CN102595051B - Automatic exposure method based on sequencing image - Google Patents
Automatic exposure method based on sequencing image Download PDFInfo
- Publication number
- CN102595051B CN102595051B CN201210071167.8A CN201210071167A CN102595051B CN 102595051 B CN102595051 B CN 102595051B CN 201210071167 A CN201210071167 A CN 201210071167A CN 102595051 B CN102595051 B CN 102595051B
- Authority
- CN
- China
- Prior art keywords
- image
- automatic exposure
- time
- exposure
- signal
- 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.)
- Active
Links
Abstract
The invention relates to the filed of image processing, and provides an automatic exposure method based on a sequencing image. The automatic exposure method comprises the steps as follows: a) in exposure time, an image sensor is utilized to photograph an image in imaging regions; b) the array point center of the image is confirmed by utilizing the Gaussian model, and the signal value of the array point center is obtained; c) a computer auxiliary device is utilized to analyze the signal value, and the signal intensity of the image is obtained; d) new exposure time is confirmed according to desired signal intensity and by utilizing the linear relation between the signal intensity and the exposure time; e) in the new exposure time, the steps a) to c) are repeated, and then step f) is repeated; and f) the coherence of the signal intensity of an image is photographed in the new exposure time and the desired signal intensity is judged, if not coincident, the steps d) to f) are repeated, and if coincident, the new exposure time is the automatic exposure time. The automatic exposure method provided by the invention achieves regulation according to the desired signal intensity, accurate and automatic exposure is realized, and accurate sequencing image data can be surely photographed.
Description
Technical field
The present invention relates to image processing field, more particularly, relate to a kind of method of automatic exposure of gene sequencing image.
Background technology
Automatic exposure is the core technology of image processing field, and whether exposure normally directly affects the effect of image, overexposure, and it is too bright that image seems, under-exposure, it is too dark that image seems.Therefore, automatic exposure is to picture signal processing critical function in current imaging system.
In gene sequencing field, multiple DNA fragmentations are fixed on globule, and globule is fixed on sequencing device, or multiple DNA fragmentation is directly fixed on sequencing device, form array point, and a globule or the DNA fragmentation that is fixed on same place are an array point.When sequencing reaction, DNA fragmentation with react with the material of fluorescent marker after, when when carrying fluorescent marker on excitation array point, each array point is luminous, now gets final product photographic images.The image of taking just as the star of intersperse in sky, a just star seemingly of each array point, the signal value of each array dot center is a key factor of decision clear picture.
In a kind of prior art, the method for automatic exposure is: step 1. whole theoretical gray areas is divided between several gray areas according to gray scale, determines gray scale desired value; Step 2. receive the view data of imageing sensor output, judge between the gray value and affiliated gray area thereof of each pixel in view data, calculate respectively the average gray in each gray area, then the average gray between several gray areas is weighted to average computation, obtains intensity-weighted average; Step 3. intensity-weighted average and gray scale desired value are compared, if meet the demands, giving tacit consent to current exposure is optimum value; If do not meet the demands, carry out step 4; Step 4., calculate the time for exposure making new advances according to the poor of intensity-weighted average and gray scale desired value controlling the time of integration according to intensity-weighted average and gray scale desired value, imageing sensor gathers image information according to the new time for exposure, execution step two.
In technique scheme, can not determine the array Dian center of sequencing image, also cannot obtain the automatic exposure time under the image intensity signal of expecting according to feature of image, cause the too strong and overexposure of sequencing image signal strength signal intensity, or signal strength signal intensity is too weak and excessively dark, thereby make automatic exposure inaccurate, finally cause the sequencing image of shooting unintelligible.
Therefore need a kind of method of automatic exposure of new gene sequencing image, can obtain the automatic exposure time under the image intensity signal of expectation, finally ensure that imageing sensor shoots sequencing image clearly.
Summary of the invention
The object of the present invention is to provide a kind of method of automatic exposure of gene sequencing image, be intended to solve the automatic exposure time under the image intensity signal that cannot accurately obtain expectation in prior art, thereby causing exposing inaccurately cannot shoot the problem of image clearly.
In order to realize goal of the invention, the method for the automatic exposure of gene sequencing image comprises the following steps: steps A. and under the time for exposure, utilize imageing sensor to take the image of adopting in graph region; Step B. utilizes Gauss model to determine array dot center in image, and obtains the signal value of array dot center; Step C. utilizes computer-aided equipment analytic signal value, obtains the signal strength signal intensity of image; Step D., according to the image intensity signal of expecting, utilizes the linear relationship of signal strength signal intensity and time for exposure, determines the new time for exposure under the image intensity signal of expecting; Step e. under the new time for exposure, after repeating step A, B and C, enter step F; Step F. judge the consistency of the signal strength signal intensity of the image of taking under the new time for exposure and the image intensity signal of expectation; Inconsistent, repeating step D, E and F; Unanimously, the new time for exposure is the automatic exposure time.
Wherein, above-mentioned imageing sensor comprises amplifying device, and the multiplication factor of this amplifying device is n, and n is natural number, is preferably n >=20.
Wherein, described computer-aided equipment is the electronic computer device arbitrarily with data processing function.Be preferably PC.
Wherein, described Gauss model is the model of setting up according to the feature of sequencing image.Wherein, the signal strength signal intensity that the image intensity signal of described expectation is predefined image.Described image is stored in two-dimensional matrix with the form of gray value.Described consistency determination methods is: in certain limit α, be also S
expect± α, is judged as consistent; Not in this scope, be judged as inconsistent.Wherein, S
expectfor the image intensity signal of expecting, the size of α determines according to actual conditions, and in the time that needs precision is high, α value can be little, and in the time that the precision of needs is not high, the value of α can be large.
Wherein, utilize s
0/ t
0=s
expect/ t
1, the new time for exposure under the image intensity signal that can expect, wherein, t
0for the initial time for exposure of setting, s
0for at t
0signal strength signal intensity under time for exposure, s
expectfor the image intensity signal of expecting, t
1for the new time for exposure.The image of taking in the new time for exposure, the signal strength signal intensity s of the image of gained
1if, s
1with the image intensity signal of expecting when inconsistent, adjusted to s the automatic exposure time
1/ t
1=s
expect/ t
2; Work as s
1when consistent with the image intensity signal of expecting, obtain the automatic time for exposure.The formula of determining the new time for exposure under the image intensity signal of expecting is: S
current/ t
current=S
expect/ t
newly.Wherein, S
currentfor current demand signal intensity, t
currentfor current time for exposure, t
newlyfor the new time for exposure.
Wherein, the signal value of array dot center is preferably the gray value of this array dot center; Or the signal value of array dot center is preferably the brightness ratio of the point of the maximum and brightness minimum of brightness in array point.
Wherein, in described array point, brightness size is the gray value at this place.
Wherein, described step B comprises the following steps: step B1. sets up Gauss model; Step B2. utilizes Gauss model to mate with image, obtains the similarity of image and Gauss model; Step B3. utilizes similarity and threshold value to determine array Dian center in image.
Wherein, described Gauss model is set up according to the shape of the size of array point and array point, and the Gauss model of setting up is two-dimensional matrix.Gauss model obtains similarity with images match successively, often completes mating of a sub-Gaussian and image, obtains a similarity matrix.According to the feature of array point in sequencing image, setting threshold.The size of threshold value is determined according to actual conditions.
Wherein, in described step B3, choose similarity matrix in maximum and threshold value comparison, when in similarity matrix, maximum value is higher than threshold value, point corresponding to this maximum is defined as array dot center in image; Lower than threshold value, cast out this point, the region that also this similarity matrix is corresponding is without array point.Have multiple maximums higher than threshold value in a similarity matrix time, the point of getting in the image that multiple peaked central points are corresponding is array dot center in image.
Wherein, the signal value of described array dot center is the gray value of this array dot center.
Wherein, choose largest value and second largest value and the threshold value comparison in similarity matrix.If second largest value is higher than threshold value, and only have one, point corresponding to this second largest value is defined as array dot center in image, casts out largest value; If second largest value is higher than threshold value, and have multiplely, the mid point of getting point corresponding to this second largest value is defined as array dot center in image; If second largest value is lower than threshold value, largest value, higher than threshold value, is defined as array dot center in image by point corresponding largest value; If largest value and second largest value are all less than threshold value, the region that this similarity matrix is corresponding is without array point.
Wherein, described step C comprises the following steps: the signal value of step C1. pair array dot center sorts; Step C2. gets front h the large point of signal value of array dot center, the signal strength signal intensity using the average of its signal value as image.
Wherein, described step C comprises the following steps: step C1 '. and the signal value of pair array dot center sorts; Step C2 '. get the average of signal value of h array dot center of the centre after sequence as the signal strength signal intensity of image.
Wherein, h is natural number, preferred 400≤h≤800.
Wherein, also comprise after described step F: step G. utilizes imageing sensor repeatedly to F, to obtain multiple automatic exposure time adopting diverse location photographic images repeating step A in graph region; Step H. utilizes the linear relationship of signal strength signal intensity and time for exposure, screens multiple automatic exposure time, obtains the best automatic exposure time.
Wherein, above-mentionedly analyze at the same same position photographic images of adopting graph region, obtain an automatic exposure time; Analyze at the same diverse location photographic images of adopting graph region, obtain same multiple automatic exposure times of adopting graph region; Adopt the diverse location of graph region in difference and adopt figure, obtain different multiple automatic exposure times of adopting in graph region; Also can get only an automatic exposure time in each graph region of adopting, multiple graph region of adopting obtain multiple automatic exposure time.
Preferably, adopting at one must an automatic exposure time in graph region.
Wherein, the number of multiple automatic exposure times is At, and At is natural number; Preferably, 3≤At≤9; Adopt the number of graph region and do not limit, be preferably 3～9.
Wherein, in step H, utilize the relation of signal strength signal intensity and time for exposure, screen multiple automatic exposure time, obtain the best automatic exposure time.The method of screening is not limit.Be preferably and adopt median method, mean value method screening, obtain the best automatic exposure time.
As from the foregoing, the present invention, by determining array dot center and the signal value thereof in sequencing image, progressively adjusts the time for exposure under the image intensity signal of expecting, the time for exposure under the image intensity signal of being expected accurately; Simultaneously, to adopting the diverse location photographic images in graph region, obtain multiple time for exposure, effectively having avoided on the one hand the signal strength signal intensity of the image that signal cancellation causes is not the real signal strength signal intensity of image, cause the time for exposure inaccurate, on the other hand, from multiple automatic exposure, determine the best automatic exposure time, ensured the accuracy of automatic exposure time.
Brief description of the drawings
Fig. 1 is the structural representation of the device of automatic exposure in gene sequencing in one embodiment of the invention;
Fig. 2 is the automatic explosion method flow chart of gene sequencing image in one embodiment of the invention;
Fig. 3 adopts graph region to adopt the distribution map of figure position in one embodiment of the invention;
Fig. 4 adopts graph region to adopt the distribution map of figure position in another embodiment of the present invention;
Fig. 5 adopts graph region to adopt the distribution map of figure position in another embodiment of the present invention;
Fig. 6 is the method flow diagram of the automatic exposure of gene sequencing image in another embodiment of the present invention;
Fig. 7 is the method flow diagram of determining array dot center in one embodiment of the invention;
Fig. 8 is the method flow diagram of determining array dot center in another embodiment of the present invention;
Fig. 9 is the method flow diagram that obtains image intensity signal in one embodiment of the invention.
Embodiment
In order to make object of the present invention, technical scheme and advantage clearer, below in conjunction with drawings and Examples, the present invention is further elaborated.
Automatic explosion method of the present invention can ensure that the device of automatic exposure photographs image clearly.In sequencing image, need to determine array Dian center in sequencing image, determine the signal strength signal intensity of whole sequencing image according to the signal value at array Dian center, then judge that the consistency of the image intensity signal of signal strength signal intensity and expectation carries out feedback regulation, finally draw the automatic exposure time.Then, according to aforesaid method, obtain multiple automatic exposure time, filter out the best automatic exposure time.This programme is through repeatedly feedback regulation, final the automatic time for exposure; And from multiple automatic exposure time, filter out the best automatic exposure time, and make the automatic exposure time more accurate, the image that has overcome the inaccurate shooting causing of traditional automatic exposure is crossed secretly or is excessively strong.
For the automatic exposure device of sequencing image, the present invention proposes an embodiment, as shown in Figure 1.Comprise light source, filtering apparatus, imageing sensor and computer-aided equipment.Wherein: described light source is can excite to adopt the luminous light source of sample that carries label in graph region, includes but not limited to mercury lamp light source and laser.Described filtering apparatus adopts for filtering stray light that light source sends and light source activation thereof the stray light that graph region sample sends when luminous, allows the light of specific wavelength to pass through.Described imageing sensor is the transducer that can experience arbitrarily optical image information and convert usable output signal to, includes but not limited to CCD and cmos image sensor.Described computer-aided equipment is the equipment that can process data, analyze, and the device of control chart image-position sensor operation is preferably personal computer and corresponding program thereof, or is chip or single-chip microcomputer and peripheral circuit and program.
In the present embodiment, adopt and in graph region, be fixed with the sample that carries label; The light that light source sends after filtration electro-optical device makes the light arrival of specific wavelength adopt graph region, other light is by filtering, it is luminous that the sample of label is carried in the optical excitation of specific wavelength, the light that sample sends arrives imageing sensor, imageing sensor gathers image, and computer-aided equipment is analyzed and processed the image gathering.
Wherein, described filtering apparatus comprises many group filters, and filter allows the light of specific wavelength to pass through.As required filtering apparatus is switched on different filters.
Based on a upper embodiment, the present invention proposes another embodiment, as shown in Figure 2.The method of the automatic exposure of gene sequencing image comprises the steps: that step S1. is under the time for exposure, utilizes imageing sensor to take the image of adopting in graph region; Step S2. utilizes Gauss model to determine array dot center in image, and obtains the signal value of array dot center; Step S3. utilizes computer-aided equipment analysis image, obtains the signal strength signal intensity of image; Step S4., according to expecting figure signal strength signal intensity, utilizes the linear relationship of signal strength signal intensity and time for exposure, determines the new time for exposure under the image intensity signal of expecting; Step S5. under the new time for exposure, repeating step S1, S2 and S3; Step S6. judges the consistency of the signal strength signal intensity of the image of taking under the new time for exposure and the image intensity signal of expectation; Inconsistent, repeating step S4, S5 and S6; Unanimously, the new time for exposure is the automatic exposure time.
In the technical scheme of the present embodiment, utilize Gauss model, accurately determine array dot center in image, then obtain signal strength signal intensity, according to the relation of the image intensity signal of signal strength signal intensity and expectation, feedback regulation, final automatic time for exposure, this technical scheme makes the automatic exposure time more accurate, has avoided the image of shooting to cross bright or excessively dark.In the present embodiment, for the consistency judgement in step S6, provide example.If the image intensity signal S expecting
expect=3000 o'clock, the signal strength signal intensity S of the image of taking under the new time for exposure
when before=2940 o'clock.In the time that the precision of needs is higher, get α=50, now, S
current< S
expect-α, judges inconsistent.In the time that the precision of needs is lower, get α=100, now, S
expect-α≤S
current≤ S
expect+ α, judgement is consistent.
Based on the embodiment shown in Fig. 1, the present invention proposes another embodiment, and the method for the automatic exposure of gene sequencing image comprises the steps: that step S1. is under the time for exposure, utilizes imageing sensor to take the image of adopting in graph region; Step S2. utilizes Gauss model to determine array dot center in image, and obtains the signal value of array dot center; Step S3. utilizes computer-aided equipment analysis image, obtains the signal strength signal intensity of image; Step S4., according to expecting figure signal strength signal intensity, utilizes the linear relationship of signal strength signal intensity and time for exposure, determines the new time for exposure under the image intensity signal of expecting; Step S5. under the new time for exposure, repeating step S1, S2 and S3; Step S6. judges the consistency of the signal strength signal intensity of the image of taking under the new time for exposure and the image intensity signal of expectation; Inconsistent, repeating step S4, S5 and S6; Unanimously, the new time for exposure is the automatic exposure time, enters step S7; Step S7. utilize imageing sensor repeatedly in difference is adopted graph region photographic images repeating step S1 to S7, obtain multiple automatic exposure time; Step S8. utilizes the relation of signal strength signal intensity and time for exposure, screens multiple automatic exposure time, obtains the best automatic exposure time.
Preferred in the present embodiment, obtain an automatic exposure time, need carry out image taking at the same position of adopting graph region; When each repeating step S1, S2, S3, S4, S5 and S6, obtain multiple automatic exposure time, while obtaining each automatic exposure time, take the position difference of adopting the image in graph region.In the preferred version of the present embodiment, at diverse location photographic images, make the view data that draws more representative, also more accurate, thereby provide reliability foundation for follow-up picture signal analysis; Also the image that effectively prevents from repeatedly taking same position, causes signal cancellation, and affects the analysis of successive image signal simultaneously, makes the time for exposure inaccurate.
In the present embodiment, the comprehensive whole graph region of adopting, adopts the image intensity signal of expectation and the signal strength signal intensity comparison under the new time for exposure, and repeatedly feedback regulation, finally obtains the automatic exposure time; Calculate respectively the automatic exposure time to clapping graph region diverse location, obtain multiple time for exposure, choose the best automatic exposure time, make under this best automatic exposure time, the image of taking is more accurate, prevents that the image of shooting from crossing data analysis dark or excessive and impact order-checking.It is more accurate that the technical program has realized automatic exposure on the whole.
Imageing sensor in above-described embodiment comprises an amplifying device, and this amplifying device is for imaging, and the multiple of this amplifying device is selected according to actual conditions.Preferably, this amplifying device is microscope or magnifying glass.。The multiplication factor of the amplifying device in this imageing sensor is larger, the image that imageing sensor is taken is more clear, but, multiplication factor is excessive, array point in the image that imageing sensor is taken is just fewer, the whole quantity of the image that region takes of taking is just more, because the time of the utilization of the data analysis of every image is almost equal, thus the excessive Efficiency Decreasing that causes data analysis of multiplication factor.
In the present embodiment, the multiplication factor n of this imageing sensor is preferably n >=20.Preferred, the multiplication factor n of this imageing sensor is 20≤n≤40.N is natural number.In the technical scheme of the present embodiment, the automatic exposure that imageing sensor carries out under suitable multiplication factor, makes imageing sensor both ensure the definition of photographic images, has ensured again the efficiency of data analysis.
The whole graph region of adopting can be divided into multiple graph region of adopting, and each graph region of adopting can obtain multiple automatic exposure time, and while obtaining the different automatic exposure time, what the image of shooting was corresponding adopts the position difference in graph region.Preferably, each graph region of adopting obtains an automatic exposure time.The number of time for exposure is not limit, preferred, and the number At of automatic exposure time is, 3≤At≤9.
Adopt position and the number of automatic exposure time in graph region for the image of taking corresponding, the present invention proposes an embodiment, as shown in Figure 3.Imageing sensor respectively adopt in graph region 1., 2., 3. position photographic images.The process of automatic exposure is: step S1., under the initial exposure time, utilizes imageing sensor to take and adopts the 1. image of interior optional position of graph region; Step S2. utilizes Gauss model to determine array dot center in image, and obtains the signal value of array dot center; Step S3. utilizes computer-aided equipment analysis image, obtains the signal strength signal intensity of image; Step S4., according to expecting figure signal strength signal intensity, utilizes the linear relationship of signal strength signal intensity and time for exposure, determines the new time for exposure under the image intensity signal of expecting; Step S5., under the new time for exposure, utilizes imageing sensor to take and adopts the 1. image of the interior position identical with photographic images position under the initial exposure time of graph region, repeating step S2 and S3; Step S6. judges the consistency of the signal strength signal intensity of the image of taking under the new time for exposure and the image intensity signal of expectation; Inconsistent, repeating step S4, S5 and S6; Unanimously, the new time for exposure is the automatic exposure time, obtains first automatic exposure time, enters step S7; Step S7., under the initial exposure time, utilizes imageing sensor to take and adopts the 2. image of interior optional position of graph region, and repeating step S2, S3, S4, S5 and S6, obtain second automatic exposure time; Under the initial exposure time, utilize imageing sensor to take and adopt the 3. image of interior optional position of graph region again, repeating step S2, S3, S4, S5 and S6, obtain the 3rd automatic exposure time; Step S8. utilizes the relation of signal strength signal intensity and time for exposure, screens multiple automatic exposure time, obtains the best automatic exposure time.
Adopt position and the number of automatic exposure time in graph region for the image of taking corresponding, the present invention proposes another embodiment, as shown in Figure 4.Imageing sensor respectively adopt in graph region 1., 2., 3., 4., 5. position photographic images.The process of automatic exposure is: step S1., under the initial exposure time, utilizes imageing sensor to take and adopts the 1. image of interior optional position of graph region; Step S2. utilizes Gauss model to determine array dot center in image, and obtains the signal value of array dot center; Step S3. utilizes computer-aided equipment analysis image, obtains the signal strength signal intensity of image; Step S4., according to expecting figure signal strength signal intensity, utilizes the linear relationship of signal strength signal intensity and time for exposure, determines the new time for exposure under the image intensity signal of expecting; Step S5., under the new time for exposure, utilizes imageing sensor to take and adopts the 1. image of the interior position identical with photographic images position under the initial exposure time of graph region, repeating step S2 and S3; Step S6. judges the consistency of the signal strength signal intensity of the image of taking under the new time for exposure and the image intensity signal of expectation; Inconsistent, repeating step S4, S5 and S6; Unanimously, the new time for exposure is the automatic exposure time, obtains first automatic exposure time; Under the initial exposure time, utilize imageing sensor to adopt graph region to shooting and 2., 3., 4., 5. carry out respectively image taking, repeating step S2, to S7, obtained for the second to the 5th automatic exposure time; Then, utilize the relation of signal strength signal intensity and time for exposure, screen multiple automatic exposure time, obtain the best automatic exposure time.
Adopt position and the number of automatic exposure time in graph region for the image of taking corresponding, the present invention proposes another embodiment, as shown in Figure 5.Imageing sensor respectively adopt in graph region 1., 2., 3., 4., 5., 6., 7., 8., 9. in optional position photographic images, carry out automatic exposure analysis, obtain first to the 9th automatic exposure time, then, recycling signal strength signal intensity and the relation of time for exposure, screen the best automatic exposure time from nine automatic exposure times, obtain the best automatic exposure time.Shown signal strength signal intensity and the relation of time for exposure can adopt the middle position method of ratio of signal strength signal intensity and time for exposure or the mean value method of ratio to select optimum exposure time.
In the present embodiment, can to same adopt graph region repeatedly repeating step S1 to S7, obtain multiple automatic exposure time, also can to same adopt graph region carry out a step S1 to S7, obtain an automatic exposure time.
Preferably, adopt graph region repeating step S1 to S7 once same, obtain an automatic exposure time.Also be that b adopts graph region and obtains b automatic exposure time.Then, in b time for exposure, filter out the best automatic exposure time.
In above-described embodiment, the diverse location of adopting in graph region is carried out to photographic images, avoid same position photographic images to cause fluorescence signal cancellation, cause the signal strength signal intensity of the image of taking lower, and can not get the correct automatic exposure time; Simultaneously, diverse location is carried out to photographic images, analyze to obtain time of multiple automatic exposure, filter out the best automatic exposure time, this technical scheme has realized the signal strength signal intensity of the image of the comprehensive whole diverse location of adopting graph region, take into full account the signal strength signal intensity of whole image, and then draw the best automatic exposure time, make the best automatic exposure time of gained more accurate.
Based on the embodiment shown in Fig. 1, Fig. 6 shows the method flow diagram of the automatic exposure of gene sequencing image in one embodiment of the invention.Step 1, under the time for exposure, utilize imageing sensor to take and adopt image in graph region; Step 2, utilize Gauss model to determine array dot center in image, and the signal value of array dot center; Step 3, utilize computer-aided equipment analysis image, obtain the signal strength signal intensity of image; Step 4, according to expecting figure signal strength signal intensity, utilize the linear relationship of signal strength signal intensity and time for exposure, determine the new time for exposure under the image intensity signal of expecting; Under the new time for exposure, repeating step one, to step 3, then, enters step 5; Step 5, judge the consistency of the signal strength signal intensity of the image of taking under the new time for exposure and the image intensity signal of expectation; Inconsistent, turn back to step 4; Unanimously, the new time for exposure is the automatic exposure time, enters step 6; Step 6, the automatic time for exposure; Then turn back to step 1, repeat abovementioned steps, obtain another automatic exposure time; Until obtain, after each automatic exposure time of adopting in graph region, entering step 7; Step 7, utilize the relation of signal strength signal intensity and time for exposure, screen multiple automatic exposure time, obtain the best automatic exposure time.
Technical scheme of the present invention is carried out respectively automatic exposure analysis to the zones of different of sequencing image, obtain the automatic exposure time of zones of different, screening optimum exposure time, the time for exposure obtaining more can represent the time for exposure in whole region, thereby make the automatic exposure of whole sequencing image more accurate, avoided some region overexposure in sequencing image and excessively bright, some regional exposure is not enough and excessively dark.
Fig. 7 shows the method flow diagram of determining array dot center in one embodiment of the invention in sequencing image.Described method comprises:
Step S21. sets up Gauss model;
Described Gauss model is accurately to quantize things according to Gaussian probability-density function (normal distribution curve), and a things is decomposed into some models forming based on Gaussian probability-density function (normal distribution curve).The foundation of the Gauss model in the present invention is to set up according to the distribution of array point in sequencing image, and this array point is the point one by one forming of when order-checking point sample.Can obtain the signal strength signal intensity of each array point, the information such as size, the shape of array point of array point through imageing sensor.
Set up Gauss model according to the shape of the array point of sequencing image and size.In the present embodiment, array point is for circular, and size is 2～8 pixels, according to the array point of sequencing image, preferably sets up the matrix of 7*7, obtains Gauss model.
Step S22. utilizes Gauss model to mate with image, obtains the similarity of image and Gauss model;
Described matching process is as follows: utilize following formula to mate
Wherein, s=1,2 ..., M-m; T=1,2 ..., N-n; W (x, y) is image array,
for the average of w (x, y); G (x, y) is Gauss model,
for the average of g (x, y).Wherein, M and the N size of presentation video matrix respectively; M and n represent the size of Gauss model.In this step, the value of γ (s, t) is [1,1], after Gauss model travels through on image array, can obtain the individual value of (M-m) * (N-n), wherein, γ (s, t) (s when maximum, t) being the point of similarity maximum, through threshold filtering, is array Dian center higher than the point of this threshold value.By coupling, the value of γ (s, t) is transformed to [1,1], irrelevant with the size of image, make as long as according to the feature of array point, definite threshold, can obtain array dot center accurately.
Step S23. determines array dot center in image according to similarity and threshold value.
Described determines that according to similarity and threshold value the method at array Dian center in image is: choose maximum and threshold value comparison in similarity matrix, when in similarity matrix, the value of maximum is higher than threshold value, point corresponding to this maximum is defined as array dot center in image; Lower than threshold value, cast out this point, the region that also this similarity matrix is corresponding is without array point.Have multiple maximums higher than threshold value in a similarity matrix time, the point of getting in the image that multiple peaked central points are corresponding is array dot center in image.
Concrete operations can be: choose largest value and second largest value and the threshold value comparison in similarity matrix.If second largest value is higher than threshold value, and only have one, point corresponding to this second largest value is defined as array dot center in image, casts out largest value; If second largest value is higher than threshold value, and have multiplely, the mid point of getting point corresponding to this second largest value is defined as array dot center in image; If second largest value is lower than threshold value, largest value, higher than threshold value, is defined as array dot center in image by point corresponding largest value; If largest value and second largest value are all less than threshold value, the region that this similarity matrix is corresponding is without array point.
The technical scheme of the present embodiment,, has realized and has accurately determined pattern matrix dot center by setting up Gauss's template according to the feature of array point.
For the signal value of array dot center, the present invention proposes an embodiment.The image that image array in the present invention is taken is stored in matrix with the form of gray value, obtains image array.In image, the gray value of array dot center is the signal value of array dot center.The technical scheme of the present embodiment, directly uses the gray value of array dot center as the signal value of array dot center, more convenient.
For the signal value of array dot center, the present invention proposes another embodiment.Signal value by the brightness ratio of the point of brightness maximum and brightness minimum in array point as array dot center.The gray value that in array point in the present embodiment, certain some brightness is this place.In the technical scheme of the present embodiment, utilize the ratio of brightness maximum and brightness minimum as the signal value of array dot center, fully take into account the whole distribution of signal, make result more accurate.
Fig. 8 shows the method flow diagram of determining array dot center in one embodiment of the invention in sequencing image.In this sequencing image, white round dot is array point, and the black color dots in array point is array dot center.The method of determining array dot center is:
Step S21. sets up Gauss model.
Described Gauss model is accurately to quantize things according to Gaussian probability-density function (normal distribution curve), and a things is decomposed into some models forming based on Gaussian probability-density function (normal distribution curve).The foundation of the Gauss model in the present invention is to set up according to the distribution of array point in sequencing image, and this array point is the point one by one forming of when order-checking point sample.Can obtain the signal strength signal intensity of each array point, the information such as size, the shape of array point of array point through imageing sensor.
Set up Gauss model according to the shape of the array point of sequencing image and size.In the present embodiment, array point is for circular, and size is 2～8 pixels, according to the array point of sequencing image, preferably sets up the matrix of 7*7, obtains Gauss model.
Step S22. utilizes Gauss model to mate with image, obtains the similarity of image and Gauss model;
Described matching process is as follows: utilize following formula to mate
Wherein, s=1,2 ..., M-m; T=1,2 ..., N-n; W (x, y) is image array,
for the average of w (x, y); G (x, y) is Gauss model,
for the average of g (x, y).Wherein, M and the N size of presentation video matrix respectively; M and n represent the size of Gauss model.
Wherein, in the time finding array dot center, Gauss model is to set up according to the shape of array point in sequencing image and size, and Gauss model mates with whole sequencing image, Gauss model often does while once coupling and obtains a matrix, is also the similarity of Gauss model and sequencing image.According to the feature of array point, setting threshold, maximizing point in matrix, (s, the t) position while being also γ (s, t) maximum, maximum threshold in this matrix, Ze Gaidianwei array dot center point.
Step S23. determines array dot center in image according to similarity and threshold value.
In image in the present embodiment, definite method of array dot center can be with reference to the method described in above-described embodiment.In the present embodiment, by the feature of Gauss model and image, Gauss model is mated with image, adopt two-dimensional correlation computing, thereby make the precision of images match higher, this is pattern matrix dot center definitely provides reliable foundation.
For in above-described embodiment for pattern matrix point determine, propose the present embodiment, the threshold value of setting is 0.0235.
Suppose that Gauss model is the matrix of 7*7, the matrix that sequencing image is 7*14, Gauss model and image carry out computing, obtain Gauss model and mate to obtain similarity with the image array of sequencing image, obtain first similarity matrix result and are:
Wherein, in this matrix, maximum is 0.0259.0.0259 is greater than threshold value, and now in sequencing image, this corresponding point is array dot center.
Mobile Gauss model on sequencing image matrix, Gauss model mates to obtain similarity with the image array of sequencing image, obtains another similarity matrix result to be:
Wherein, in this matrix, maximum is 0.0215.0.0215 is less than threshold value, now the Gai Dianbushi array dot center of correspondence in sequencing image.
In the present embodiment, Gauss model does and mates respectively at whole sequencing image, obtains multiple matrixes, selects the maximum in each matrix, according to the feature of sequencing image, utilizes threshold value, and multiple maximums are filtered, and obtains the array dot center of sequencing image.The technical scheme of the present embodiment has realized and has disturbed the filtration of array point, thereby obtains array dot center accurately.
For in above-described embodiment for pattern matrix point determine, the present invention proposes another embodiment, the threshold value of setting is 0.0235.
Suppose that Gauss model is the matrix of 7*7, the matrix that sequencing image is 7*14, Gauss model and image carry out computing, obtain Gauss model and mate to obtain similarity with the image array of sequencing image, obtain first similarity matrix result and are:
Wherein, in this matrix, Second Largest Value is 0.0251, has four identical values.0.0251 is greater than threshold value, and now getting this is that the mid point of point corresponding to value is as array dot center.
Mobile Gauss model on sequencing image matrix, Gauss model mates to obtain similarity with the image array of sequencing image, obtains another similarity matrix result to be:
Wherein, in this matrix, Second Largest Value is 0.0209, totally four Second Largest Values, and 0.0209 is less than threshold value.Get again first value 0.0215 and be less than threshold value, in this point that now in sequencing image, this matrix is corresponding, do not have array dot center, also there is no array point.
In the present embodiment, Gauss model does and mates respectively at whole sequencing image, obtains multiple matrixes, determines the array dot center of sequencing image according to matrix.The technical scheme of the present embodiment has realized and has disturbed the filtration of array point, thereby obtains array dot center accurately.
Fig. 9 shows one embodiment of the invention and obtains the method flow diagram of image intensity signal.The method comprises: the signal value of step S31. pair array dot center sorts; Step S32. gets front h the large point of signal value of array dot center, the signal strength signal intensity using the average of its signal value as image.
In the present embodiment, to the sequence of signal value can according to from small to large or from greatly to sort method, pick out front h the large point of signal value of array dot center, get the average of signal value of this h array dot center as the signal strength signal intensity of image.In the present embodiment, the array point that the signal value of array dot center is lower may not be the array point of sequencing image, may be interference signal.Get the average of signal value of front h the point that the signal value of array dot center is large as the signal strength signal intensity of image, can ensure the accuracy of the signal strength signal intensity of image, avoid interference signal and participate in picture appraisal, and make the signal strength signal intensity of image inaccurate.
For the method for obtaining image intensity signal, the present invention proposes another embodiment.Described method comprises:
Step S31 ': the signal value of pair array dot center sorts;
Step S32 ': get the average of signal value of h array dot center of the centre after sequence as the signal strength signal intensity of image.
In the present embodiment, effectively prevent that signal value due to the array dot center that causes of error of taking is excessive or signal value is too small, the average of the signal value in the middle of adopting represents the signal strength signal intensity of whole image, make the signal strength signal intensity of image more accurate, thereby the data that make to participate in subsequent analysis are more reliable, for the accuracy of whole automatic exposure provides guarantee.
For adopting median method to obtain the best automatic exposure time, the present invention proposes an embodiment.The signal strength signal intensity of gained r automatic exposure time and image is respectively: t1, s1; T2, s2; Tr, sr.s1/t1＝A1；s2/t2＝A2；……；sr/tr＝Ar。A1 to automatic exposure time corresponding to the median in Ar be optimum exposure time.
The technical scheme of the present embodiment, adopts the signal strength signal intensity of unit interval to compare, and gets the median of the signal strength signal intensity of unit interval, and automatic exposure is more accurate on the whole to make whole sequencing image, avoids some region overexposure, some regional exposure deficiency.
For adopting mean value method to obtain the best automatic exposure time, the present invention proposes an embodiment.Gained f automatic exposure time, be respectively: t1, t2 ..., tf.Remove t1 to minimum and maximum in tf, get the mean value of remaining automatic exposure time, this mean value is the best automatic exposure time.
The technical scheme of the present embodiment, take into full account the difference existing between each several part in sequencing image, the appropriateness of relatively exposing of the regional in whole image, avoids having occurred some region overexposure, some regional exposure deficiency, makes the automatic exposure of sequencing image more accurate.
It should be noted that the typical application of the present invention is not limited to the automatic exposure of gene sequencing image, similarly also can apply method set forth in the present invention at other in image processing field.
The foregoing is only preferred embodiment of the present invention, not in order to limit the present invention, all any amendments 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 (10)
1. a method for the automatic exposure of gene sequencing image, is characterized in that, said method comprising the steps of:
A. under the time for exposure, utilize imageing sensor to take the image of adopting in graph region;
B. utilize Gauss model to determine array dot center in image, and obtain the signal value of array dot center;
C. utilize computer-aided equipment analytic signal value, obtain the signal strength signal intensity of image;
D. according to the image intensity signal of expecting, utilize the linear relationship of signal strength signal intensity and time for exposure, determine the new time for exposure under the image intensity signal of expecting;
E. under the new time for exposure, after repeating step A, B and C, enter step F;
F. judge the consistency of the signal strength signal intensity of the image of taking under the new time for exposure and the image intensity signal of expectation;
Inconsistent, repeating step D, E and F;
Unanimously, the new time for exposure is the automatic exposure time.
2. the method for the automatic exposure of gene sequencing image according to claim 1, is characterized in that, described step B comprises the following steps:
B1. set up Gauss model;
B2. utilize Gauss model to travel through and mate with image, obtain the similarity of image and Gauss model;
B3. utilize similarity and threshold value to determine array dot center in image.
3. the method for the automatic exposure of gene sequencing image according to claim 2, is characterized in that, in described step B3, when in similarity matrix, the value of maximum is higher than threshold value, point corresponding to this maximum is defined as array dot center in image.
4. the method for the automatic exposure of gene sequencing image according to claim 1, is characterized in that, described imageing sensor comprises amplifying device, and the multiplication factor of this amplifying device is n, and n is natural number, n >=20.
5. the method for the automatic exposure of gene sequencing image according to claim 1, is characterized in that, the signal value of array dot center is the gray value of this array dot center.
6. according to the method for the automatic exposure of the gene sequencing image described in any one in claim 1 to 5, it is characterized in that, described step C comprises the following steps:
C1. the signal value of pair array dot center sorts;
C2. get front h the large point of signal value of array dot center, the signal strength signal intensity using the average of its signal value as image.
7. according to the method for the automatic exposure of the gene sequencing image described in any one in claim 1 to 5, it is characterized in that, described step C comprises the following steps
C1 '. the signal value of pair array dot center sorts;
C2 '. get the average of signal value of h array dot center of the centre after sequence as the signal strength signal intensity of image.
8. according to the automatic explosion method of the gene sequencing image described in any one in claim 1 to 5, it is characterized in that, after described step F, also comprise:
G. utilize imageing sensor repeatedly to F, to obtain multiple automatic exposure time adopting diverse location photographic images repeating step A in graph region;
H. utilize the relation of signal strength signal intensity and time for exposure, screen multiple automatic exposure time, obtain the best automatic exposure time.
9. the method for the automatic exposure of gene sequencing image according to claim 8, is characterized in that, the number of described multiple automatic exposure times is At, and At is natural number, 3≤At≤9.
10. the method for the automatic exposure of gene sequencing image according to claim 8, is characterized in that, in described step H, utilizes the relation of signal strength signal intensity and time for exposure, adopts median method or mean value method screening, obtains the best automatic exposure time.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210071167.8A CN102595051B (en) | 2012-03-16 | 2012-03-16 | Automatic exposure method based on sequencing image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210071167.8A CN102595051B (en) | 2012-03-16 | 2012-03-16 | Automatic exposure method based on sequencing image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102595051A CN102595051A (en) | 2012-07-18 |
| CN102595051B true CN102595051B (en) | 2014-10-22 |
Family
ID=46483216
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210071167.8A Active CN102595051B (en) | 2012-03-16 | 2012-03-16 | Automatic exposure method based on sequencing image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102595051B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2003014400A1 (en) * | 2001-08-08 | 2003-02-20 | Applied Precision, Llc | Time-delay integration imaging of biological specimens |
| CN1483142A (en) * | 2000-12-26 | 2004-03-17 | ���ְ�˹��ѧ��ҵ��ʽ���� | Fluorescent intensity measuring method and apparatus |
| CN1934450A (en) * | 2004-01-15 | 2007-03-21 | 内诺斯佩尔公司 | Nanoparticle imaging system and method |
| WO2008077407A1 (en) * | 2006-12-22 | 2008-07-03 | Aalborg Universitet | Light induced material deposition by molecular immobilization |
-
2012
- 2012-03-16 CN CN201210071167.8A patent/CN102595051B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1483142A (en) * | 2000-12-26 | 2004-03-17 | ���ְ�˹��ѧ��ҵ��ʽ���� | Fluorescent intensity measuring method and apparatus |
| WO2003014400A1 (en) * | 2001-08-08 | 2003-02-20 | Applied Precision, Llc | Time-delay integration imaging of biological specimens |
| CN1934450A (en) * | 2004-01-15 | 2007-03-21 | 内诺斯佩尔公司 | Nanoparticle imaging system and method |
| WO2008077407A1 (en) * | 2006-12-22 | 2008-07-03 | Aalborg Universitet | Light induced material deposition by molecular immobilization |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102595051A (en) | 2012-07-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111368342B (en) | Image tampering identification model training method, image tampering identification method and device | |
| CN108764372A (en) | Construction method and device, mobile terminal, the readable storage medium storing program for executing of data set | |
| CN102982336B (en) | Model of cognition generates method and system | |
| CN108875821A (en) | The training method and device of disaggregated model, mobile terminal, readable storage medium storing program for executing | |
| CN113807210A (en) | Remote sensing image semantic segmentation method based on pyramid segmentation attention module | |
| CN108419028B (en) | Image processing method, image processing device, computer-readable storage medium and electronic equipment | |
| CN104504744B (en) | A kind of true method and device of plan for synthesizing license plate image | |
| CN102867418A (en) | Method and device for judging license plate identification accuracy | |
| EP2154630A1 (en) | Image identification method and imaging apparatus | |
| CN105283902A (en) | Image processing device, image processing method, and image processing program | |
| CN108039044B (en) | Vehicle intelligent queuing system and method based on multi-scale convolutional neural network | |
| CN101142812A (en) | Image processing device and method, program and recording medium | |
| CN113822951B (en) | Image processing method, device, electronic equipment and storage medium | |
| CN112818732B (en) | Image processing method, device, computer equipment and storage medium | |
| CN111626960A (en) | Image defogging method, terminal and computer storage medium | |
| CN113824884B (en) | Shooting method and device, shooting equipment and computer readable storage medium | |
| CN110363720A (en) | A kind of color enhancement method, apparatus, equipment and the storage medium of image | |
| CN112712519A (en) | Non-contact type machine-made sandstone powder content intelligent real-time detection method and device | |
| CN106034202A (en) | Adjusting method and adjusting device for video splicing camera | |
| CN108184075A (en) | For generating the method and apparatus of image | |
| CN105959593A (en) | Exposure method for camera device and camera device | |
| CN113177438A (en) | Image processing method, apparatus and storage medium | |
| CN115049675A (en) | Generation area determination and light spot generation method, apparatus, medium, and program product | |
| CN112861987A (en) | Target detection method under dark light environment | |
| CN102595051B (en) | Automatic exposure method based on sequencing image |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |