Image signal high-speed analysis method and system and medium

Abstract

The invention discloses an image signal high-speed analysis method and system and a medium, and the analysis method comprises the steps: adjusting the pose of a sensor until a plurality of light sensing lines of the sensor are perpendicular to the long axis direction of a laser image, so as to enable image data collected by the light sensing lines to directly serve as sampling data; sampling data of different areas of the laser image in the long axis direction are collected in a grouped mode so as to correspond to a plurality of defocusing amount sampling points; effective data in all the sampling data are screened, so that a plurality of effective sampling points are obtained, and the defocusing amount corresponding to each effective sampling point is obtained through calculation; and analyzing the standard deviation of the defocusing amounts of the effective sampling points, and screening to obtain a plurality of effective defocusing amounts so as to output the final defocusing amount. According to the method, the defocusing amount sampling rates in the near-focus state and the far-focus state can be kept consistent, the sampling data participating in defocusing amount calculation can be dynamically adjusted by analyzing the effective data in all the sampling data and the standard deviation of the defocusing amount of the effective sampling points, and the calculation amount is reduced.

Description

Technical Field

[0001] The present invention belongs to the field of machine vision, and in particular relates to a method, system and medium for high-speed analysis of image signals. Background Art

[0002] The current mainstream microscope autofocus methods can be divided into passive and active.

[0003] Passive methods often use the clarity of the microscopic image collected by the detection camera to directly calculate the defocus amount of the object surface, and then feed it back to the Z axis to move the objective lens up and down for focusing. The defocus amount sampling rate of this type of method is limited by the image frame rate of the detection camera, which is generally only in the hundreds of Hz level.

[0004] The active method usually inserts an additional sensor that can detect the amount of defocus in the microscope optical path. For example, a coaxial laser light source is used to project the laser directly onto the sample surface. The image sensor that receives the signal analyzes the state of the laser reflected back from the sample surface to determine the size and direction of the defocus of the objective lens. The defocus of the objective lens can be detected in real time, and the defocus signal is transmitted to the Z-axis actuator for defocus compensation to ensure that the target object surface is always within the depth of field of the objective lens and the microscopic imaging is always clear. Therefore, the active laser autofocus method has the advantages of high signal-to-noise ratio, not easily affected by ambient light, large focusing range, and fast focusing speed.

[0005] The current mainstream active laser autofocus method usually uses multiple sampling image segments or multiple sampling points to sample the defocus signal to enhance the focus stability. Here, an area array image sensor is required to perform high-speed analysis of multi-point / segment laser image signals. Traditional area array image sensors need to read row data row by row, and can only perform sampling point analysis in each column direction after all row data are read, so as to realize the defocus analysis of multiple sampling points of the laser image. Therefore, the sampling rate of the defocus signal is limited by the frame rate of the area array image sensor. The maximum sampling rate is generally in the kHz level, and the sampling rate at the far focus will be further reduced to the hundreds of Hz level as the laser image signal is broadened. With the development of motion control technology, there are Z-axis actuators with a response bandwidth of kHz level, such as piezoelectric ceramic Z-axis actuators. This kHz high response bandwidth Z-axis actuator can be used to solve the focus tracking needs of target surfaces with high-frequency fluctuations. Accordingly, the defocus sampling rate of the focus sensor needs to reach more than 10kHz to achieve stable focus feedback.

[0006] Therefore, in order to solve the problem that the defocus sampling rate is limited by the frame rate of the area array image sensor and thus leads to a low sampling rate, the present invention provides a method, system and medium for high-speed analysis of image signals. Summary of the invention

[0007] The object of the present invention is to overcome the above problems existing in the prior art, and to provide an image signal high-speed analysis method, system and medium.

[0008] To achieve the above technical object and reach the above technical effect, the present invention is realized through the following technical solutions: An image signal high-speed analysis method, which analyzes the acquisition data of the laser image in the autofocus system to output the defocus amount. The analysis method includes: Adjust the pose of the sensor until several sensing lines of the sensor are perpendicular to the long axis direction of the laser image, so that the image data collected by the sensing lines can be directly used as sampling data; Group and collect the sampling data of different regions along the long axis direction of the laser image, so as to respectively correspond to several defocus amount sampling points; Screen the valid data among all the sampling data, so as to obtain several valid sampling points, and calculate the defocus amount corresponding to each valid sampling point; Analyze the standard deviation of the defocus amounts of the valid sampling points, and screen to obtain several valid defocus amounts, so as to output the final defocus amount; Wherein, the acquisition ranges of several sensing lines of the sensor cover the entire laser image.

[0009] Further, after grouping and collecting the sampling data of different regions along the long axis direction of the laser image, when reading the sampling data corresponding to the current region, synchronously analyze whether the sampling data corresponding to the previous region is valid data.

[0010] Further, screening the valid data among all the sampling data includes: calculating the maximum gray value in each group of sampling data, and analyzing whether the maximum gray value is greater than the gray threshold. If so, it is judged as valid data, otherwise it is invalid data.

[0011] Further, the defocus amount corresponding to each valid sampling point is obtained by calculating the offset of the gray centroid of each valid sampling point from the focus reference zero point.

[0012] Further, the valid defocus amount is the defocus amount of the valid sampling points within the range corresponding to the standard deviation.

[0013] Further, the final defocus amount is the average value of all valid defocus amounts.

[0014] Further, the sensor is a multi-line linear image sensor, and the height of the multi-line linear image sensor corresponds to the long axis span of the laser image.

[0015] Further, the sensor is a area array image sensor, and includes several regions of interest corresponding to each group of sampling data. Each region of interest corresponds to a single row of sensing lines or multiple rows of continuous sensing lines.

[0016] The present invention also provides an image signal high-speed analysis system, comprising: A posture adjustment module is used to adjust the posture of the sensor until several sensitive light lines of the sensor are perpendicular to the long axis direction of the laser image, so that the image data collected by the sensitive light lines can be directly used as sampling data; A grouping acquisition module is used to collect sampling data of different areas of the laser image along the long axis direction in groups, so as to correspond to a number of defocus amount sampling points respectively; An effective analysis module is used to screen effective data from all sampled data, thereby obtaining a number of effective sampling points, and calculating the defocus amount corresponding to each effective sampling point; The defocus analysis module is used to analyze the standard deviation of the defocus amount of the effective sampling points, and screen out a number of effective defocus amounts to output the final defocus amount.

[0017] The present invention also provides a computer-readable storage medium, comprising a computer program, wherein the computer program implements the above-mentioned analysis method when executed by a processor.

[0018] The beneficial effects of the present invention are: (1) The present invention makes the long axis direction of the light-sensitive line perpendicular to the laser image, so that the image data collected by the light-sensitive line will directly correspond to the sampling data. Therefore, the corresponding sampling point can be calculated after the corresponding data of the current light-sensitive line is read, and the corresponding data of the next light-sensitive line can be read synchronously. This not only eliminates the resource consumption of the dynamic ROI, but also eliminates the need to calculate each sampling point after all the corresponding data of all light-sensitive lines are read. It can realize reading and calculation at the same time, ensuring that the defocus sampling rate in the near-focus and far-focus states remains consistent, thereby improving the stability of the working rate of the focusing system. In actual work, only a small amount of The defocus amount can be calculated by the sensitive light line, such as K=10, and the total number of pixel rows in the ROI area is usually higher than 20. It can be seen that compared with the original ROI method, the laser image provided by the present invention has a lower data volume, and the rates of transmission to the processing chip and calculation of the defocus amount are faster, thereby further improving the sampling rate; by flexibly collecting sampling data of different areas of the laser image along the long axis direction by grouping, a one-to-one correspondence between the defocus amount sampling points and the sensitive light lines is achieved; by analyzing the valid data in all the sampling data and the standard deviation of the defocus amount of the valid sampling points, the sampling data involved in the defocus amount calculation can be dynamically adjusted to reduce the amount of calculation.

[0019] (2) The present invention can select a multi-line array sensor to significantly reduce the amount of data compared to an area array image sensor, thereby increasing the defocus sampling rate to above 10 kHz, solving the problem that a conventional single-line array sensor can only achieve high-speed acquisition of a single-point defocus signal but lacks focusing stability; it can also be implemented using multiple regions of interest of an area array image sensor. By selecting multiple rows of pixel signals of the area array image sensor for output, the amount of data can be reduced, and high-speed, multi-point defocus sampling can be achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings: Figure 1 is a flow chart of the analysis method in the present invention; Figure 2 is a schematic diagram of the optical path of the autofocus system in the present invention; Figure 3 It is a schematic diagram of the beam cross section corresponding to different object distances in the automatic focusing system of the present invention; Figure 4 It is a schematic diagram of a laser image of a conventional area array image sensor in an autofocus system; Figure 5 is an image schematic diagram of a laser image of a multi-line array image sensor in an autofocus system; Figure 6 It is a schematic diagram of an image of a laser image in an autofocus system after the area array image sensor is set using multiple regions of interest; Figure 7 It is a structural schematic diagram of the image collection device in the present invention; Figure 8 It is a structural block diagram of the analysis system in the present invention.

[0021] In the figure: 1-laser; 2-cylindrical lens; 3-baffle; 4-first beam splitter; 5-second beam splitter; 6-microscope objective; 7-motor; 8-focusing lens; 9-image sensor; 10-object to be measured. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Example 1

[0023] This embodiment first provides a high-speed image signal analysis method, which analyzes the collected data of the laser image in the auto-focus system to output the defocus amount, wherein the focus optical path of the auto-focus system is as follows: Figure 2 As shown, the details are as follows: The autofocus system includes: a laser 1, a cylindrical lens 2, a baffle 3, a first beam splitter 4, a second beam splitter 5, a microscope objective 6, a motor 7, a focus lens 8, an image sensor 9, and a measured object 10. The circular parallel light emitted by the laser 1 enters the cylindrical lens 2, and after being modulated by the cylindrical lens 2, becomes an asymmetric light beam that is collimated in a direction parallel to the paper surface and diverges in a direction perpendicular to the paper surface; after passing through the baffle 3, the asymmetric light beam will lose half of its energy and propagate only on one side of the optical axis, and after being reflected by the first beam splitter 4, it reaches the second beam splitter 5, and after being reflected by the second beam splitter 5, it enters one side of the microscope objective 6, and is projected on the surface of the measured object 10 by one side of the microscope objective 6. Due to the reflection of the surface of the measured object, the laser is collected again by the other side of the microscope objective 6, and after being continuously reflected by the second beam splitter 5 and refracted by the first beam splitter 4, it reaches the focus lens 8, and after being converged by the focus lens 8, it is received by the image sensor 9. The defocus amount of the microscope objective lens 6 is calculated according to the state of the light spot received by the image sensor 9, and the defocus amount is converted into a control signal of the motor 7, and the motor 7 is driven to drive the microscope objective lens 6 to move, thereby realizing automatic focusing.

[0024] Assume that the focal length of the microscope objective lens 6 is f and the object distance is L; when L increases from less than f to greater than f, the cross-sectional shape of the light beam passing through the microscope objective lens 6 changes, as shown in FIG. Figure 3 As shown, when L is less than f, the cross section of the light beam is an upper semi-ellipse, and the larger L is, the smaller the minor axis length of the ellipse is. When L is equal to f, in the direction parallel to the paper surface, the light beam incident on the microscope objective lens 6 is parallel light, which converges to a point after being modulated by the microscope objective lens 6. In the direction perpendicular to the paper surface, the light beam incident on the microscope objective lens 6 is divergent light, and is a line segment at the focal position of the microscope objective lens 6. Therefore, the cross section of the light beam is a line with a line width of l. When L is greater than f, the cross section of the light beam is a lower semi-ellipse, and the larger L is, the larger the minor axis length of the ellipse is.

[0025] like Figure 4 FIG. 1 is a schematic diagram of a laser image of a conventional area array image sensor in an autofocus system. Figure 2The image sensor 9 in the image sensor uses an area array image sensor to process the laser image signal reflected from the sample surface. The area array image sensor has N rows of pixels (r1, r2...rN), and the row direction of the area array image sensor is parallel to the long axis direction of the laser image. When calculating the defocus amount of a traditional laser image, it is necessary to accumulate and collect the grayscale values ​​of all pixels in the entire laser image area and calculate the corresponding defocus amount. In order to reduce the amount of calculation and increase the sampling rate, the laser image can be divided into several sub-images by column. The laser line signal in each sub-image can be processed independently, and the defocus amount data of K sampling points (s1, s2...sK) can be calculated.

[0026] Depend on Figure 3 It can be seen that near the focus of the objective lens, that is, the focus reference zero point, the width of the laser image in the short axis direction is small and the laser line is thin; away from the focus reference zero point, the laser image will be widened and the laser line will be thick. Therefore, near the focus reference zero point, only a smaller number of lines of image area size are needed to fully capture the laser image signal; away from the focus reference zero point, a larger number of lines of image area size are required to capture the complete laser image signal. High-speed area array image sensors need to use global shutter technology. Global shutter area array image sensors have the characteristics of fewer output lines and a higher image frame rate. Therefore, Figure 3 The corresponding laser image signal processing method can use dynamic ROI (region of interest) technology to reduce the number of output image lines near the focus reference zero point to achieve the highest image frame rate, that is, the defocus sampling rate. However, when the defocus is far, the number of image lines increases, the image frame rate decreases, the defocus sampling rate decreases, and the focusing speed decreases.

[0027] Therefore, it can be seen that even if the dynamic ROI technology is used to output only the region of interest corresponding to the laser image, it is still necessary to first read several rows of data corresponding to the dynamic ROI row by row, and then perform sampling point analysis in each column direction after all rows of data have been read, thereby realizing the defocus analysis of multiple sampling points of the laser image.

[0028] In order to further improve the processing speed of laser image signals, Figure 1 As shown, the analysis methods include: Step 1: Adjust the sensor posture until the sensor's several light-sensitive lines are perpendicular to the long axis direction of the laser image, so that the image data collected by the light-sensitive lines can be directly used as sampling data, wherein the collection range of the sensor's several light-sensitive lines covers the entire laser image.

[0029] The multiple light-sensitive lines of the sensor are perpendicular to the long axis of the line laser, so that the multiple light-sensitive lines can directly divide the line laser image signal. For example, a single line of light-sensitive lines corresponds to a defocus sampling point. The line data formed by each light-sensitive line will be sent to the processing chip for reading and processing. At this time, the processing chip can calculate the corresponding sampling point after reading the current line of data, and can also read the next line of data synchronously. When all the line data are read, the calculation of all sampling points can be completed. Figure 4 In comparison, it not only eliminates the resource consumption of dynamic ROI, but also eliminates the need to wait until all row data are read before calculating the sampling points in all column directions. The row data reading of the sensor is combined with the sampling point calculation, so that the collected data of each sensitive line can be directly used as the sampling data, which effectively improves the sampling efficiency.

[0030] Step 2: Collect sampling data of different areas of the laser image along the long axis direction in groups to correspond to a number of defocus sampling points respectively.

[0031] As a specific implementation of the present invention, each group of sampling data corresponds to a single line of sensitive light or multiple lines of continuous sensitive light. A single line of sensitive light can correspond to one defocus amount sampling point, or multiple lines of continuous sensitive light can correspond to one defocus amount sampling point.

[0032] like Figure 5 As shown, each pixel column (that is, the sensitive light line at this time) collects a part of the sub-image of the line laser segment, corresponding to a defocus sampling point. K defocus sampling points require that the linear array image sensor has K sensitive light lines. Multi-line linear array sensors can achieve a defocus sampling rate of 10kHz or even 100kHz, and can maintain the maximum line frequency in both near-focus and far-focus states without performing dynamic ROI operations. With a high-response bandwidth Z-axis actuator, ultra-fast focus response speed can be achieved, improving the adaptability of focus tracking on surfaces with large surface height difference change rates.

[0033] As a specific implementation of the present invention, after collecting the sampling data of different areas of the laser image along the long axis direction in groups, when reading the sampling data corresponding to the current area, it is synchronously analyzed whether the sampling data corresponding to the previous area is valid data.

[0034] The corresponding sampling points can be calculated after the corresponding data of the current sensitive line is read, and the corresponding data of the next sensitive line can be read synchronously. This not only eliminates the resource consumption of the dynamic ROI, but also eliminates the need to wait until the corresponding data of all sensitive lines are read before calculating each sampling point. It can realize reading and calculation at the same time, ensuring that the defocus sampling rate in the near focus and far focus states remains consistent, thereby improving the stability of the focusing system's working speed.

[0035] On the basis of canceling the dynamic ROI, the present invention still retains the advantage of reducing the amount of calculation, cleverly sets a one-to-one correspondence between the defocus amount sampling points and the light-sensitive lines, and further solves the defect of the dynamic ROI with a large amount of data when far-focus, and always maintains a stable defocus amount sampling rate in the near-focus and far-focus states. Not only that, since the corresponding sampling points can be calculated after each reading of the corresponding data of the current light-sensitive line, only a small amount of cache space for caching image data is needed to meet the cache needs, which significantly reduces resource consumption even compared with the dynamic ROI.

[0036] Since the sampling rate is more than 10 times higher than that of conventional area array image sensor detection solutions, the computing resources required for high-speed image processing per unit time will also increase synchronously. In order to make full use of the high line frequency of the linear array sensor, in the process of using multiple sampling points to calculate the defocus amount, according to the morphological information of the laser contour image received by the image sensor, the effective pixel columns can be screened, and the number of pixel columns involved in the defocus amount calculation can be dynamically adjusted to reduce the amount of calculation. The steps for pixel column screening and defocus amount calculation are as follows: Step 3: Filter valid data from all sampled data to obtain a number of valid sampling points, so as to calculate the defocus value corresponding to each valid sampling point.

[0037] As a specific implementation of valid data analysis, screening valid data from all sampled data includes: calculating the maximum grayscale value in each group of sampled data to analyze whether the maximum grayscale value is greater than the grayscale threshold, if so, it is judged as valid data, otherwise it is invalid data, as follows: like Figure 5 As shown, the current image frame is first analyzed to calculate the maximum grayscale value Gmax_i of each column of K pixel columns (c1, c2...cK), i=1, 2,..., K. If Gmax_i is greater than the grayscale threshold G_valid of the valid laser image signal, the data of the corresponding pixel column is judged to be valid, otherwise it is invalid, thereby obtaining M valid pixel columns c_i.

[0038] In order to calculate the defocus amount corresponding to each valid sampling point, the offset between the grayscale centroid of each valid sampling point and the focus reference zero point can be calculated as follows: Continue to process M valid pixel columns c_i and calculate the grayscale centroid P_i of each valid pixel column:

[0039] Get the defocus value of M effective sampling points

[0040] Where, j=1,2,...,n; G i(j) is the grayscale value of different positions of the effective sampling point, i=1,2,...,M; P_ref is the pixel coordinate of the focus reference zero point; A_pf is the conversion factor between pixel coordinate and defocus amount.

[0041] Step 4: Analyze the standard deviation of the defocus values ​​at the valid sampling points, and screen out several valid defocus values ​​to output the final defocus value.

[0042] In order to screen and obtain a number of effective defocus values, as a specific implementation of the present invention, the effective defocus value is the defocus value of the effective sampling point within the range corresponding to the standard deviation, which is as follows: According to the standard deviation σ of the defocus value F_i of the effective sampling point F , we can further select L more reasonable effective defocus values ​​Fv_x to satisfy:

[0043] Where x=1,2,...,L.

[0044] After L effective defocus amounts Fv_x are screened, as a specific implementation of the present invention, the final defocus amount is the mean of all effective defocus amounts, that is, the mean Fv_mean of Fv_x is calculated as the final single-sampling defocus amount output.

[0045] As a specific embodiment of the present invention, the sensor can be a multi-line array image sensor, and the height of the multi-line array image sensor corresponds to the long axis span of the laser image; it can also be an area array image sensor, and includes a number of regions of interest corresponding to each group of sampling data, and each region of interest corresponds to a single line of sensitive light or multiple lines of continuous sensitive light.

[0046] When using an area array image sensor, a global shutter area array image sensor can be used for simulation, and multiple regions of interest of the global shutter area array image sensor need to be enabled. Figure 6 As shown, the area array image sensor is rotated 90 degrees, and the long side direction of each of the K regions of interest (a1, a2...aK) is perpendicular to the long axis direction of the line laser focus image signal. Each region of interest can select a single row or multiple continuous rows of laser images for processing. Due to the reduction in the amount of image data to be processed, the frame rate of the area array image sensor will be greatly improved. When each region of interest contains only one row of pixels, the frame rate of the image sensor reaches the highest. Similarly, under this relative position relationship between the image sensor and the laser image signal, it is no longer necessary to perform dynamic ROI operations. The microscopic focusing device can maintain the maximum defocus sampling rate in both near-focus and far-focus states. Example 2

[0047] Another aspect of the present invention is to provide a sensor image acquisition device for acquiring a semi-elliptical laser image projected on the surface of an object to be measured in an autofocus system to calculate a defocus amount corresponding to the laser image, the device comprising: At least two linear array sensors are used to collect samples of at least two areas in the long axis direction of the laser image according to preset defocus amount sampling points; At least two one-dimensional optical fiber arrays are arranged along the long axis direction of the laser image and perpendicular to the long axis of the laser image, and are used to transmit the image information of each defocus amount sampling point in the long axis direction of the laser image to the linear array sensor, and the area for collecting the laser image by the linear array sensor is expanded by adjusting the spacing between the one-dimensional optical fiber arrays; A fixture is used to independently fix the input ends of the one-dimensional optical fiber array at preset intervals to correspond to the defocus amount sampling points, and to align and couple the output ends of the one-dimensional optical fiber array with the corresponding linear array sensor; Wherein, the linear array sensor is a single-row pixel linear array image sensor, and the defocus amount corresponding to the laser image is calculated by analyzing the defocus amounts of effective sampling points among all defocus amount sampling points.

[0048] In order to fully collect laser image signals with a length of millimeters, a conventional multi-line array image sensor, such as a typical linear array image sensor with a 2K resolution, 256 lines, and a pixel size of 3.5um, can only cover laser images with a length of less than 1mm, and there is a partial signal loss. Moreover, the defocus amount sampling of 256 light-sensitive lines at a frequency of 10kHz at the same time requires too much calculation. In order to solve the problem of insufficient sampling area of ​​this multi-line array image sensor, a linear array image sensor encapsulated with several sparsely spaced light-sensitive lines can be used for processing, where the light-sensitive lines are spaced D_line. K light-sensitive lines can sample laser image signals with a length of (K-1)×D_line, and D_line can be adjusted to adapt to laser lines of different lengths. However, this special linear array image sensor usually needs to be customized. The present invention is based on a two-dimensional transmission fiber array and several single-line array image sensors spliced ​​into a multi-line array image sensor, where a single-line array image sensor is a single-row pixel array image sensor, as follows: like Figure 7 As shown, a two-dimensional optical fiber array is used to relay the laser image signal at the focal plane of the focusing lens in the optical path of the laser focus sensor to multiple independent single-line array image sensors at the back end.

[0049] First, the two-dimensional fiber array is composed of K one-dimensional fiber arrays FA_i, i = 1, 2, ..., K. Each FA_i consists of N one-dimensional fiber arrays with a core diameter of several microns. f N adjacent optical fibers are spliced ​​together. fThe larger the value, the stronger the far-focus laser image signal that can be collected, and the larger the focusing range of the laser focus sensor. As a specific embodiment of the present invention, each one-dimensional optical fiber array is evenly spaced along the long axis direction of the laser image to uniformly collect image information of the laser image.

[0050] K one-dimensional fiber arrays correspond to K defocus sampling points. When the number of sampling points is constant, the uniformity and integrity of the laser line signal sampling can be guaranteed by adjusting the interval D_line of the one-dimensional fiber array so that (K-1)×D_line is equal to the length of the laser image.

[0051] After the laser image signals are collected at the entrances of K one-dimensional fiber arrays, they are segmented and sampled, and flexibly transmitted to the K single-line array image sensors at the back end via the fiber bundle assembly line. The outlet of the i-th one-dimensional fiber array FA_i and the sensitive light of the single-line array image sensor LS_i are directly aligned and coupled. The coupling method includes mechanical structure limit alignment and fixation, and the mechanical three-axis adjustment structure pre-aligns the sensitive light before encapsulating and bonding the FA_i outlet and the linear array image sensor. Since the defocus signal extraction only requires the centroid coordinates of the illumination distribution of the laser image signal, and no image details are required, the fiber array outlet of FA_i and the sensitive light of the linear array image sensor LS_i do not need to be precisely aligned less than the pixel size, and the lateral offset of several pixels has no obvious effect on the image signal processing.

[0052] The fixture is divided into a first fixing component and a second fixing component. The first fixing component is used to independently fix the input end of the one-dimensional optical fiber array to correspond to the defocus sampling point. For example, it includes a plurality of independently arranged adjustable substrates. The adjustable substrates have a plurality of optical fiber grooves or holes for installing optical fibers arrayed along a straight line direction. The optical fiber can be fixed by clamping or bonding so that the optical fiber is aligned with the defocus sampling point, such as perpendicular to the image area corresponding to the defocus sampling point; the second fixing component is used to complete the alignment and coupling of the output end of the one-dimensional optical fiber array with the corresponding linear array sensor. For example, the output end of the one-dimensional optical fiber array can be aligned with the corresponding linear array sensor and then packaged and bonded to achieve alignment and coupling.

[0053] When the K single-line array image sensors LS_i synchronously complete the acquisition of the segmented laser image signals, the image processing chip at the back end calculates the K defocus signals. The calculation can also be performed by a host computer or manually or by other computing entities. For the specific calculation algorithm, refer to the description of the above analysis method.

[0054] As a specific implementation of the present invention, analyzing the defocus amounts of valid sampling points among all defocus amount sampling points includes: Collecting sampling data of the corresponding areas of each one-dimensional optical fiber array in groups to correspond to a number of defocus amount sampling points respectively; Filter valid data from all sampled data to obtain a number of valid sampling points, so as to calculate the defocus amount corresponding to each valid sampling point; The standard deviation of the defocus values ​​at the effective sampling points is analyzed, and several effective defocus values ​​are screened to output the final defocus value.

[0055] By aligning and coupling several independently arranged one-dimensional optical fiber arrays with corresponding single-line array image sensors, sampling can be performed by adaptively adjusting according to the length of the laser image. This not only solves the problem of insufficient sampling area of ​​traditional multi-line array image sensors, but also effectively combines the characteristic advantages of multi-line array sensors to improve the defocus sampling rate. Example 3

[0056] Another aspect of the present invention further provides a map collection device, comprising: Such as the above-mentioned map collection device; The lens is arranged in front of the one-dimensional optical fiber array along the light propagation path. Example 4

[0057] like Figure 8 As shown, the present invention also provides a high-speed image signal analysis system, comprising: A posture adjustment module is used to adjust the posture of the sensor until several sensitive light lines of the sensor are perpendicular to the long axis direction of the laser image, so that the image data collected by the sensitive light lines can be directly used as sampling data; A grouping acquisition module is used to collect sampling data of different areas of the laser image along the long axis direction in groups, so as to correspond to a number of defocus amount sampling points respectively; An effective analysis module is used to screen effective data from all sampled data, thereby obtaining a number of effective sampling points, and calculating the defocus amount corresponding to each effective sampling point; The defocus analysis module is used to analyze the standard deviation of the defocus amount of the effective sampling points, and screen out a number of effective defocus amounts to output the final defocus amount. Example 5

[0058] Another aspect of the present invention provides a computer-readable storage medium, comprising a computer program, wherein the computer program implements the above-mentioned analysis method when executed by a processor.

[0059] In practical applications, the computer-readable storage medium may adopt any combination of one or more computer-readable media. The computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium may be, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, device or device.

[0060] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0061] The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0062] Computer program code for performing the operation of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0063] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0064] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements all fall within the scope of the present invention to be protected.

Claims

1. A high-speed image signal analysis method, which analyzes the collected data of the laser image in the auto-focus system to output the defocus value, characterized in that: Analytical methods include: Adjust the sensor posture until several sensing lines of the sensor are perpendicular to the long axis direction of the laser image, so that the image data collected by the sensing lines can be directly used as sampling data; Collecting sampling data of different regions of the laser image along the long axis direction in groups to correspond to a number of defocus amount sampling points respectively; Filter valid data from all sampled data to obtain a number of valid sampling points, so as to calculate the defocus amount corresponding to each valid sampling point; Analyze the standard deviation of the defocus values ​​at the effective sampling points, and screen out a number of effective defocus values ​​to output the final defocus value; Among them, the collection range of several sensitive light rays of the sensor covers the entire laser image.

2. The high-speed image signal analysis method according to claim 1, characterized in that: After collecting the sampling data of different areas of the laser image along the long axis direction in groups, when reading the sampling data corresponding to the current area, it is simultaneously analyzed whether the sampling data corresponding to the previous area is valid data.

3. The high-speed image signal analysis method according to claim 1, characterized in that: Screening valid data from all sampled data includes: calculating the maximum grayscale value in each group of sampled data to analyze whether the maximum grayscale value is greater than a grayscale threshold, and if so, determining it as valid data, otherwise, determining it as invalid data.

4. The high-speed image signal analysis method according to claim 3, characterized in that: The defocus amount corresponding to each valid sampling point is obtained by calculating the offset between the grayscale centroid of each valid sampling point and the focus reference zero point.

5. The high-speed image signal analysis method according to claim 4, characterized in that: The effective defocus amount is the defocus amount of the effective sampling points within the corresponding range of the standard deviation.

6. The high-speed image signal analysis method according to claim 5, characterized in that: The final defocus value is the average of all valid defocus values.

7. A high-speed image signal analysis method according to any one of claims 1 to 6, characterized in that: The sensor is a multi-line array image sensor, and the height of the multi-line array image sensor corresponds to the long axis span of the laser image.

8. A high-speed image signal analysis method according to any one of claims 1 to 6, characterized in that: The sensor is an area array image sensor and includes a plurality of regions of interest corresponding to each group of sampling data, each region of interest corresponding to a single line of sensitive light or a plurality of continuous lines of sensitive light.

9. A high-speed image signal analysis system, characterized in that: include: A posture adjustment module is used to adjust the posture of the sensor until several sensitive light lines of the sensor are perpendicular to the long axis direction of the laser image, so that the image data collected by the sensitive light lines can be directly used as sampling data; A grouping acquisition module is used to collect sampling data of different areas of the laser image along the long axis direction in groups, so as to correspond to a number of defocus amount sampling points respectively; An effective analysis module is used to screen effective data from all sampled data, thereby obtaining a number of effective sampling points, and calculating the defocus amount corresponding to each effective sampling point; The defocus analysis module is used to analyze the standard deviation of the defocus amount of the effective sampling points, and screen out a number of effective defocus amounts to output the final defocus amount.

10. A computer-readable storage medium comprising a computer program, characterized in that: When the computer program is executed by a processor, the analysis method according to any one of claims 1 to 8 is implemented.

Images