JPH0785055B2 - Band sequence pattern automatic reading method and device - Google Patents

Description

TECHNICAL FIELD The present invention relates to a band sequence pattern automatic reading method and apparatus, and more particularly to a band sequence pattern suitable for reading a base band sequence pattern of a gene imaged on an X-ray filter. The present invention relates to an automatic reading method and device.

[Background of the Invention]

The gene of an organism is composed of a substance called deoxyribonucleic acid (DNA), and has a structure in which four kinds of organic bases of adenine (A), cytosine (C), guanine (G) and thymine (T) are arranged in a double helix. Has been elucidated and is generally known. The shotgun method is a method for determining the nucleotide sequence of this DNA. In this method, DNA strands are disassembled with a reagent that chemically cleaves each base, and each base is arrayed according to a length difference by electrophoresis, and each base is imaged on a film by an X-ray image. As shown in FIG. 2A, each of the base codes A, C, G, and T on the film 3 on which the X-ray image was projected extends in a direction substantially perpendicular to the lane within one lane. A plurality of each existing band 4 is arranged, and the gene can be analyzed by analyzing the base sequence on this film.

For this analysis, the base sequence pattern on the film is input to the computer. Conventionally, a digitizer has been used as an input device for this purpose, the X-ray film 3 is placed on the digitizer, and the position of the band 4 on the film is sequentially pointed with a pen or a cursor to input the output signal of the digitizer to a computer. There is a device to do.

However, in the conventional apparatus, since input is performed by sequentially pointing the bands on the X-ray film with a pen or the like, it takes time and effort to input all the bands on the X-ray film, and if input operation is performed for a long time. There are many typos.

Therefore, there has been a demand for the development of an automatic reader for nucleotide sequence patterns.

[Object of the Invention]

An object of the present invention is to provide a band position automatic reading method and apparatus.

[Outline of Invention]

In the first invention of the present application, a pattern having a plurality of bands in a plurality of lanes is photoelectrically converted, and on a line passing through a plurality of points in the lane vertical direction based on the grayscale data of each photoelectrically converted pixel. The sum (histogram) of the gray-scale data (pixel data) of the pixel is detected, the point representing each lane existence area, for example, the center position of each lane is detected from the histogram, and the detected center of each lane is detected. The grayscale data of the picture elements on the position is collected to obtain the spectrum of each lane, and the plurality of maximum positions are detected as band positions from the spectrum.

By doing this, the lane position can be accurately determined.
The band representative position, for example, the center position can be accurately obtained.

In a particularly preferred embodiment of the present invention, a histogram in the lane vertical direction is obtained for each minute section in the lane direction, and the position of the lane area representative point, for example, the center position is obtained for each section. By doing so, the band position can be accurately obtained even when the lane is not necessarily a straight line.

Further, in the embodiment of the present invention, the positions of the lane center point and the both end points of the lane are obtained as the points representing the lane area, and the center position and both end positions of each band are obtained based on these. Thereby, even if the band is inclined with respect to the lane direction, the band position can be accurately obtained.

Further, in the second invention of the present application, as a device suitable for carrying out the first invention of the present application, a first memory having storage locations for different positions in the vertical direction of the lane and pixel data photoelectrically converted A second memory for storing the two-dimensional position of each pixel, and the pixel data on the one line in synchronization with the photoelectric conversion of the pixel on one line in the lane vertical direction of the pattern. Means for adding corresponding lane vertical position data already written to the first memory and writing the result again to the first memory to form histogram data for each lane vertical position; Based on the histogram of the memory, the position of a point representing the existence area of each lane is detected, pixel data for pixels on a line passing through this detection point is read from the second memory, The resulting spectral data for lanes, comprising means for detecting the maximum position of the spectral data as a band position of each lane.

By doing so, the band positions in a plurality of lanes on the pattern can be detected at high speed.

Further, in the embodiment of the present invention, the adding means obtains the histogram for each predetermined section in the lane direction. At this time, the second memory is composed of two areas for storing the pixel data in the two sections, and when the histogram is obtained for the new section, these two areas are sequentially switched and used.

By doing so, the second memory does not need to store all pixel data on the pattern, which simplifies the device.

Example of Invention

1 (A) and 1 (B) show a block system diagram of an embodiment of a band arrangement pattern reading device according to the present invention and a schematic configuration of an image scanner (5), respectively. In FIG. 1A, the image scanner 5 is provided with, for example, a transparent glass 2 of about A3 size.

X-ray film 3 in which the base sequences of genes are imaged by X-ray images
Is placed on the transparent glass 2. As shown in FIG. 2 (A), the X-ray film 3 has an image in which bands 4 shown by solid lines corresponding to four kinds of base codes A, C, G and T sequenced by the shotgun method are arranged. It is shadowed. Normally, the X-ray film 3 has a plurality of base codes imaged in the X-axis direction with one base code A, C, D, G, T as one set.
Only one set is shown in FIG.
In the X-ray film 3 described above, the longitudinal direction of the lane which is the effective range of each of the base codes A, C, G and T is Y of the image scanner 5.
It is positioned so as to substantially coincide with the axial direction.

A one-dimensional image sensor 6 is provided corresponding to the position of the transparent glass 2. The light from the one-dimensional light emitting element 7 is applied to the X-ray film 3, and the reflected light of one line of the X-ray film 3 in the X-axis direction is incident on the image sensor 6. The image sensor 6 photoelectrically converts the reflected light for one line in pixel units, and then incorporates the charge coupled device (CC
D) (not shown) or the like and serially take out the electric signal thus taken out to, for example, an 8-bit digital signal A /
D-convert and output. The image sensor 6 and the light emitting element 7 are moved at a constant speed in the Y-axis direction by a drive mechanism (not shown) to detect and read the X-ray film 3. Further, the reading device is provided with operation keys 9 such as a start key.

The image scanner 5 and the operation keys 7 are shown in FIG.
The interface circuit 8 shown in FIG. The interface circuit 8 includes an image scanner 5 and a CPU 12
It is a circuit to interface with.

The interface circuit 8, CPU12, ROM14, RAM150,
The disk device 16 and the interface unit 17 are connected to each other via a bus line 13.

The interface circuit 8 has the configuration shown in FIG. In FIG. 1C, the control circuit 30 generates operation control signals for other circuits in the interface circuit 8 according to the address signals and control signals input from the CPU 12 to the terminals 31, 32 via the bus 13. Command register 33 is a terminal
A control signal from the CPU 12 for the image scanner 5 and its driving mechanism input to the memory 34 is stored, and this is supplied to the image scanner 5 and its driving mechanism from a terminal 35 in accordance with an operation signal from the control circuit 30. Status register 36
Stores the status signals of the image scanner 5 and its driving mechanism input from the terminal 37 and the status signals inside the interface circuit 8 such as the control circuit 30 and the address counter 38. It is supplied to the CPU 12 via the bus line 13.

The address counter 38 counts the clock supplied from the control circuit 30 to generate an address corresponding to the pixel position on one line in the X-axis direction, and the carry of the X counter to count the pixels in the Y-axis direction. It is composed of a Y counter 38b which generates an address corresponding to a position. The output addresses of the X counter 38a and the Y counter 38b are supplied to the image memory 40, and the output address of the X counter 38a is further supplied to the addition memory 41.

The image memory 40 stores the output pixel data of the image sensor 6 input from the terminal 42 according to the address, and supplies the pixel data stored under the control of the control circuit 30 to the CPU 12 from the read terminal 43. Further, the pixel data input to the terminal 42 is supplied to the adder 44, where it is read from the addition memory 41 and added to the addition pixel data latched by the latch circuit 45. The added pixel data output by the adder 44 is stored in the addition memory 41 again. Also,
The addition pixel data stored in the addition memory 41 is the control circuit 30.
Is supplied to the CPU 12 from the read terminal 46 under the control of.

FIG. 3 shows a flow chart of an embodiment of the band array reading program executed by the CPU 12. When the operation signal generated by pressing the start key in the operation key 7 is supplied to the CPU 12 via the interface circuit 8, the CPU
12 starts execution of the process shown in FIG. 3 (A), and the process of reading the X-ray film 3 by the image sensor 6 and the process of spectrum collection are executed (step 100).

In step 100, the process shown in FIG. 3 (B) is executed. In the figure, in step 130, the CPU 12 supplies a control signal to the image sensor 6 via the interface circuit 8, whereby the image sensor 6 is moved by the drive mechanism in the Y-axis direction at a distance d ₁ (d ₁ is, for example, several tens of mm). Only you can move at a constant speed. The image sensor 5 has a small distance d ₂
Every time it moves (for example, d ₂ = 1/16 mm), pixel data (detection data) for one line in the X-axis direction is serially output. The pixel data is supplied to the interface circuit 8 and sequentially written in the image memory built in the interface circuit 8. Further, this pixel data is added by the adder 44 in the interface circuit 8 to the addition pixel data read from the address of the addition memory 41 corresponding to each detection position of one line in the X-axis direction, and the addition memory 4 is added. It is stored in the same address as the reading of.

That is, the image sensor 6 reads the pixel data for the distance d ₁ at the same time in parallel to the address corresponding to each detection position of one line in the X-axis direction of the addition memory 41 at the minute distance d ₂ in the Y-axis direction.
The addition data, which is the sum of the pixel data for each, is stored and a histogram is obtained.

Hereinafter, the process of step 130 will be described in more detail.

The pixel data serially transmitted as shown in FIG. 4 (A) is input to the terminal 42 shown in FIG. 1 (C) and supplied to the image memory 40 and the adder 44. In synchronization with this pixel data, the X counter 38a of the address counter 38 changes its output address value as shown in FIG. 4 (B). Further, the control circuit 30 supplies an operation control signal (write strobe) as shown in FIG. 4 (C) to the addition memory 41, and the addition memory 41 writes in the low level period of the write strobe and reads in the high level period. Do. The same write strobe as in FIG. 4 (C) is supplied to the image memory 40, and the image memory 40 is written during the low level period.

The addition memory 41 has an address corresponding to each detection position of one line in the X-axis direction, and when the CPU 12 instructs the image sensor 6 to start reading, the contents of all addresses are cleared to zero. Thereafter, the addition memory 41 reads the addition image data stored at the address designated by the X counter 38a during the low level period of the write strobe shown in FIG. 4 (C). The latch circuit 45 is shown in FIG. 4 (D).
A latch pulse with the write strobe inverted is supplied from the control circuit 30 as shown in (4), and the above-mentioned added pixel data is latched at the rising edge of this latch pulse, that is, at the end of reading from the addition memory 41. As a result, the data is supplied from the latch circuit 45 to the adder 44. The added pixel data changes as shown in FIG. 4 (E). Therefore, the adder 44
The addition pixel data, which is the addition value of the pixel data from the terminal 44 and the addition pixel data from the latch circuit 45 obtained in step (4), changes as shown in FIG. The addition pixel data is written in the addition memory 41 during the low level period of the strobe shown in FIG.

Thus each pixel position in the X-axis direction when moving the image sensor 6 by a distance d ₁ to the address of the addition memory 41 corresponding to the (detected position), the sum of pixel data for each small distance d ₁ in the Y-axis direction The added pixel data, that is, the histogram is stored.

The image memory 40 is divided into a first area 41 and a second area 42 according to the most significant bit of the output address of the Y counter 38b, and the CPU 12 first instructs the image sensor 6 to read. When the value of the above most significant bit is “0”
The first area 41 is selected with, and this value changes each time the reading instruction is given. The address of the address counter 38 supplied to the image memory 40 is incremented each time pixel data is input from the terminal 42, and the pixel data (while the image sensor 6 is initially moved by the distance d ₁ as described above) ( For example, all the data in the partial area 31 in FIG. 2) is stored in the first area 41. Note that all pixel data (for example, data in the partial area 32 in FIG. 2) obtained while the image sensor 6 moves the distance d ₁ is stored in the second area 42 next.

In this way, the pixel data from the X-ray film 3 is read and written in the image memory 40, and at the same time, the pixel data in the predetermined section (distance d ₁ ) in the Y-axis direction is added for each pixel position unit in the X-axis direction. Then, it is written in the addition memory 41 and a histogram is generated. The above is the contents of step 130. The generation of the above histogram is performed at high speed by hardware such as the adder 44, whereby the image memory 40
It can be processed in parallel with the writing of the pixel data to the pixel.

Further, the histogram is generated by the addition memory 41, the single adder 44, and the latch circuit 45 attached thereto, and the number of parts is small and the circuit configuration is simple.

It is stored in the addition memory 41. When converted into an analog amount, the histogram has a waveform as shown in FIG.
In step 131, the CPU 12 differentiates the histogram of the addition memory 41 in the X direction to obtain a differential value whose analog conversion value has a waveform as shown in FIG.
It is stored in a work area (not shown) in the RAM 15. Further, with reference to the maximum positive value and the maximum negative value of the differential value, for example, a positive reference value THH and a negative reference value THL, which are half the respective values, are set. After this point, the point where the differential value exceeds the positive reference value THH and the point where it is less than the negative reference THL x _LA , x _RA , x _LC , x _RC , x _LG , x
The values of _RG , x _LT , and x _RT are calculated and recognized as the positions of the left and right ends in the X-axis direction of the lanes of the base codes A, C, G, and T in the partial area 31 (FIG. 2) (step 131). ). Subsequently, the values of the midpoints x _MA , x _MC , x _MG , x _MC of the points x _LA , x _RA , x _LC , x _RC , x _LG and x _RG , x _LT and x _RT are calculated. These points x _Li
.About.x _Ri (i = A, C, G, T) is recognized as a sampling point (step 132).

After that, the CPU 12 moves the image sensor 6 further in the Y-axis direction by the distance d ₁ so that the pixel data output from the image sensor 6 is sequentially stored in the second area 42 of the image memory 40, and the histogram is stored in the addition memory 41. The added pixel data as is stored (step 133). Thereafter, as in step 131, the left end, the right end, and the right end of each lane x ′ _Li , x ′ _Ri (i
= A, C, G, T) and each value is recognized as the position of the lane of each base code (step 134). Further, the midpoint _x'Mi (i = A, C, G, T) of each lane in the second partial area 32 (Fig. 2) is calculated based on the values of the above points, and all sampling points _{x'Li, Mi , Ri} (i = A, G, C, T) values are obtained.

Here, even if there is no band to be extracted in a specific lane in the added area, the gray value in the lane is larger than the gray value of the background, and therefore a significant difference is obtained in the addition histogram. , You can ask for a lane.
However, if the difference in gray value is less than the sensitivity of the detector and the difference cannot be found in the addition histogram, the boundary value of the lane found before that may be used. If the small section for which the lane is not found is the first small section, the lanes corresponding to the other found bases may be proportionally distributed. In other words, if two lanes cannot be found, a section that cannot be found may be divided into two and the center thereof may be the center of the lane.

The sampling points x _Li , x _Mi , x _Ri and _x'Li , x ' _Mi ,
Since x ′ _Ri (i = A, C, G, T) is obtained based on the added value during the Y-axis direction distance d ₁ , the Y-axis of the first and second partial regions 31 and 32 is Positions of the left, center, and right ends of each lane at the center position in the direction p _Li , p _Mi , p _Ri , p ' _Li , p' _Mi , p ' _Ri ,
It can be regarded as p ′ _Ri (i = A, C, G, T). Therefore, if the pixel data on the line connecting the points p _Li and p ′ _Ri , p _Mi and p ′ _Mi , and p ′ _Ri and p ′ _Mi are read from the image memory 40, the pixels at the left end, the center, and the right end of the lane are respectively read. A data string (spectrum) is obtained.

By the way, at this point, the pixel data of the first and second partial regions 31 and 32 (FIG. 2) of the film 3 are stored in the first and second regions 41 and 42 of the image memory 40, respectively. . Therefore, CPU 12 sequentially reads the point p _Li and p 'point than the point p _Li pixel data of the corresponding position on the straight line connecting the _Li (address) p' _Li direction, sequentially writing the predetermined address in the RAM15 Then, the spectra are collected between the points p _LA and p ′ _LA . Similarly, spectra are collected between points p _MA and p ′ _{MA and} between points p _RA and p ′ _RA . Do the same for the other lanes. (Step 136). As a result, each base code A, C, G, T is spectrally collected at both ends and the center of the band. Thereafter, the first area 41 and the second area 42 of the image memory 40 are switched and used (step 13).
7). Next, the image sensor 6 scans to the right end of the transparent glass 2 shown in FIG. 1 (A) to determine whether or not the spectrum acquisition of the entire surface of the X-ray film 3 is completed (step
138), and if not completed, move to step 133. Here, since the first area 41 and the second area 42 of the image memory 40 are switched in step 137, in step 133 the image memory 40 in which the previous pixel data was written.
The pixel data of the area 33 (FIG. 2) at the next distance d ₁ is sequentially written in an area (for example, the first area 41) different from the internal area (for example, the second area 42). In this way, step 33 ~
By repeating 37, the spectrum of the entire surface of the X-ray film 3 is collected. Thus step 100
Ends.

In the above step 100, 12 kinds of spectra S _Li , S _Mi , S in total of 3 kinds for each base code A, C, G, T
_Ri (i = A, C, G, T) is collected and stored in the RAM 15 in the order of the Y-axis position of each pixel. In the next step 200, band positions are detected by analyzing the spectra S _Li , S _Mi and S _Ri (i = A, C, G, T). FIG. 3C shows a detailed flowchart of this process 200. By the way, the spectrum S _Li , S
_Mi and S _Ri (i = A, C, G, T) are analog quantities S
When converted into (y), the waveform is as shown in FIG. 6 (A). In step 230 of FIG. 3 (C), the CPU 12 reads each spectrum S (y) from the RAM 15 and differentiates the spectrum S (y) so that the analog conversion value becomes a waveform shown in FIG. 6 (B). y) is calculated. After that, with reference to the maximum positive value and the maximum negative value of the differential spectrum S ′ (y) as a reference, for example, a positive reference value THH that is a half value of the maximum positive value and a maximum negative value, and a negative reference value THH Each of the reference values THL of is set (step 131).

Next, the value y ₁ in the Y-axis direction in which the differential spectrum S ′ (y) changes from a value equal to or greater than the positive reference value THH to less than the positive reference value THH,
Y ₄ , y ₇ , ..., And the differential spectrum S ′ (y) changes from a value equal to or greater than the negative reference value THL to a value less than the negative reference value THL.
The values y ₃ , y ₆ , y ₉ , ... in the axial direction are obtained and the peak existing range y ₁
〜Y ₃ , y ₄ 〜y ₆ , y ₇ 〜y ₉ , ... are detected (step 13
2). After that, the differential spectrum S ′ (y) is 0 for the peak existence range y _{1 to} y ₃ , and therefore the value y _{2 in the} Y-axis direction at which the spectrum S (y) is maximized is obtained. peak existence range _{y 4 ~y 6, y 7 ~y} 9, a differential spectrum S '(y) is 0 at ...... respectively, thus, Y-axis direction of the value y ₅ to spectrum S (y) is the maximum, y ₈ ... Is required (step 133). The value obtained here y ₃ ,
The positions of peak points are y ₅ , y ₈ , .... In this way, the analysis spectrum SS (y) representing the band position is obtained. The analysis spectrum SS (y) is stored in the RAM 15 as a table in which an address is added in pixel units in the Y-axis direction, the value of the band existence position is “1”, and the value of the band non-existence position is “0”. . Thus, the differential spectrum S '(y) is used to detect the band existence range, and the differential spectrum S' (y) is 0 within this band existence range, and therefore the spectrum S (y) becomes maximum. By detecting the point as the band position, the existence position of the band can be accurately specified regardless of the density of the band and the size of the band width.

After the peak is detected, the positive reference value THH is compared with a predetermined value α (step 134), and when the positive reference value THH is large, the routine proceeds to step 135. In step 135, the positive reference value TH
Each of H and the negative reference value THL is updated to, for example, 1/2 of the value so far, and the process proceeds to step 132. As a result, peak detection is repeated using the new positive reference value THH and negative reference value THL (step 133), but the existing range y _{1 to} y ₃ , y _{4 to} y ₆ obtained in the previous step 132 is repeated. , y _{7 to} y ₉ are not subjected to peak detection processing. Positive reference value THH in step 134
When is smaller than the predetermined value α, the processing of FIG. 3 (C) for one spectrum is completed. The processing of FIG. 3 (C) is performed by the spectrum S _Li , S _Mi , S _Ri (i = A, C, G, T).
Analysis spectra SS _Li , SS _Mi , and SS _Ri (i = A, C, G, T) that are performed for all the bands and are band detection information are obtained.

In the next step 300, the band position is normalized by the routine shown in FIG. 3 (D).

In FIG. 3 (D), the base code C is RAM15.
The analysis spectrum SS _{MC in} is searched in order of the Y-axis direction,
The address of the analysis spectrum SS _{MC having} the value "1", that is, the value y _{51 of the} position in the Y-axis direction shown in FIG. 6 is obtained (step 330). In step 331, the analysis spectrum with the value "1" is displayed.
It is determined whether SS _MC has been searched, and if searched, the process proceeds to step 332. In step 332, the analysis spectrum SS _LC , in the vicinity of the value y ₅₁ obtained in this search (Y-axis direction)
The value Y coordinate y ₄₁ , y _{61 of the} point where each value of SS _LR becomes “1” is obtained, and the left end, center, and right end positions (y ₄₁ , y) of band C are obtained.
₅₁ , y ₆₁ ) is detected. Of course, when the values y ₄₁ and y ₆₁ cannot be obtained, the value y ₅₁ is recognized as an error and the existence of the band C is not detected. After that, the inclination θ _{C1 of the} straight line passing through the three points y ₄₁ , y ₅₁ , and y ₆₁ on the detected band C ₁ is calculated (step 333).

Next, using the slope y _C1 with the value y ₅₁ , the boundary between the base codes C and G
The value (normalized value) of the y-axis direction position C _IN of band C ₁ in II (FIG. 7) is calculated (step 334). After that, the process proceeds to step 330, and steps 330 to 334 are repeatedly executed to obtain the normalized value C _2N , ... Spectrum SS
_{When the MC} search is completed, the process of FIG. 3 (D) for the base code C is completed, and thereafter it is determined that the process of FIG. 3 (D) is not executed for the base code G (step 33).
5), the normalized positions g _1N , g _2N , ... _Are similarly obtained. After that, for base code A, the analysis spectrum SS
_{The MA} is searched for in RAM15 (step 336),
The value of the Y axis direction position of the analysis spectrum SS _MA , for example, y ₂₁
Is obtained (step 337), the values y ₁₁ and y _{31 at} which the values of the analysis spectra SS _LA and SS _RA are “1” are obtained and the position of the band A ₁ (y ₁₁ , y ₂₁ , y _{31) is obtained.} ) Is detected (step 338). Further, in step 338, the slope θ _{a1 of the} straight line passing through the three values y ₁₁ , y ₂₁ , and y ₃₁ of band A ₁ is obtained (step 339), and the base code is calculated using the value y ₃₁ and the slope θ _a1.
A quasi-normalized value α _{1 at} the boundary I between A and C is calculated (step 340).

Here, since the boundary is the boundary of the lane obtained earlier, it may be inclined. Even in that case, the band obtained in step 339 can be regarded as a straight line with a value of θ _a1 , so that the quasi-normalized value α ₁ can be easily obtained by finding the intersection of two straight lines.

Next, the band Ci (of the base code C at the position most recent to the quasi-normalized value α ₁ in the large part and the small part in the y-axis direction from the quasi-normalized value α ₁ which is the value at the y-axis direction is Ci ( In this example, the gradient θc _i (θc _{1 in} this example) of C ₁ ) is found, and using these, the lane of the base code C is normalized to α ₁
The inclination θ _a c ₁ of the position corresponding to is calculated (step 34
1). Further, the Y coordinate of the normalized position a _IN of the band A _{1 on} the boundary II is obtained using the quasi-normalized value α ₁ and the slope θ _a c ₁ (step 342). After that, the process shifts to step 335 and steps 336 to 342 are repeatedly executed, and the Y coordinates of the normalized positions a _2N , a _3N , ... Of the bands A ₂ and A ₃ are obtained. After that (step 343), the process of FIG. 3 (E) is similarly performed on the base code T to obtain the normalized value t _1N , ....

Normalized value a obtained by the processing of FIGS. 3 (D) and (E)
_1N , c _1N , g _1N , t _1N, etc. are values in the Y-axis direction on the boundary II. The area SD in the RAM 15 has an address in the Y-axis direction on the boundary II, and the normalized position a _1N ...
The base code data "A" is stored in the address corresponding to ..., and similarly the normalized positions c _1N , ... and g _1N , ... and t _1N ,
…… Base code data “C” in the address for each
"G" and "T" are stored respectively.

In this way, the band 4 of each of the base codes A, C, G, T is normalized on the boundary II according to the inclination of each band, so that even if each band is inclined, the band arrangement order of each base code is Can be read accurately.

Thus, step 22 in FIG. 3 (A) is completed.

After this, in step 400 of FIG. 3A, the area SD in the RAM 15 is
The base code data of is corrected. Figure 3 (E)
Shows a detailed flowchart of the correction processing of the base code data. By the way, the band interval (Y-axis direction) and the Y-axis direction position of adjacent bands which are read from the X-ray film 3 imaged by the shotgun method and normalized on the boundary line shown in the destruction II of FIG. The value has a relationship as shown in FIG. In step 430 of FIG. 3 (E), the area S
The band intervals of adjacent bands are sequentially obtained from the difference in the addresses where the base code data of D is stored, and this band interval is compared with the minimum band interval D _min in FIG.
When the band interval obtained here is smaller than the minimum band interval D _min, Y of the adjacent bands for which the band interval is obtained is determined.
The value of the axial position, that is, the base code data of the band in which the address value of the area SD is small is erased from the area SD as noise. The above noise is canceled. The noise canceling process is performed on the entire area SD.

Next, using the area SD in which the noise elimination processing has been performed, the value of the position in the Y-axis direction, that is, the address is sequentially increased, and the band interval between adjacent bands is obtained. The band spacing characteristics obtained by translating the characteristics of FIG. 7 in the Y-axis direction (left-right direction in FIG. 7) according to the band spacing of the obtained bands are calculated, and have the addresses corresponding to the Y-axis direction in the RAM 15. The reference band intervals are sequentially stored in the area B for the band interval function (step 431).

The band spacing function BY is obtained in this way because the value in the Y-axis direction in FIG. 7 changes depending on the mounting position of the X-ray film 3 on the transparent glass 2.

Next, the address of the area SD is sequentially increased to obtain the band interval between the adjacent bands, and the reference read from the area B of the address corresponding to the value of the midpoint of the adjacent bands in the Y-axis direction. It is compared with an allowable band interval obtained by multiplying the band interval by a predetermined constant (step 432). When the band interval obtained from the address of the area SD individually is larger than the allowable band interval, it is determined that there is a band detection omission, and the step 433,
Band rediscovery is performed at 434 as follows. In step 433, the base codes A, C, G, which correspond to the positions of the correction points, which are the midpoints of the bands adjacent to each other in the area SD, in the Y-axis direction.
The inclination of each band of T is obtained by the same method as that for detecting the band position described above, and the center of each band of the base codes A, C, G and T corresponding to the above correction point (X in FIG. 4 (A)) The value of the Y-axis direction position (inverse normalization correction point) for _MA , X _MC , X _MG , X _MT ) is obtained.

The value of the denormalization correction points of the spectrum S ₂ in step 434 after which, the value of the spectrum S ₂ recognized a band in the denormalization correction point near, is compared with the value of example X _2, the denormalization When the value of the correction point is large, the inverse normalization correction point is set as the band candidate point of the base code A. Similarly, the band candidate points of the base codes C, G, T are obtained for the spectra S ₅ , S ₈ , S ₁₁ , respectively. When a plurality of band candidate points are obtained in this way, the spectrum values at the respective band candidate points are compared. As a result, the base code having the maximum spectrum value is determined as the corrected base code, and the corrected base code data is stored at the address corresponding to the correction point in the area SD. In addition, here, when it is known in advance from the nature of the film that the spectrum value increases or decreases depending on the contents of the bands before and after, it is effective to make a determination in consideration of the condition. For example, when the adjacent band is A, the value of adjacent C is reduced by 30%, or when the ghost appears (the value of 30% increases), the value of C is corrected. I will compare it after doing it. When it is determined in step 432 that the band interval obtained from the address of the area SD is smaller than the allowable band interval, or after step 434 is completed, the process proceeds to step 435, where the band interval check for the entire Y-axis direction is performed. It is determined whether or not the process is completed, and if not completed, the process proceeds to step 432. In this way steps 432-43
By repeating step 4, the band interval check is executed for the entire Y-axis direction, and the process of FIG. 3 (E) ends.

In step 500, the base code data of the area SD is checked based on the above-mentioned existence probability, the erroneously detected base code data is erased, and the base code data of the missed detection is corrected.

After this, the base code data stored in the area SD is set to Y
Disk device that reads in the order of addresses corresponding to the axial direction
The data is stored in 16 and sent from an interface circuit 17 to an external computer (not shown).

In the above example, the reading of the band sequence information of the base code of the X-ray imaged gene was described as an example, but the present invention is not limited to this, and the reading of the band sequence information in a chromatograph of amino acids and the like is performed. Alternatively, a band that moves by electrophoresis or the like may be directly detected by a fixed sensor.

〔The invention's effect〕

As described above, according to the present invention, the band arrangement pattern can be automatically read. Furthermore, in the method of obtaining a histogram for each constant section in the longitudinal direction of the lane and detecting the band position from it, it is possible to accurately detect the band position in each lane even if each lane meanders. It has features. Further, in the apparatus according to the present invention, since the histogram can be obtained in parallel with the collection of the grayscale data, the processing speed can be increased.

[Brief description of drawings]

FIG. 1 (A) is a block diagram of an automatic reading device for a band array pattern according to the present invention, and FIG. 1 (B) is a diagram showing a main part of an image scanner used in the device of FIG. 1 (A). FIG. 1C is a block diagram of an interface circuit of the device of FIG. 1A, and FIG. 2A is a diagram showing an example of a band array pattern read by the device of FIG. 1A. 2 (B) is a diagram showing another example of the band arrangement pattern of FIG. 3 (A), FIGS. 3 (A) to 3 (E) are operation flow diagrams of the apparatus of FIG. 1 (A), FIG. 4 is a time chart of the circuit of FIG. 1 (C), FIG. 5 is a diagram for explaining a method of detecting a lane position representative point by the device of FIG. 1 (A), and FIG. FIG. 7A is a diagram for explaining the band position detection method in the lane by the device, and FIG. 7 is a band position by the device of FIG. 1A. Diagram for explaining the normalization, FIG. 8 is a diagram for explaining the apparatus correction band position detection error due to the first view (A). 2 ... glass plate, 3 ... film, 4 ... band, 6 ...
… Photoelectric conversion circuit, 7… Light source

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Tetsuo Kurisu 6-81, Onoe-cho, Naka-ku, Yokohama-shi, Kanagawa Hitachi Software Engineering Co., Ltd. (56) Reference JP-A-57-204437 (JP, A) Special Kai 57-139646 (JP, A) JP 58-45540 (JP, A) JP 57-91444 (JP, A) JP 60-253848 (JP, A) JP 55-34383 (JP JP, B2)

Claims (9)

[Claims] 1. A two-dimensional arrangement in which a plurality of lanes each having a plurality of band patterns arranged substantially along a first direction are arranged adjacent to each other in a second direction having a vertical relationship with the first direction. In the automatically arranged band array pattern reading method, light emitted from a light source is irradiated in the first direction of an arbitrary lane to generate grayscale data, the grayscale data is stored, and the grayscale data is stored in the second direction. On the other hand, the frequency of occurrence of the grayscale data satisfying a predetermined criterion is counted to create a histogram for each predetermined period, the characteristic points of the histogram for each predetermined period are detected, and the histogram intersects the line connecting the characteristic points. A band arrangement characterized in that grayscale data corresponding to a band pattern is detected from the stored grayscale data to obtain position information of the band pattern along the first direction. Automatic column pattern reading method. 2. The above-mentioned characteristic point is the above-mentioned second of the above-mentioned arbitrary lane.
The band array pattern automatic reading method according to claim 1, wherein the point is a point representing a midpoint position of the length in the direction of. 3. The band arrangement according to claim 1, wherein the characteristic points of the arbitrary lane are points that represent positions of both ends of the length of the arbitrary lane in the second direction. Automatic pattern reading method. 4. A differential curve is obtained by differentiating the histogram, the sex maximum value and the negative maximum value of the differential curve are detected, and a constant relationship is established with respect to the positive maximum value and the negative maximum value. Calculating a positive reference value and a negative reference value having
Each of the points indicating the positive reference value and the negative reference value in the differential curve is set as a point representing the both end positions, and the middle point of the points representing the both end positions is set as the middle point position. The band array pattern automatic reading method according to claim 3, characterized in that it is a representative point. 5. The band arrangement pattern according to any one of claims 1 to 4, wherein the position of the maximum grayscale data of the grayscale data is used as position information of the band pattern. Automatic reading method. 6. A two-dimensional structure in which a plurality of lanes, each of which has a plurality of band patterns arranged substantially along a first direction, are arranged adjacent to each other in a second direction which is perpendicular to the first direction. Object having a band array arranged in a uniform manner, a plurality of light sources arranged in the second direction and scanning in the first direction, and grayscale data of the object using the light emitted from the light source. And a photoelectric conversion means for outputting
Memory and the grayscale data corresponding to the plurality of light sources at a predetermined timing and the data stored in the first memory are added to obtain the grayscale corresponding to the band pattern arranged in the first direction. An adding means for forming a histogram showing the appearance frequency of data in the first memory; a second memory for sequentially storing the grayscale data corresponding to the band pattern in synchronization with the photoelectric conversion by the photoelectric converting means; The representative point at the position of the grayscale data corresponding to the band pattern in the second direction is detected from the histogram, and the grayscale data corresponding to the band pattern intersecting on the line connecting the representative points arranged in the first direction. From the second memory, and processing means for detecting the position of the band pattern along the first direction. Characteristic band array pattern automatic reader. 7. The first memory holds a set of grayscale data corresponding to the band pattern in one section of a plurality of sections arranged in the first direction by the photoelectric conversion, respectively. And a second area, and in synchronization with the output of the grayscale data of each of the plurality of sections by the photoelectric conversion means, in the plurality of sections alternately in the first and second storage areas. And a means for writing a grayscale data set, wherein the first memory resets the stored contents thereof in synchronization with the output of the grayscale data set for each of the plurality of sections by the photoelectric conversion means. Means for holding the histogram relating to the grayscale data corresponding to a plurality of the band patterns in one section in the first memory, and the processing means includes the photoelectric conversion means. In synchronization with the output of each grayscale data set of the plurality of sections, the position of a point that represents the position of the arbitrary lane in each section is detected, and the detected one section Which is on a line connecting a first point representing the arbitrary lane position in the above and a second point representing the position of the arbitrary lane in the section one previous to the one section previously detected. The position of the band pattern is detected, the grayscale data for the band pattern at the detected position is read from the first and second regions, and the plurality of sections read from the second memory are read. 7. The device according to claim 6, which is means for detecting the position of the band pattern in the arbitrary lane along the first direction based on the grayscale data corresponding to the band pattern. Band alignment pattern automatic reader. 8. The first memory is sequentially stored in synchronization with the photoelectric conversion of light applied to the band pattern on one straight line in the second direction by the photoelectric conversion means. Data is output, and the adding means photoelectrically converts the light emitted by the photoelectric converting means and applied to the plurality of band patterns on one straight line in the second direction. 8. The automatic band array pattern reader according to claim 7, wherein the grayscale data and the data sequentially read from the first memory are sequentially input. 9. The automatic band sequence pattern reader according to any one of claims 6 to 8, wherein the band sequence pattern is a base band sequence pattern of a gene.