and system for detecting elliptical target in image | |
| DE102020215209A1 (en) | Time-to-digital converter and conversion process | |
| WO2012034596A1 (en) | Improvements in or relating to automatic detection of the number of lanes into which a road is divided | |
| CN105512206A (en) | Outlier detection method based on clustering | |
| CN104992454A (en) | Regionalized automatic-cluster-change image segmentation method | |
| CN104102833B (en) | Based on the tax index normalization found between compact district and fusion calculation method | |
| CN103218837B (en) | The histogrammic method for drafting of a kind of unequal interval based on empirical distribution function | |
| CN106295500A (en) | A kind of repetition dither signal and normal signal separation method | |
| CN103353883A (en) | Big data stream type cluster processing system and method for on-demand clustering | |
| CN104636748A (en) | License plate recognition method and device | |
| CN106991815B (en) | Traffic jam control method | |
| CN115372995A (en) | Laser radar target detection method and system based on European clustering | |
| CN104851105B (en) | A kind of improved froth images dividing method based on watershed transform | |
| CN105574363A (en) | Feature selection method based on SVM-RFE (Support Vector Machine-Recursive Feature Elimination) and overlapping degree | |
| CN106201949B (en) | The analysis method of eye figure shake | |
| CN111504333B (en) | Method and device for adjusting elevation of uplink and downlink roads in three-dimensional map | |
| CN104881670A (en) | Rapid target extraction method used for SAR azimuth estimation | |
| CN106528507A (en) | Chinese text similarity detection method and device | |
| CN110110685A (en) | Data are filed method, apparatus and server |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20131204 |