Method for positioning gene sequencing chip reaction well in image

Abstract

The invention relates to the field of gene sequencing, in particular to a method for positioning a gene sequencing chip reaction well in an image. The method comprises the steps that S1, a chip fluorescence picture is shot and obtained; and S2, carrying out standardized operation on the fluorescence picture to correct the fluorescence picture into a standard posture. And S3, respectively searching a vertical edge and a transverse edge of the chip, and determining a column index col of the vertical edge and a row index row of the transverse edge of the chip in the image so as to determine the position of the shot chip part in the whole chip in the image. And S4, positioning the reaction well and calculating the coordinates of the reaction well. And S5, according to the region identifier of the image, starting from the reaction well closest to the edge line, endowing each reaction well with a global reaction well index. According to the method, the standard row and column pixel spacing of the reaction wells and the actual row and column positions of the reaction wells do not need to be known, the reaction wells are obtained through self-adaptive calculation through a statistical method, the method can be suitable for biochips with staggered rows, and the function of deducing the global position index of the reaction wells in the whole chip from a local chip area is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of gene sequencing image recognition, and in particular to a method for positioning reaction wells of a gene sequencing chip in an image. Background Art

[0002] In gene sequencing, after a biochip is placed on a platform, a random angle will be generated between the biochip and the platform, and the generated random angle will cause coordinate deviation of the field of view (equivalent to the reaction well described in the present patent technology), so it needs to be corrected.

[0003] The commonly used correction method is as follows: Given the width and height of the field of view and the theoretical row and column positions of the field of view, calculate the theoretical abscissa and theoretical ordinate of the field of view; Given the width and height of the field of view and the actual row and column positions of the field of view, calculate the actual abscissa and actual ordinate of the field of view; Through the theoretical coordinates and actual coordinates of several fields of view, obtain the rotation matrix by the least squares method, and then calculate the rotation angle. The image can be corrected for coordinates through the rotation angle.

[0004] However, this method has certain limitations. It requires the known standard width and height of the field of view and the actual row and column positions of the field of view. Therefore, this method can only be applied to biochips with a standard grid distribution, cannot be applied to biochips with a staggered row distribution, and does not have the function of inferring the global position index of the reaction wells (fields of view) in the entire chip from a local chip area. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to design a method to solve the existing technical problems.

[0006] To solve the above technical problems, the present invention provides a method for positioning reaction wells of a gene sequencing chip in an image, specifically including the following steps:

[0007] Step S1: Take and obtain a fluorescence picture of the chip.

[0008] For the microchip of second-generation gene sequencing, the reaction wells are distributed in the chip according to certain rules. After adding a fluorescent reagent, a fluorescence reaction will occur in the reaction wells, presenting bright holes, and the brightness is related to the reaction degree.

[0009] The taken fluorescence picture cannot cover the entire chip under the premise of meeting the resolution. Without the condition based on calibration, based on 1 local fluorescence picture, it is necessary to obtain the global coordinate index of the reaction wells therein and the corresponding image pixel coordinates, so as to form a coordinate correspondence with the electrical signals synchronously obtained by the chip, and then analyze the correlation between the fluorescence intensity of each reaction well and the electrical signal.

[0010] Chip fluorescence images at the four corners of the chip (upper left corner, upper right corner, lower left corner, lower right corner), with the requirement that the fluorescence images contain two fluorescence edges. The reaction wells in the chip are circular and arranged and distributed according to a fixed rule, such as Figure 1 shown.

[0011] Step S2: Perform a normalization operation on the fluorescence image, and correct the fluorescence image to a standard posture by rotation.

[0012] Step S3: Search for the vertical edge and horizontal edge of the chip respectively, determine the column index col of the vertical edge of the chip in the image and the row index row of the horizontal edge, so as to determine the position of the chip part captured in the image in the whole chip (upper left corner, lower left corner, upper right corner, lower right corner).

[0013] Step S4: Locate the reaction wells and calculate the reaction well coordinates (X a , Y b ).

[0014] Step S5: According to the region identifier of the image (upper left corner, upper right corner, lower left corner, lower right corner), starting from the reaction well closest to the edge line, assign a global reaction well index (IDX_X a , IDX_Y b ) to each reaction well.

[0015] Further in the present invention, in step S2, the process of the fluorescence image normalization operation is: rotate the image at equal angles within a certain range, calculate the cumulative minimum gray value mz after each rotation, traverse all rotation angles, and correct at the angle with the minimum cumulative minimum gray value.

[0016] Further in the present invention, the calculation formula of the cumulative minimum gray value mz is:[[]]

[0017]

[0018] where z a,n is the gray value of the nth pixel in the a-th minimum gray value row, A is the number of minimum gray value rows in the image, and N is the number of pixel values in each row of the image (i.e., the image width);

[0019] The condition for the t-th row in the image to satisfy the minimum gray value row is:[[]]

[0020]

[0021] where y t,n is the gray value of the nth column in the t-th row of the image, span is the window size, and M is the number of pixel values in each column of the image (i.e., the image height).

[0022] Further in the present invention, in step S3, finding the vertical edges of the chip includes the following steps:

[0023] Step (1): Perform a sliding window operation along the width direction of the image to obtain a number of vertical bands;

[0024] Step (2): Calculate the average gray value c_μ of each vertical band i ;

[0025] Step (3): Calculate the standard deviation c_s of each vertical band i ;

[0026] Step (4): Calculate the coefficient of variation c_cv of each vertical band i ,

[0027] Step (5): Determine the column index col of the vertical edge:

[0028] col = argmin(c_cv1, c_cv2, c_cv3,..., c_cv i ), where i is the number of the vertical band.

[0029] Further in the present invention, in step S3, finding the horizontal edges of the chip includes the following steps:

[0030] Step (1): Perform a sliding window operation along the height direction of the image to obtain a number of horizontal bands;

[0031] Step (2): Calculate the average gray value r_μ of each horizontal band i ;

[0032] Step (3): Calculate the standard deviation r_s of each horizontal band i ;

[0033] Step (4): Calculate the coefficient of variation r_cv of each horizontal band i ,

[0034] Step (5): Determine the row index row of the horizontal edge:

[0035] row = argmin(r_cv1, r_cv2, r_cv3,..., r_cv i ). i is the number of the horizontal band.

[0036] Further in the present invention, step S4 specifically includes the following steps:

[0037] Step S41: Divide the reaction well image by rows and calculate the row coordinates of the reaction wells;

[0038] Step S42: Stitch the reaction well images of odd rows and even rows respectively;

[0039] Step S43: Split the stitched images of odd rows and even rows by columns respectively, and calculate the column coordinates of the reaction wells;

[0040] Step S44: Calculate the standard row spacing sta_h and column spacing sta_w of the reaction wells;

[0041] Step S45: Deduce the coordinates (X a , Y b ) of the reaction wells.

[0042] In the present invention, further, in Step S41, all reaction well parts are extracted from the image and segmented by rows. Specifically, during implementation, taking the rows with extremely low gray levels as the row segmentation lines, the images between two adjacent segmentation lines are segmented out as the reaction well images of one row, and I row segmentation lines row_dv i are obtained, where i = 1, 2, 3,..., I; I - 1 reaction well row images are obtained from the I row segmentation lines, and their row coordinates Y i are defined as follows: Y i = (row_dv i + row_dv i+1 ) / 2, i = 1, 2, 3,..., I - 1.

[0043] In the present invention, further, in Step S43, taking the columns with extremely low gray levels as the column segmentation lines:

[0044] For the odd - row image pic odd , J column segmentation lines are obtained, and the column segmentation lines are defined as col_odd_dv j , where j = 1, 2, 3,..., J; J - 1 reaction well column images are obtained from the J column segmentation lines, and their column coordinates X_odd j are defined as follows:

[0045] X_odd j = (col_odd_dv j + col_odd_dv j+1 ) / 2, j = 1, 2, 3,..., J - 1;

[0046] For the even - row image pic even , K column segmentation lines are obtained, and the column segmentation lines are defined as col_even_dv k , where k = 1, 2, 3,..., K; K - 1 reaction well column images are obtained from the K column segmentation lines, and their column coordinates X_even k are defined as follows:

[0047] X_evenk =(col_even_dv k +col_even_dv k+1 ) / 2, k = 1, 2, 3, ..., K - 1.

[0048] In the present invention, further, in step S44, the calculation formula for the standard spacing is:

[0049] The calculation formula for the standard row spacing sta_h of the reaction wells is:

[0050] where diff_Y i is the row spacing;

[0051] The calculation formula for the standard column spacing sta_w of the reaction wells is:

[0052]

[0053] where diff_X_odd j is the column spacing of the odd - row images, and diff_X_even k is the column spacing of the even - row images.

[0054] In the present invention, further, in step S45, the method for calculating the coordinates of the reaction wells is as follows: Let the coordinates of the reaction well at the origin of the odd rows be (X_odd1, Y1), and let the coordinates of the reaction well at the origin of the even rows be (X_even1, Y1). For the coordinates (X a , Y b ) of any reaction well, where a is the coordinate index of the current reaction well in the X - direction of the chip, and b is the coordinate index of the current reaction well in the Y - direction of the chip, there are:

[0055] When b is odd:

[0056] When b is even:

[0057] Advantages of the present invention:

[0058] (1) The method for positioning the reaction wells of the gene - sequencing chip in the image of the present invention does not require the known standard row - and - column pixel spacings of the reaction wells (equivalent to the standard width and height of the field of view), but is adaptively calculated through statistical methods.

[0059] (2) The method for positioning the reaction wells of the gene - sequencing chip in the image of the present invention does not require the actual row - and - column positions of the reaction wells (field of view), but is adaptively calculated through statistical methods.

[0060] (3) The method for positioning the reaction wells of the gene - sequencing chip in the image of the present invention can be applied to biochips with staggered - row distributions.

[0061] (4) The method for positioning the reaction wells of the gene sequencing chip of the present invention in the image has the function of inferring the global position index of the reaction wells (field of view) in a local chip area in the entire chip. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] The following further clarifies the specific embodiments of the present invention with reference to the drawings.

[0063] Figure 1 It is the arrangement distribution diagram of the reaction wells in the chip in the embodiment of the present invention.

[0064] Figure 2 It is the flow chart of the method for positioning the reaction wells of the gene sequencing chip of the present invention in the image.

[0065] Figure 3(a) is the fluorescence image before rotation in the embodiment of the present invention.

[0066] Figure 3(b) is the optimal image after rotation and correction to the standard posture in the embodiment of the present invention.

[0067] Figure 4 It is the fluorescence image for determining the chip edge in the embodiment of the present invention.

[0068] Figure 5 It is the schematic diagram of splitting the reaction well image row by row in the embodiment of the present invention.

[0069] Figure 6(a) is the schematic diagram after splicing the reaction well images of odd rows in the embodiment of the present invention.

[0070] Figure 6(b) is the schematic diagram of splitting the spliced odd-row image column by column in the embodiment of the present invention.

[0071] Figure 7(a) is the schematic diagram after splicing the reaction well images of even rows in the embodiment of the present invention.

[0072] Figure 7(b) is the schematic diagram of splitting the spliced even-row image column by column in the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0073] Embodiment 1

[0074] Combined with Figure 2 , the method for positioning the reaction wells of the gene sequencing chip of this embodiment in the image specifically includes the following steps:

[0075] Step S1: Photograph and obtain the chip fluorescence picture.

[0076] For the microchip of second-generation gene sequencing, the reaction wells are distributed in the chip according to certain rules. After adding the fluorescent reagent, a fluorescent reaction will occur in the reaction wells, presenting bright holes, and the brightness is related to the reaction degree.

[0077] The captured fluorescence images cannot cover the entire chip while meeting the resolution requirements. Without calibration, based on a single local fluorescence image, it is necessary to obtain the global coordinate indices of the reaction wells and their corresponding image pixel coordinates therein, so as to form a coordinate correspondence with the electrical signals synchronously acquired by the chip, and then analyze the correlation between the fluorescence intensity and the electrical signals of each reaction well.

[0078] Fluorescence images of the four corners of the chip (upper left, upper right, lower left, lower right), and it is required that the fluorescence images contain two fluorescence edges. The reaction wells in the chip are circular and arranged in a fixed pattern, such as Figure 1 shown.

[0079] Step S2: Perform a normalization operation on the fluorescence image and correct the fluorescence image to a standard pose by rotation.

[0080] In the present invention, the standard pose means that the connection line of the centers of the reaction wells in each row is horizontal. The captured fluorescence images are not necessarily in the standard pose, so it is necessary to rotate the image to correct it to the standard pose.

[0081] Preferably in this embodiment, in step S2, the process of normalizing the fluorescence image is as follows: Rotate the image at equal angles within a certain range, calculate the cumulative minimum gray value mz after each rotation, traverse all rotation angles, and correct it at the angle with the minimum cumulative minimum gray value.

[0082] If the pixel brightness values of each row in the image are added together, a one-dimensional sequence with a length equal to the height of the rotated image can be obtained. When rotated to the optimal angle, the data waveform of this one-dimensional sequence will show a phenomenon similar to many sine or cosine waveforms with equivalent amplitudes, and the peak distance and valley distance between adjacent sine or cosine waveforms are also approximately equal.

[0083] In two-dimensional space, rotation can be defined by a single angle θ. By convention, a positive angle represents counterclockwise rotation. The matrix for rotating the column vector of Cartesian coordinates counterclockwise about the origin is:

[0084]

[0085] Preferably in this embodiment, the calculation formula for the cumulative minimum gray value mz is:

[0086]

[0087] where z a,n is the gray value of the nth pixel in the a-th minimum gray row, A is the number of minimum gray rows in the image, and N is the number of pixel values in each row of the image (i.e., the width of the image).

[0088] The condition for the t-th row in the image to satisfy the minimum gray row is:

[0089]

[0090] where y t,n is the gray value of the t-th row and n-th column of the image, span is the window size, and M is the number of pixel values in each column of the image (i.e., the height of the image).

[0091] Specifically, in this embodiment, the fluorescence image before rotation is shown in Fig. 3(a), and the optimal image corrected to the standard posture by rotation is shown in Fig. 3(b).

[0092] Step S3: Search for the vertical edge and horizontal edge of the chip respectively, determine the column index col of the vertical edge of the chip in the image and the row index row of the horizontal edge, so as to determine the position (upper left corner, lower left corner, upper right corner, lower right corner) of the captured chip part in the full chip in the image.

[0093] Preferably, in this embodiment, in step S3, searching for the vertical edge of the chip includes the following steps:

[0094] Step (1): Perform a sliding window operation along the width direction of the image to obtain a number of vertical bands.

[0095] Specifically, in this embodiment, set the sliding window width to w pixels and slide along the x direction with a step size of 1. If M is the height of the image and N is the width of the image, then N - w + 1 vertical bands can be obtained.

[0096] Step (2): Calculate the average gray value c_μ of each vertical band i .

[0097] Specifically, in this embodiment, the average gray value c_μ i of each vertical band is calculated by the formula:

[0098]

[0099] where y m,n is the gray value of the m-th row and n-th column of the image.

[0100] Step (3): Calculate the standard deviation c_s of each vertical band i .

[0101] Specifically, in this embodiment, the standard deviation c_s i of each vertical band is calculated by the formula:

[0102]

[0103] where y m,n is the gray value of the m-th row and n-th column of the image.

[0104] Step (4): Calculate the coefficient of variation c_cv for each vertical strip i :

[0105]

[0106] Step (5): Determine the column index col of the vertical edge:

[0107] col = argmin(c_cv1, c_cv2, c_cv3,..., c_cv N-w+1 ).

[0108] Preferably in this embodiment, in step S3, finding the horizontal edges of the chip includes the following steps:

[0109] Step (1): Perform a sliding window operation along the height direction of the image to obtain a number of horizontal strips.

[0110] Specifically in this embodiment, set the sliding window height to h pixels and slide along the y direction with a step size of 1. M is the height of the image and N is the width of the image, then M - h + 1 horizontal strips can be obtained.

[0111] Step (2): Calculate the average gray value r_μ of each horizontal strip i .

[0112] Specifically in this embodiment, the average gray value r_μ i of each horizontal strip is calculated by the formula:

[0113]

[0114] where y m,n is the gray value of the m-th row and n-th column of the image.

[0115] Step (3): Calculate the standard deviation r_s of each horizontal strip i .

[0116] Specifically in this embodiment, the standard deviation r_s i of each horizontal strip is calculated by the formula:

[0117]

[0118] where y m,n is the gray value of the m-th row and n-th column of the image.

[0119] Step (4): Calculate the coefficient of variation r_cv of each horizontal strip i :

[0120]

[0121] Step (5): Determine the column index row of the horizontal edge:

[0122] row = argmin(r_cv1, r_cv2, r_cv3,..., r_cv M-h+1 )。

[0123] Specifically, in this embodiment, the values of the set sliding window widths w and h are both 5 pixels. According to the column index col of the vertical edge of the chip and the row index row of the horizontal edge in the image, the position of the captured chip part in the image within the entire chip can be determined, as shown in Figure 4 the figure.

[0124] Step S4: Locate the reaction well and calculate the reaction well coordinates (X a , Y b ).

[0125] Preferably, in this embodiment, step S4 specifically includes the following steps:

[0126] Step S41: Split the reaction well image by rows and calculate the row coordinates of the reaction well.

[0127] Preferably, in this embodiment, in step S41, all reaction well parts are extracted from the image and split by rows. Specifically, when implementing, the minimum gray level rows are used as the row splitting lines, and the image between two adjacent splitting lines is split out as an image of one row of reaction wells. I row splitting lines are obtained, and the row splitting lines are defined as row_dv i , i = 1, 2, 3,..., I; specifically as shown in Figure 5 the figure.

[0128] For the I row splitting lines, I - 1 row images of the reaction wells can be obtained, and their row coordinates are defined as Y i ,

[0129] Y i = (row_dv i + row_dv i+1 ) / 2, i = 1, 2, 3,..., I - 1.

[0130] Step S42: Stitch the reaction well images of the odd rows and even rows respectively.

[0131] Specifically, in this embodiment, the reaction well images of the odd rows are stitched into an odd row image pic odd , as specifically shown in Figure 6(a); the reaction well images of the even rows are stitched into an even row image pic even , as specifically shown in Figure 7(a).

[0132] Step S43: Split the odd row stitched image and the even row stitched image by columns respectively and calculate the column coordinates of the reaction well.

[0133] Preferably, in this embodiment, when specifically implementing step S43, the minimum gray level columns are used as column division lines:

[0134] For the odd - numbered row image pic odd , J column division lines are obtained, and the column division lines are defined as col_odd_dv j , j = 1, 2, 3,..., J; as specifically shown in Figure 6(b).

[0135] J column division lines obtain J - 1 reaction well column images, and their column coordinates are defined as X_odd j :

[0136] X_odd j =(col_odd_dv j +col_odd_dv j+1 ) / 2, j = 1, 2, 3,..., J - 1;

[0137] For the even - numbered row image pic even , K column division lines are obtained, and the column division lines are defined as col_even_dv k , k = 1, 2, 3,..., K; as specifically shown in Figure 7(b).

[0138] K column division lines obtain K - 1 reaction well column images, and their column coordinates are defined as X_even k :

[0139] X_even k =(col_even_dv k +col_even_dv k+1 ) / 2, k = 1, 2, 3,..., K - 1.

[0140] Step S44: Calculate the standard reaction well row spacing sta_h and column spacing sta_w.

[0141] Preferably, in this embodiment, in step S44, the calculation formula for the standard spacing is:

[0142] The calculation formula for the standard reaction well row spacing sta_h is:

[0143] Among them, diff_Y i is the row spacing.

[0144] Specifically, in this embodiment, the calculation formula for the row spacing diff_Y i is:

[0145] diff_Y i =|Y i+1 -Y i, i = 1, 2, 3, ..., I - 2.

[0146] The calculation formula for the standard reaction well column spacing sta_w is:

[0147]

[0148] Wherein, diff_X_odd j is the column spacing of the odd - numbered row images, and diff_X_even k is the column spacing of the even - numbered row images.

[0149] Specifically in this embodiment, the calculation formula for the column spacing diff_X_odd of the odd - numbered row images j is:

[0150]

[0151] Specifically in this embodiment, the calculation formula for the column spacing diff_X_even of the even - numbered row images k is:

[0152] diff_X_even k = |X_even k+1 - X_even k |, k = 1, 2, 3, ..., K - 2.

[0153] Step S45: Deduce the coordinates (X a , Y b ) of the reaction well.

[0154] Preferably in this embodiment, in step S45, the method for deducing the coordinates of the reaction well is: Let the coordinates of the reaction well at the origin of the odd - numbered rows be (X_odd1, Y1), and let the coordinates of the reaction well at the origin of the even - numbered rows be (X_even1, Y1). For the coordinates (X a , Y b ) of any reaction well, where a is the coordinate index of the current reaction well in the X - direction of the chip, and b is the coordinate index of the current reaction well in the Y - direction of the chip, there are:

[0155] When b is odd:

[0156] When b is even:

[0157] Specifically in this embodiment, (X a , Y b ) needs to satisfy being inside the chip edge line. Specifically in implementation, there are:

[0158]

[0159] Coordinate points that do not meet the requirements will be deleted.

[0160] Step S5: According to the reaction wells determined within the specified range, the image coordinates (X a , Y b ) of which have been determined in the previous step, based on the regional identification of the image (upper left corner, upper right corner, lower left corner, lower right corner), starting from the reaction well closest to the edge line, a global reaction well index (IDX_X a , IDX_Y b ) can be assigned to each reaction well.

[0161] Many specific details have been set forth in the above description to facilitate a full understanding of the present invention. However, the above description is only a preferred embodiment of the present invention, and the present invention can be implemented in many other ways different from those described herein. Therefore, the present invention is not limited by the specific embodiments disclosed above. At the same time, any person skilled in the art can make many possible changes and modifications to the technical solution of the present invention, or modify it into an equivalent embodiment with equivalent changes, without departing from the scope of the technical solution of the present invention. All simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A method for positioning a reaction well of a gene sequencing chip in an image, characterized in that: It includes the following steps: Step S1: Take and obtain the fluorescence image of the chip; Step S2: Perform standardization operation on the fluorescence image, and correct the fluorescence image to a standard posture by rotation; Step S3: Search for the vertical edge and horizontal edge of the chip respectively, determine the column index col of the vertical edge of the chip in the image and the row index row of the horizontal edge, so as to determine the position of the chip part captured in the image in the whole chip; Step S4: Locate the reaction well and calculate the coordinates (X a , Y b ) of the reaction well; Step S5: According to the regional identification of the image, starting from the reaction well closest to the edge line, assign a global reaction well index (IDX_X a , IDX_Y b ) to each reaction well.

2. The method for positioning the gene sequencing chip reaction well in the image according to claim 1, wherein: In Step S2, the process of the fluorescence image standardization operation is as follows: Rotate the image at equal angles within a certain range, calculate the cumulative minimum gray value mz after each rotation, traverse all rotation angles, and correct at the angle with the minimum cumulative minimum gray value.

3. The method for positioning the gene sequencing chip reaction well in the image according to claim 2, wherein: The calculation formula of the cumulative minimum gray value mz is: where z a,n is the gray value of the nth pixel in the a-th minimum gray level row, A is the number of minimum gray level rows in the image, and N is the number of pixel values in each row of the image; The condition for the t-th row in the image to satisfy the minimum gray value row is: t = span + 1, span + 2, span + 3,..., span + (M - 2*span), where y t,n is the gray value of the nth column in the tth row of the image, span is the window size, and M is the number of pixel values in each column of the image.

4. The method for positioning a gene sequencing chip reaction well in an image according to claim 1, wherein: In Step S3, searching for the vertical edge of the chip includes the following steps: Step (1): Perform a sliding window operation along the width direction of the image to obtain several vertical bands; Step (2): Calculate the average gray value c_μ of each vertical band i ; Step (3): Calculate the standard deviation c_s of each vertical strip i ; Step (4): Calculate the coefficient of variation c_cv of each vertical band i , Step (5): Determine the column index col of the vertical edge: col = argmin(c_cv1, c_cv2, c_cv3,..., c_cv i ), where i is the number of the vertical band.

5. The method for positioning the gene sequencing chip reaction well in the image according to claim 1, characterized in that: In Step S3, searching for the horizontal edge of the chip includes the following steps: Step (1): Perform a sliding window operation along the height direction of the image to obtain several horizontal bands; Step (2): Calculate the average gray value \(r_{\mu}\) of each horizontal band i ; Step (3): Calculate the standard deviation r_si of each horizontal band; Step (4): Calculate the coefficient of variation \(r_{cv}\) for each horizontal band i , Step (5): Determine the column index row of the horizontal edge: row = argmin(r_cv1, r_cv2, r_cv3,..., r_cv i ). i is the number of the horizontal band.

6. The method for positioning the reaction wells of the gene sequencing chip in the image according to claim 1, characterized in that: In Step S4, it specifically includes the following steps: Step S41: Divide the reaction well image by rows and calculate the row coordinates of the reaction wells; Step S42: Stitch the reaction well images of odd rows and even rows respectively; Step S43: Divide the odd-row stitched image and the even-row stitched image by columns respectively and calculate the column coordinates of the reaction wells; Step S44: Calculate the standard reaction well row spacing sta_h and column spacing sta_w; Step S45: Calculate the coordinates (X a , Y b ) of the reaction well.

7. The method for positioning the gene sequencing chip reaction well in the image according to claim 6, characterized in that: In step S41, taking the minimum gray level row as the row division line, the image between two adjacent division lines is segmented out as a row of reaction well images, and I row division lines row_dv are obtained. i , where i = 1, 2, 3,..., I; I row division lines obtain I - 1 row images of reaction wells, and their row coordinates Y i are defined as follows: Y i =(row_dv i +row_dv i+1 ) / 2, where i = 1, 2, 3,..., I - 1.

8. The method for positioning the reaction wells of the gene sequencing chip in the image according to claim 6, characterized in that: In Step S43, use the minimum gray value column as the column division line: For the odd - row image pic odd , obtain J column division lines, and the column division lines are defined as col_odd_dv j , j = 1, 2, 3,..., J; J column division lines obtain J - 1 reaction well column images, and their column coordinates X_odd j are defined as follows: X_odd j = (col_odd_dv j + col_odd_dv j+1 ) / 2, j = 1, 2, 3, ..., J - 1; For the even - numbered row image pic even , obtain K column division lines, and the column division lines are defined as col_even_dv k , k = 1, 2, 3,..., K; K column division lines obtain K - 1 reaction well column images, and their column coordinates X_even k are defined as follows: X_even k = (col_even_dv k + col_even_dv k+1 ) / 2, k = 1, 2, 3,..., K - 1。 9. The method for positioning a gene sequencing chip reaction well in an image according to claim 8, wherein: In Step S44, the calculation formula of the standard spacing is: The calculation formula of the standard reaction well row spacing sta_h is: Among them, diff_Y i is the line spacing; The calculation formula of the standard reaction well column spacing sta_w is: Among them, diff_X_odd j is the column pitch of the odd-row images, and diff_X_even k is the column pitch of the even-row images.

10. The method for positioning the reaction wells of the gene sequencing chip in the image according to claim 9, wherein: In step S45, the method for calculating the coordinates of the reaction wells is as follows: Let the coordinates of the reaction well at the origin of the odd rows be (X_odd1, Y1), and let the coordinates of the reaction well at the origin of the even rows be (X_even1, Y1). For the coordinates (X a , Y b ) of any reaction well, there is: When b is odd: When b is an even number:

Images