A method for judging the clarity of laser focus

Abstract

The present invention belongs to the field of imaging, and specifically relates to a laser focus clarity determination method, which is used to solve the problem that the existing clarity evaluation method can only not cover all edge information of the central pixel, resulting in easy fall into the local optimal solution. The laser focus clarity determination method takes the focus circle center as the center, emits N uniformly distributed concentric radial rays around the focus, calculates the gradient of the pixel point passed by each ray, and uses the gradient sum of all rays as the determination basis of the laser focus clarity.

Description

Technical Field

[0001] The invention belongs to the field of imaging, and in particular relates to a method for distinguishing the clarity of a laser focal spot. Background Art

[0002] On the SILEX-II laser device, the high spatial resolution imaging system can be used for physical measurements during aiming and shooting. When adjusting the posture and focal spot of the ellipsoidal plasma mirror (EPM), the imaging mirror group is adjusted to follow and image the secondary focal spot of the main laser focused by the EPM. How to quickly and clearly image the focal spot becomes a key factor restricting the next step of adjusting the target position. In actual experiments, the focal spot needs to be focused and imaged multiple times. The use of manual focusing will be affected by the electromagnetic radiation on site and affect the overall experimental efficiency of the device.

[0003] Traditional focusing evaluation functions are divided into spatial domain method, frequency domain method, entropy function method, statistical function method and other methods. Spatial domain method includes grayscale variance and grayscale gradient; grayscale gradient includes SMD function, Brenner function, Laplacian function, Prewitt function, energy gradient function and SML function; frequency domain method includes Fourier transform method, discrete cosine transform method and wavelet transform method; statistical function method includes Menmay function, Range function and power spectrum function; other methods include quadtree block method, structural texture method, domain related function method and support vector machine method. Most of the above methods take the entire image as the calculation target and add a lot of non-edge redundant information.

[0004] In response to the above problems, some researchers have made improvements from the following two aspects:

[0005] First, the traditional method does not emphasize the edge information of the image evaluation target, and incorporates irrelevant information such as non-edge information, resulting in increased computational redundancy and weakened anti-interference ability. Mo Chunhong et al. proposed to first use the local variance of the image as the local threshold to distinguish edge pixels from non-edge pixels, then calculate a new standard deviation of the entire image as the global threshold to weaken the influence of noise and background, and finally use the gradient focusing algorithm to calculate the focus value of the image after threshold preprocessing to evaluate the clarity (Mo Chunhong, Liu Bo, Ding Lu, et al. A gradient threshold automatic focusing algorithm [J]. Infrared and Laser Engineering, 2014 (1)); Ni Jun et al. proposed to first transform the target area to obtain the gradient of the typical target area, detect the edge line and its direction of the target, numerically fit the gradient, and calculate the edge sharpness distribution (Ni Jun, Yuan Jiahu, Wu Qinzhang. Optical image clarity judgment based on edge features [J]. Chinese Laser, 2009, 36 (001): 172-176).

[0006] Secondly, although the spatial grayscale gradient can better obtain edge information, most of the existing edge detection operators only extract the gradients in the horizontal and vertical directions of the image. When actually focusing, the actual gradient direction of the image is unknown. Zhang Laixian et al. made improvements on this basis and proposed to calculate the eight-field grayscale gradient of the central pixel and use the sum of the maximum grayscale gradient of each pixel in the automatic focusing area and the product of the maximum grayscale and the minimum gradient as a new evaluation function (Zhang Laixian, Sun Huayan, Guo Huichao, et al. Automatic focusing algorithm based on image grayscale gradient maximum value accumulation [J]. Journal of Photonics, 2013, 42(005): 605-610). Although the above method increases the edge information in eight directions by calculating the eight-field gradient of the central pixel, the edge information of the laser focus is far more than the grayscale gradient in eight directions. Summary of the invention

[0007] The purpose of the present invention is to solve the shortcoming that the existing clarity evaluation method can only not cover all the edge information of the central pixel and is easy to fall into the local optimal solution, and to provide a laser focus clarity judgment method.

[0008] In order to solve the deficiencies of the above-mentioned prior art, the present invention provides the following technical solutions:

[0009] A method for determining the clarity of a laser focal spot, which is special in that it comprises the following steps:

[0010] Step (1): For the focal spot imaged on the target surface of the sCMOS camera, the focal spot center (P, Q) is calculated by the least square method, and then the original image coordinate system (X, Y) is translated and transformed into the focal spot coordinate system (X′, Y′) with the focal spot center (P, Q) as the origin;

[0011] Step (2): With the focal spot center (P, Q) as the center, emit S uniformly distributed concentric radial rays in all directions;

[0012] Step (3): Count the pixels that each concentric radial ray passes through, calculate the grayscale gradient on the concentric radial ray, and sum the gradients on all concentric radial rays to obtain the clarity evaluation function Z;

[0013] In step (3), the grayscale gradient on the concentric radial ray is calculated, and the specific process of summing the gradients on all concentric radial rays to obtain the clarity evaluation function Z is as follows:

[0014] Assume that the gray value of the pixel (i, j) hit by a concentric radial ray k is expressed as Calculate the grayscale gradient of the pixel hit by the concentric radial ray and accumulate the results of all concentric radial rays to obtain the final clarity evaluation function Z:

[0015] Furthermore, in step (1), the conversion relationship between the original image coordinate system (X, Y) and the focal spot coordinate system (X′, Y′) is:

[0016] Furthermore, in step (2), the S uniformly distributed concentric radial rays can be expressed in the original image coordinate system as:

[0017] Furthermore, in step (3), the criterion for determining the pixel point through which each concentric radial ray passes is: if the distance from the pixel point to a concentric radial ray is less than It means that the concentric radial ray passes through the pixel point; assuming that the image size is M×N, the pixel point coordinates are set to (i, j), and the distance d from the concentric radial ray to the pixel point in the original image coordinate system can be expressed as

[0018]

[0019] Compared with the prior art, the present invention has the following beneficial effects:

[0020] The invention discloses a method for distinguishing the clarity of a laser focal spot. The method takes the center of a focal spot as the center, emits N uniformly distributed concentric radial rays around the focal spot, calculates the gradient of the pixel points passed by each concentric radial ray, and takes the gradient sum of all the concentric radial rays as a basis for distinguishing the clarity of the laser focal spot. The method overcomes the drawback of the existing spatial grayscale gradient method that only considers the edge information of the central pixel in eight directions, and considers that the radial direction of the edge of the focal spot in reality is the maximum gradient of the grayscale. Comparative experiments prove that the clarity evaluation function of the method has a higher clarity ratio, and the single peak and smoothness of the curve are better, and has stronger robustness and adaptability when used for automatic focusing of the laser focal spot. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 A schematic diagram of the structure of a high spatial resolution imaging system in one embodiment of the present invention;

[0022] Figure 2 for Figure 1 A schematic diagram of a process for determining the clarity of a laser focal spot in an embodiment;

[0023] Figure 3 It is the normalized result of the comparative experiment between the traditional focusing evaluation function and the present invention. DETAILED DESCRIPTION

[0024] The present invention will be further described below in conjunction with the accompanying drawings and exemplary embodiments.

[0025] Reference Figure 1,A laser focus clarity determination method is based on a high spatial resolution imaging system, such as Figure 1 As shown, the high spatial resolution imaging system includes a laser, an ellipsoidal plasma mirror, an imaging lens group, an sCMOS camera, a motion control component and a host computer; the laser emitted by the laser is focused by the ellipsoidal plasma mirror to form a secondary focal spot, and then imaged on the target surface of the sCMOS camera through the imaging lens group; the motion control component is used to control the imaging lens group to make precise movements along the axial direction, and it includes a driver, a stepper motor and a grating ruler with an accuracy of 0.2μm; the driver input end is connected to the host computer, and the output end is connected to the stepper motor; the stepper motor is connected to the imaging lens group; the grating ruler is connected to the input end of the host computer.

[0026] The host computer is used to receive the actual position value fed back by the grating ruler and compare it with the preset position value stored in the host computer. The host computer issues instructions to the driver based on the deviation between the two and adjusts the imaging lens group with an adjustment accuracy within 0.5μm.

[0027] A method for determining the clarity of a laser focus spot comprises the following steps:

[0028] Step (1): Refer to Figure 2 For the focal spot imaged on the target surface of the sCMOS camera, the focal spot center (P, Q) is calculated by the least squares method, and then the original image coordinate system (X, Y) is translated into the focal spot coordinate system (X′, Y′) with the focal spot center (P, Q) as the origin:

[0029] The conversion relationship between the original image coordinate system (X, Y) and the focal spot coordinate system (X′, Y′) is as follows:

[0030]

[0031] Step (2): With the focal spot center (P, Q) as the center, emit S uniformly distributed concentric radial rays in all directions:

[0032] S evenly distributed concentric radial rays cut the circle where the focal spot is located into S equal parts. The concentric radial rays can be expressed in the original image coordinate system as:

[0033]

[0034] Step (3): Count the pixels that each concentric radial ray passes through, and calculate the grayscale gradient on the concentric radial ray. Figure (c) shows a local enlarged view of the edge of a concentric radial ray. The gradients on all concentric radial rays are summed to obtain the clarity evaluation function Z:

[0035] If the distance from a pixel to a concentric radial ray is less than It means that the concentric radial ray passes through the pixel point; assuming that the image size is M×N, the pixel point coordinates are set to (i, j), and the distance d from the concentric radial ray to the pixel point can be expressed in the original image coordinate system as:

[0036]

[0037] Assume that the gray value of the pixel (i, j) hit by a concentric radial ray k is expressed as Calculate the grayscale gradient of the pixel hit by the concentric radial ray and accumulate the results of all concentric radial rays to obtain the final clarity evaluation function Z:

[0038]

[0039] The present invention is applied to a high spatial resolution imaging system on a SILEX-II laser device. In the process of adjusting the imaging system for follow-up and imaging the secondary focal spot focused by the main laser through an ellipsoidal plasma mirror, the image of the secondary focal spot is automatically focused. The present invention performs curve fitting near the focus, finds the optimal solution interval, and performs multiple iterations until the step length is less than 0.5 μm, thereby avoiding the problem that the hill climbing method and the Fibbonacci method are prone to fall into the local optimal solution.

[0040] Reference Figure 3 , the SMD function, SML function, discrete cosine transform (DCT), Vollaths function, multi-directional maximum gradient (MMG), area method (Area) and the present invention are selected from the traditional methods for comparative experiments. From the normalized results, it can be seen that the present invention has better unimodality and sensitivity near the focus, and is suitable for the needs of small focusing amounts.

[0041] The above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit them. For ordinary professional and technical personnel in the field, the specific technical solutions recorded in the aforementioned embodiments can be modified, or some of the technical features therein can be replaced by equivalents, and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions protected by the present invention.

Claims

1. A method for determining the clarity of a laser focus, characterized in that: The steps include: Step (1): For the focal spot imaged on the target surface of the sCMOS camera, the focal spot center (P, Q) is calculated by the least square method, and then the original image coordinate system (X, Y) is translated into the focal spot coordinate system (X′, Y′) with the focal spot center (P, Q) as the origin; The conversion relationship between the original image coordinate system (X, Y) and the focal spot coordinate system (X′, Y′) is: Step (2): With the focal spot center (P, Q) as the center, S uniformly distributed concentric radial rays are emitted to the surroundings; the S uniformly distributed concentric radial rays are expressed in the original image coordinate system as: Step (3): Count the pixels that each concentric radial ray passes through, calculate the grayscale gradient on the concentric radial ray, and sum the gradients on all concentric radial rays to obtain the clarity evaluation function Z; The judgment criteria for the pixel point that each concentric radial ray passes through is: if the distance from the pixel point to a concentric radial ray is less than It means that the concentric radial ray passes through the pixel point; assuming that the image size is M×N, the pixel point coordinates are set to (i, j), and the distance d from the concentric radial ray to the pixel point is expressed in the focal spot coordinate system as The specific process of calculating the grayscale gradient on the concentric radial ray and summing the gradients on all concentric radial rays to obtain the clarity evaluation function Z is as follows: Assume that the gray value of the pixel (i, j) hit by a concentric radial ray k is expressed as Calculate the grayscale gradient of the pixel hit by the concentric radial ray and accumulate the results of all concentric radial rays to obtain the final clarity evaluation function Z:

Images