A method for automatic adjustment of light source parameters for an LDI apparatus
By automatically adjusting the light source parameters in LDI devices and optimizing the light source parameters using image quality and swarm intelligence algorithms, the problem of light source adjustment relying on human experience in existing technologies is solved, achieving efficient and accurate light source parameter settings that are adaptable to different production environments.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- JIANGSU YSPHOTECH INTERGRATED CIRCUIT EQUIP CO LTD
- Filing Date
- 2023-11-03
- Publication Date
- 2026-07-21
Smart Images
Figure CN117471870B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of LDI equipment technology, and in particular to an automatic adjustment method for light source parameters of LDI equipment. Background Technology
[0002] Photolithography equipment is a core component of the equipment required for integrated circuit production. It prints the designed integrated circuit pattern onto photosensitive materials through light exposure. This manufacturing process is widely used in semiconductor devices, wafer manufacturing, chip production, and printed circuit boards. LDI (Laser Direct Imaging) equipment is primarily used to complete the exposure process on photosensitive materials. LDI technology is a type of direct imaging technology, and its image quality is clearer than traditional exposure techniques. Therefore, LDI equipment has a significant advantage in the manufacturing of mid-to-high-end devices.
[0003] Accurate target positioning is a crucial step in the use of LDI equipment, directly impacting the quality of the exposure process. Target position calculation relies on a vision imaging system and digital image processing methods; therefore, the configuration of the vision imaging system and the selection of the image processing method are key factors in ensuring the target positioning accuracy meets requirements. Since equipment status, exposure environment, and products vary during production, the settings of the vision imaging system must be adjusted accordingly. Selecting appropriate light source brightness and color is a critical operation, as the appropriateness of the light source settings determines the imaging quality, thus affecting target position calculation. Currently, LDI equipment often relies on operators manually adjusting the light source in the vision imaging system based on experience. This method depends on manual experience, resulting in low accuracy, low efficiency, and poor versatility. Summary of the Invention
[0004] This application addresses the aforementioned problems and technical needs by proposing an automatic adjustment method for light source parameters in LDI devices. The technical solution of this application is as follows:
[0005] An automatic adjustment method for light source parameters in LDI devices, the method comprising:
[0006] The target image under the current light source parameters is acquired using the camera in the LDI device. The fitness value is determined based on the image quality of the target image. The better the image quality of the target image, the higher the fitness value is obtained.
[0007] The next iteration is performed based on the current fitness value until the iteration termination condition is met. Then, the light source parameter corresponding to the maximum fitness value is output as the target light source parameter.
[0008] The light source in the LDI device is adjusted to the target light source parameters, thus completing the adjustment of the light source in the LDI device.
[0009] The further technical solution is that the light source parameters are the three-channel RGB values of the light source, and the three-channel RGB values of the light source parameters are binary encoded to obtain the corresponding binary sequence form of color code.
[0010] The further technical solution is that the obtained color code includes 24 bits, the lowest 8 bits are the binary sequence corresponding to the B channel value of the light source parameter, the middle 8 bits are the binary sequence corresponding to the G channel value of the light source parameter, and the highest 8 bits are the binary sequence corresponding to the R channel value of the light source parameter.
[0011] The further technical solution is that the next iteration operation based on the current fitness value includes: determining the fitness value of the color code corresponding to each light source parameter at the current time;
[0012] The swarm intelligence algorithm is used to select and retain the M color codes with the largest current fitness values, where M is an integer.
[0013] Using a swarm intelligence algorithm, Cr*M color codes are randomly selected from the reserved M color codes according to the crossover probability Cr as parent codes. Any two parent codes are cross-recombined to form a new color code.
[0014] The swarm intelligence algorithm is used to mutate the color codes that have completed the selection and crossover operations and then proceed to the next iteration.
[0015] A further technical solution involves cross-combining the bits of any two parent class codes to form a new color code, including for any first parent class code and second parent class code:
[0016] The first parent code and the second parent code are randomly divided into high-order segments and low-order segments respectively. The number of bits in the high-order segment of the first parent code and the number of bits in the high-order segment of the second parent code are equal. The number of bits in the low-order segment of the first parent code and the number of bits in the low-order segment of the second parent code are equal.
[0017] A new color code is formed by concatenating the high-order segment of the first parent code and the low-order segment of the second parent code.
[0018] A further technical solution involves using a swarm intelligence algorithm to perform mutation operations on the color codes that have completed the selection and crossover operations, including:
[0019] Based on the mutation probability Mr, randomly select Mr*K color codes from the K color codes retained after selection and crossover, and randomly invert several bits in each selected color code to form a new color code.
[0020] A further technical solution involves determining the fitness value of the color coding based on the image quality of the target image, including:
[0021] The sharpness index of the target image is obtained by calculating the degree of pixel value deviation of local pixels in the target image, and the sharpness index of the target image is used as the fitness value corresponding to the current light source parameters.
[0022] A further technical solution involves calculating the sum of the sub-sharpness indices between adjacent pixels and between spaced pixels in the target image, and using the degree of pixel value deviation of local pixels in the target image as the sharpness index of the target image.
[0023] A further technical solution involves calculating the sum of the sub-sharpness index between adjacent pixels in the row direction and the sub-sharpness index between pixels spaced apart in the row direction in the target image. The formula for calculating the sharpness index of the target image is as follows:
[0024]
[0025] Where a1 and a2 are positive definite parameters, I(x,y) is the pixel value of the pixel at coordinate (x,y) in the target image, I(x,y-1) is the pixel value of the pixel at coordinate (x,y-1) in the target image, I(x+1,y) is the pixel value of the pixel at coordinate (x+1,y) in the target image, I(x+2,y) is the pixel value of the pixel at coordinate (x+2,y) in the target image, and n is the total number of pixels in the target image.
[0026] The further technical solution is that the positive definite parameter a1+a2=1.
[0027] A further technical solution is that the automatic adjustment method for light source parameters includes:
[0028] When the number of iterations reaches the iteration threshold, or when a fitness value reaches the fitness value threshold, the iteration termination condition is determined to be met.
[0029] The further technical solution is that the swarm intelligence algorithm used is any one of the genetic algorithm, particle swarm algorithm, and ant colony algorithm.
[0030] The beneficial technical effects of this application are:
[0031] This application discloses an automatic adjustment method for light source parameters in LDI equipment. The method evaluates the adaptability of light source parameters based on the image quality of the actual target image and uses a swarm intelligence algorithm to search for the optimal solution of the light source parameters. This allows for automatic iterative calculation and finding a set of light source parameters that match the current production products and processing environment, resulting in better imaging effects for the visual imaging system of the LDI equipment. This method has a high degree of automation and high adjustment efficiency, avoiding the problems of low accuracy and low efficiency of manual adjustment of light source parameters, and has good versatility in different application scenarios. Attached Figure Description
[0032] Figure 1 This is a schematic diagram of the hardware structure of an LDI device.
[0033] Figure 2 This is a flowchart of a method for automatically adjusting light source parameters according to an embodiment of this application.
[0034] Figure 3 This is a flowchart of a method for automatically adjusting light source parameters according to another embodiment of this application.
[0035] Figure 4 This is a diagram illustrating how a new color code is obtained by crossing two parent color codes in an example.
[0036] Figure 5 This is a schematic diagram illustrating how two color codes are mutated to obtain new color codes in an example. Detailed Implementation
[0037] The specific embodiments of this application will be further described below with reference to the accompanying drawings.
[0038] This application discloses an automatic adjustment method for light source parameters in LDI devices. This method is implemented based on the hardware structure of the LDI device. Please refer to [reference needed]. Figure 1The schematic diagram shows that in the LDI equipment, the work platform 1 is used to place the LDI equipment exposure production products, such as PCBs and wafers. The work platform 1 has high displacement accuracy. The visual imaging system of the LDI equipment includes a camera 2, a light source 3, and a light source controller 4. (1) The camera 2 is generally a high-precision CMOS camera with more than 1 million pixels. The camera 2 is also equipped with a visual imaging lens 5, which is used to magnify the imaging resolution of the production products on the work platform 1 and magnify the image of the production products. The camera 2 is used to collect the image of the production products on the work platform 1 and feed the image back to the host computer 6. (2) The light source 3 is used to illuminate the production products when the camera 2 collects the image of the production products. By reasonably setting the color and brightness of the light source, the collected image of the production products can better reflect the characteristics of the production products. In order to achieve a better lighting effect, the light source 3 adopts a ring light source structure. (3) The light source controller 4 is connected to and controls the light source 3. The light source controller 4 can be a current digital controller, which contains multiple control ports. It can control the opening / closing of the light source 3, and can also control the color and brightness of the light source 3. The host computer 6 connects to and controls the light source controller 4. The host computer 6 is mainly used to calculate and process the data information collected by the hardware device and send control commands to the hardware.
[0039] When the LDI equipment is in operation, it first completes the pre-processing, including the host computer 6 controlling the camera 2 to turn on and initialize the camera parameters. The host computer 6 also controls the light source 3 to turn on via the light source controller 4. Then, the production product is placed on the work platform 1, and the host computer 6 controls the work platform 1 to move, thereby moving the target on the production product into the field of view of the camera 2. After the pre-processing is completed, the camera 2 can be used to acquire target images and transmit the acquired images to the host computer 6 in real time for target positioning and subsequent exposure processing. The automatic light source parameter adjustment method of this application is based on this existing hardware architecture of the LDI equipment. The host computer 6 can perform the automatic light source parameter adjustment method, and this automatic light source parameter adjustment method is executed after the pre-processing is completed. That is, after the pre-processing is completed, the automatic light source parameter adjustment method of this application is executed to adjust the light source. Then, the camera 2 is used to acquire target images and transmit the acquired images to the host computer 6 in real time for target positioning and subsequent exposure processing.
[0040] Please refer to Figure 2 The flowchart shown illustrates the automatic adjustment method for light source parameters in LDI devices, which includes the following steps:
[0041] Step 110: Use the camera in the LDI device to acquire the target image under the current light source parameters, and determine the fitness value of the color coding based on the image quality of the target image. The better the image quality of the target image, the higher the fitness value obtained.
[0042] The light source parameters are used to characterize the color and brightness of light source 3. When the light source parameters are different, the color and brightness of the corresponding light source will also be different.
[0043] Step 120: Perform the next iteration operation based on the current fitness value until the iteration termination condition is met, then output the light source parameter corresponding to the maximum fitness value as the target light source parameter.
[0044] The iteration termination condition is determined when the number of iterations reaches the iteration threshold, or when a fitness value reaches the fitness value threshold. The fitness value threshold is a pre-set empirical value, which can be set according to different device models or the required accuracy. The iteration number threshold can be customized, for example, set to 100.
[0045] Step 130: Control the light source in the LDI device to adjust to the target light source parameters, thus completing the adjustment of the light source in the LDI device.
[0046] This automatic light source parameter adjustment method evaluates the adaptability of the light source parameters based on the image quality of the actual target image and iteratively searches for the optimal solution of the light source parameters. In this way, it can automatically calculate and find a set of light source parameters that match the current production products and processing environment, so that the visual imaging system of LDI equipment has better imaging effect. This method has a high degree of automation and high adjustment efficiency, avoiding the problems of low accuracy and low efficiency of manual adjustment of light source parameters, and has good versatility in different application scenarios.
[0047] The light source parameters used in this application can be characterized by the values of the light source in a color space. Common color spaces include RGB, HSV, HSL, and CMYK. To more conveniently and simply represent the color and brightness of the light source, and for ease of adjustment, in one embodiment, the light source parameters used are the three-channel RGB values of the light source. Correspondingly, the light source 3 in the LDI device uses a light source with independent adjustment function for the three RGB channels. The host computer 6 can independently adjust the values of the three RGB channels of the light source 3 through the light source controller 4. Therefore, after determining the three-channel RGB values of the light source, controlling the light source 3 to display according to these three-channel RGB values can achieve the required color and brightness. The color and brightness of the light source 3 can be changed by changing the values of one or more color channels.
[0048] In one embodiment, the iterative operation in step 120 is based on a swarm intelligence algorithm. The swarm intelligence algorithm used can be any one of genetic algorithm, particle swarm optimization, or ant colony optimization.
[0049] Based on the characteristics of the light source parameters being the three-channel RGB values of the light source and the iterative operation based on a swarm intelligence algorithm, in another embodiment, the implementation of each step of the automatic adjustment method for the light source parameters includes the following, please refer to... Figure 3 :
[0050] Step 210: Randomly generate N light source parameters and encode each light source parameter to obtain the corresponding color code, thus obtaining an initial population containing N color codes.
[0051] N is an integer parameter, and the value of N can be set by the user. Generally, the value range is 20 to 100.
[0052] The encoding method can be varied, as long as it ensures a one-to-one correspondence between the light source parameters and the color codes. In one embodiment, binary encoding is used, that is, the three-channel RGB values of the light source parameters are binary encoded to obtain the corresponding binary sequence color code. When binary encoding the three-channel RGB values, a convenient encoding method is to independently binary encode the value of each channel. The value range of each channel is 0 to 255, occupying 8 bits, which can be represented as 00000000 to 11111111. Therefore, the color code obtained by encoding each light source parameter includes 24 bits. The lowest 8 bits are the binary sequence corresponding to the B channel value of the light source parameter, the middle 8 bits are the binary sequence corresponding to the G channel value of the light source parameter, and the highest 8 bits are the binary sequence corresponding to the R channel value of the light source parameter. For example, in one instance, when the light source parameter is {255,0,0}, the corresponding color code is 111111110000000000000000.
[0053] Step 220: For each color code in the current population, control the light source in the LDI device to adjust to the light source parameters corresponding to the color code, and use the camera in the LDI device to acquire the target image under the current light source parameters. Calculate the pixel value deviation of local pixels in the target image to obtain the sharpness index of the target image, and use the sharpness index of the target image as the fitness value of the corresponding color code.
[0054] One method for calculating the pixel value deviation of local pixels in a target image to obtain the sharpness index of the target image is to calculate the sum of the sub-sharpness index between adjacent pixels and the sub-sharpness index between spaced pixels in the target image, and use the pixel value deviation of local pixels in the target image as the sharpness index of the target image.
[0055] When calculating the sum of the sub-sharpness index between adjacent pixels and the sub-sharpness index between pixels spaced apart in the row direction of the target image, the formula for calculating the sharpness index of the target image is:
[0056]
[0057] in, It is a sub-resolution metric between adjacent pixels in the row direction of a target image. is the sub-sharpness index between pixels spaced in the row direction in the target image. a1 and a2 are positive definite parameters used for weighted calculation of the two sub-sharpness indices. Through actual adjustment and verification, the optimal parameter setting is a1 + a2 = 1. I(x,y) is the pixel value of the pixel at coordinates (x,y) in the target image, I(x,y-1) is the pixel value of the pixel at coordinates (x,y-1) in the target image, I(x+1,y) is the pixel value of the pixel at coordinates (x+1,y) in the target image, I(x+2,y) is the pixel value of the pixel at coordinates (x+2,y) in the target image, and n is the total number of pixels in the target image.
[0058] Step 230: Utilize a swarm intelligence algorithm to select based on the fitness values of each color code in the current population. This includes selecting and retaining the M color codes with the highest fitness values in the current population, where M is an integer parameter and M ≤ N.
[0059] Step 240: Perform crossover on the color codes in the current population using a swarm intelligence algorithm. This includes randomly selecting Cr*M color codes from the retained M color codes as parent codes according to a preset crossover probability Cr, and performing bit-wise crossover and recombination on any two parent codes to form new color codes, which are then added to the current population. The crossover probability Cr typically ranges from 0.4 to 0.99.
[0060] When forming a new color code by cross-combining the bits of any two parent codes, for any first and second parent codes: First and second parent codes are randomly divided into high-order segments and low-order segments respectively. The number of bits in the high-order segment of the first parent code is equal to the number of bits in the high-order segment of the second parent code, and the number of bits in the low-order segment of the first parent code is equal to the number of bits in the low-order segment of the second parent code. The high-order and low-order segments of the first and second parent codes are then concatenated to form a new color code. Similarly, the high-order and low-order segments of the second and first parent codes are concatenated to form a new color code.
[0061] For example, please refer to Figure 4The diagram shows that the first parent code is 101100010001101011011110, and the second parent code is 011010100101010011111111. These two parent codes are randomly divided into a high-order segment of 10 bits and a low-order segment of 14 bits, which are then concatenated to form two new color codes: 101100010001010011111111 and 011010100101101011011110.
[0062] Step 250: Mutate the color codes in the current population using a swarm intelligence algorithm. This includes randomly selecting Mr*K color codes from the K color codes retained after selection and crossover in the current population according to a preset mutation probability Mr, and randomly inverting several bits in each selected color code to form a new color code, which is then added to the current population. K is an integer parameter. The mutation probability Mr typically ranges from 0.005 to 0.01.
[0063] When inverting bits in each color code, this includes inverting a random number of bits at random positions. The number of inverted bits for any two color codes can be equal or unequal, and the inverted bits can overlap or be different. For example, please refer to... Figure 5 Invert two bits in 101100010001101011011110 and invert three bits in 011010100101101011011110.
[0064] Step 260: Determine if the iteration termination condition has been reached. If the termination condition has not been reached, use the population obtained after selection, crossover, and mutation as the next generation population and return to step 220 for the next iteration. If the iteration termination condition has been reached, use the light source parameter corresponding to the color code with the highest fitness value in the population obtained after selection, crossover, and mutation as the target light source parameter.
[0065] Step 270: Control the light source in the LDI device to adjust to the target light source parameters, and complete the adjustment of the light source in the LDI device.
[0066] The above are merely preferred embodiments of this application, and this application is not limited to the above embodiments. It is understood that other improvements and variations that can be directly derived or conceived by those skilled in the art without departing from the spirit and concept of this application should be considered to be included within the protection scope of this application.
Claims
1. A method for automatically adjusting light source parameters in LDI equipment, characterized in that, The automatic adjustment method for light source parameters includes: A target image is acquired using a camera in an LDI device under the current light source parameters. The sum of the sub-sharpness indices between adjacent pixels in the row direction and between pixels spaced apart in the row direction is calculated. The pixel value deviation of local pixels in the target image is used as the sharpness index of the target image. This sharpness index is then used as the fitness value corresponding to the current light source parameters. The better the image quality of the target image, the higher the fitness value. The formula for calculating the sharpness index of the target image is as follows: in, and These are positive definite parameters. The coordinates in the target image The pixel value of the pixel at that location. The coordinates in the target image The pixel value of the pixel at that location. The coordinates in the target image The pixel value of the pixel at that location. The coordinates in the target image The pixel value of the pixel at that location. It is the total number of pixels contained in the target image; The next iteration is performed based on the current fitness value until the iteration termination condition is met. Then, the light source parameter corresponding to the maximum fitness value is output as the target light source parameter. The light source in the LDI device is adjusted to the target light source parameters, thus completing the adjustment of the light source in the LDI device.
2. The automatic adjustment method for light source parameters according to claim 1, characterized in that, The step of performing the next iteration operation based on the current fitness value includes: Determine the fitness value of the color encoding corresponding to the current multiple light source parameters; The swarm intelligence algorithm is used to select and retain the M color codes with the largest current fitness values, where M is an integer. Using a swarm intelligence algorithm, Cr*M color codes are randomly selected from the reserved M color codes according to the crossover probability Cr as parent codes. Any two parent codes are cross-recombined to form a new color code. The swarm intelligence algorithm is used to mutate the color codes that have completed the selection and crossover operations and then proceed to the next iteration.
3. The automatic adjustment method for light source parameters according to claim 2, characterized in that, The swarm intelligence algorithm is used to perform mutation operations on the color codes that have completed the selection and crossover operations, including: Based on the mutation probability Mr, randomly select Mr*K color codes from the K color codes retained after selection and crossover, and randomly invert several bits in each selected color code to form a new color code.
4. The automatic adjustment method for light source parameters according to claim 2 or 3, characterized in that, The light source parameters are the three-channel RGB values of the light source. The three-channel RGB values of each light source parameter are binary encoded to obtain the corresponding binary sequence color code.
5. The automatic adjustment method for light source parameters according to claim 2 or 3, characterized in that, The swarm intelligence algorithm used can be any one of the genetic algorithm, particle swarm algorithm, and ant colony algorithm.
6. The automatic adjustment method for light source parameters according to claim 1, characterized in that, The positive definite parameter .
7. The method for automatically adjusting light source parameters according to any one of claims 1-3 and 6, characterized in that, The automatic adjustment method for light source parameters includes: When the number of iterations reaches the iteration threshold, or when a fitness value reaches the fitness value threshold, the iteration termination condition is determined to be met.