Penetrating scattering medium quantum imaging system and method based on light modulation and genetic algorithm

Abstract

The invention relates to the technical field of quantum imaging, in particular to a light modulation and genetic algorithm-based penetrating scattering medium quantum imaging system and method.The light modulation and genetic algorithm-based penetrating scattering medium quantum imaging system comprises a generation measurement unit, an optimization shaping unit and a verification evaluation unit. Measuring the initial modulation light by using a detector to obtain an initial image, obtaining a histogram of the initial image to obtain a maximum value and a minimum value of intensity distribution of the initial image, calculating a contrast ratio of the initial image according to the maximum value and the minimum value of the intensity distribution of the initial image, and defining a target function according to a contrast ratio value of the initial image; and the control center introduces a genetic algorithm according to a target function to optimize the initial modulation signal, finds out a modulation signal which maximizes the difference between a bright region and a dark region in the intensity distribution of the initial image, and performs peak signal-to-noise ratio calculation on a final image generated corresponding to the modulation signal to evaluate the imaging quality.

Description

Technical Field

[0001] The present invention relates to the field of quantum imaging technology, and in particular to a system and method for quantum imaging through scattering media based on light modulation and genetic algorithm. Background Art

[0002] Quantum imaging is an important technology that can effectively reduce the impact of noise on imaging quality, especially in low-light conditions. By utilizing the quantum interference effect, this can improve the detectability of the signal and reduce interference with the imaged object. It is very suitable for use in medical and biological research and can acquire data without damaging the sample.

[0003] In many practical applications, the imaging process is often disturbed by scattering media, resulting in a decrease in imaging quality. For example, in medical imaging, the scattering of light by human tissue can make the image blurred and difficult to accurately diagnose diseases. Traditional imaging methods often do not work well when facing scattering media and cannot effectively penetrate the scattering media and obtain clear images. Quantum imaging technology, as an emerging imaging technology, has unique advantages. By modulating the incident light and combining it with a genetic algorithm to find a modulation signal that maximizes the difference between bright and dark areas in the initial image intensity distribution, the imaging quality can be effectively improved. In order to solve this problem, we provide a quantum imaging system and method for penetrating scattering media based on light modulation and genetic algorithms. Summary of the invention

[0004] The purpose of the present invention is to provide a quantum imaging system and method for penetrating scattering media based on light modulation and genetic algorithm to solve the problems raised in the above background technology.

[0005] To achieve the above-mentioned purpose, a quantum imaging system for penetrating scattering media based on light modulation and genetic algorithm is provided, including a generation and measurement unit, an optimization and shaping unit, and a verification and evaluation unit;

[0006] The generating and measuring unit sends the incident light generated by the laser to the spatial light modulator, and the spatial light modulator modulates the incident light according to the initial modulation signal sent by the control center to form initial modulated light, and uses the detector to measure the initial modulated light to obtain the size and pixel intensity value of the initial image;

[0007] The optimization shaping unit obtains the intensity distribution of the initial image according to the size and pixel intensity value of the initial image, and feeds back the initial modulation signal corresponding to the initial image to the control center, and the control center introduces a genetic algorithm to iteratively update and optimize the initial modulation signal until a modulation signal that maximizes the difference between the bright area and the dark area in the intensity distribution of the initial image is obtained;

[0008] The verification and evaluation unit sends the modulation signal to the spatial light modulator through the control center. The spatial light modulator modulates the incident light according to the modulation signal to form modulated light. The modulated light passes through the scattering medium to form a final image. Finally, the imaging quality is evaluated by comparing the peak signal-to-noise ratio of the initial image and the final image.

[0009] As a further improvement of the technical solution, the spatial light modulator in the generating and measuring unit modulates the incident light according to the modulation signal sent by the control center to form the modulated light, specifically including:

[0010] The spatial light modulator is a liquid crystal spatial light modulator, which uses the birefringence characteristics of liquid crystal molecules to control the phase of incident light. When the arrangement of the liquid crystal molecules changes under the action of different electric fields, their refractive index will also change accordingly, causing the phase of the incident light to change when passing through the liquid crystal layer, forming modulated light.

[0011] As a further improvement of the technical solution, the steps of optimizing the operation of the shaping unit are as follows:

[0012] A1. Analyze the intensity distribution of the initial image according to the size and pixel value intensity of the initial image, obtain the maximum and minimum values ​​of the intensity distribution of the initial image, and calculate the contrast value of the initial image according to the maximum and minimum values ​​of the intensity distribution of the initial image, and then define the objective function according to the contrast value of the initial image, and feed it back to the control center, wherein the objective function is a modulation signal corresponding to the maximum contrast that maximizes the difference between the bright area and the dark area in the intensity distribution of the initial image;

[0013] A2. The control center introduces a genetic algorithm to optimize the initial modulation signal, that is, randomly extracts a group of initial modulation signals of spatial light modulators to form a genetic algorithm population, each modulation unit of the spatial light modulator represents an individual of the genetic algorithm population, and for the genetic algorithm population individual, uses its corresponding modulation unit to modulate the incident light to obtain optimized modulated light, then passes the optimized modulated light through the scattering medium to form an optimized image, and uses a detector to measure the intensity distribution of the optimized image, and finally, calculates the contrast value of the optimized image according to the intensity distribution of the optimized image, and uses the contrast value of the optimized image as the fitness value for the genetic algorithm population individual;

[0014] A3. Sort individuals according to their fitness values, and select individuals with high fitness values ​​as parents for crossover operations to generate new offspring individuals. Then perform mutation operations on the newly generated offspring individuals and add them to the population to replace individuals with low fitness values ​​to form a new population. Perform the next iteration until the maximum contrast value changes less than the contrast value threshold in several consecutive iterations, then stop the iterative optimization.

[0015] As a further improvement of the technical solution, the method of calculating the contrast value of the initial image according to the maximum and minimum values ​​of the intensity distribution of the initial image in S1 specifically includes:

[0016] Assume that the initial image has L different intensity levels, the intensity levels are 0 to L-1, the size of the initial image is M×N through the detector measurement, and the intensity value of each pixel is I i，j , where i=1, 2, ..., M, j=1, 2, ..., N, and the intensity value of each pixel is I i，j Substitute the formula to calculate the histogram H(k) of the initial image. The specific calculation formula is:

[0017]

[0018] Where k = 0, 1, ..., L-1, δ is the function of the histogram H(k) of the initial image. i，j -k=0,δ(I i，j -k)=1, otherwise 0, traverse the histogram H(k) of the initial image, and find the first non-zero k value, which is the minimum value I min , find the last non-zero k value, which is the maximum value I max , the maximum value I of the initial image intensity distribution max and minimum value I min Substitute the formula to calculate the contrast value C of the initial image, then the specific calculation formula is:

[0019]

[0020] The contrast value C of the initial image is used for the difference between the bright area and the dark area in the initial image. Maximizing this contrast value parameter is used as the objective function. By adjusting the modulation signal of the spatial light modulator, the modulated initial image has a higher contrast value.

[0021] As a further improvement of the technical solution, the control center in S2 introduces a genetic algorithm to optimize the modulation signal, specifically including:

[0022] The number of pixels of the spatial light modulator is recorded as A, where each pixel can control the phase of the incident light. The modulation signal emitted by the control center is now set to a vector of length A, where each element represents the phase value of a pixel, and the phase value range is set to [0, 2π]. According to the phase value range, B different vectors of length A are randomly selected as the initial population of the genetic algorithm. For the first individual in the initial population, its modulation signal is set to be the vector [a1, a2, ..., a A ], where a iIt is represented as the phase value of the i-th pixel. The spatial light modulator adjusts the phase of the incident light according to these phase values ​​to obtain the optimized modulated light. The contrast value of the image is calculated according to the intensity distribution of the optimized modulated light and used as the fitness value of the individual. The specific calculation formula is:

[0023]

[0024] Among them, f represents the fitness value of the individual.

[0025] As a further improvement of the technical solution, the method of sorting individuals according to fitness values ​​in S3, selecting individuals with high fitness values ​​as parents for crossover operation, and generating new offspring individuals specifically includes:

[0026] The fitness value f of each individual in the statistical genetic algorithm population i , sort each individual in the genetic algorithm population from large to small, find two individuals with high fitness values ​​as parents for crossover operation, and the crossover operation is performed as follows:

[0027] Assume that the modulation signals of the two parent individuals are vectors D = [d1, d2, ..., d A ] and the vector E = [e1, e2, ..., e A ], randomly select the crossover point as l, and the two offspring individuals generated after crossover are D'=[d1,d2,...,d l , d l+1 , ..., e A ] and E' = [e1, e2, ..., e l , e l+1 , ..., d A ].

[0028] As a further improvement of the technical solution, the method of performing mutation operation on the newly generated offspring individuals in S3 and adding them to the population to replace individuals with low fitness values ​​to form a new population specifically includes:

[0029] For the offspring individual vector D' and vector E', randomly select an element j in the vector for mutation, and randomly generate a new value j' from the phase value range [0, 2π] to replace the original element value, that is, update the element j in the individual vector D' to j', and perform the same operation on the vector E'. For the two mutated offspring individuals, use their corresponding modulation signals to modulate the incident light, let the modulated light pass through the scattering medium, and then use the detector to measure the intensity distribution, and calculate the contrast value of the image based on the intensity distribution as the fitness value of the two offspring individuals. Add the two mutated offspring individuals to the current population, traverse all individuals in the population, and sort the individuals according to the fitness value. If the fitness value of the newly added offspring individual is higher than that of the individual with the lowest fitness value, replace the individual with the lowest fitness value with the newly added offspring individual to form a new population.

[0030] As a further improvement of the technical solution, the method of stopping the iterative optimization in S3 until the maximum contrast value changes less than the contrast value threshold in several consecutive iterations specifically includes:

[0031] Set the contrast value threshold T. In each iteration, calculate the fitness values ​​of all individuals in the population and record the maximum contrast value C in this iteration. t After each iteration, the maximum contrast value C of the current iteration is t The maximum contrast value C with the previous iteration t-1 Compare them and calculate the absolute value dC of the difference between them, where dC = |C t -C t-1 |, if the absolute value of the difference is less than the preset contrast value threshold T for 5 consecutive times, it is considered that the stop condition is met and the iterative optimization is stopped.

[0032] As a further improvement of the technical solution, the verification and evaluation unit evaluates the imaging quality by comparing the peak signal-to-noise ratio of the initial image and the final image, specifically including:

[0033] The size of the initial image and the final image obtained by the detector is M×N, the initial image is recorded as I, the final image is recorded as I', and the peak signal-to-noise ratio of the initial image and the final image is obtained by calculating the mean square error of the initial image and the final image. The calculation formula of the mean square error of the initial image and the final image is:

[0034]

[0035] Wherein, i represents the row index of the initial image and the final image, j represents the column index of the initial image and the final image, i∈[0,M-1], j∈[0,N-1], and α represents the mean square error between the initial image and the final image;

[0036] Substitute the mean square error of the initial image and the final image into the formula to calculate the peak signal-to-noise ratio of the initial image and the final image. The specific calculation formula is:

[0037]

[0038] Among them, I max It is represented as the maximum value of the intensity distribution of the initial image. A reference image similar to the initial image is found from the image database. The peak signal-to-noise ratio β' of the initial image and the reference image is calculated by the above method. β' is used as the peak signal-to-noise ratio of the initial image. If β is greater than β', it means that the imaging quality of the final image is higher than that of the initial image.

[0039] A second object of the present invention is to provide a method for realizing a quantum imaging system for penetrating scattering media based on light modulation and genetic algorithm as described in any one of the above, comprising the following steps:

[0040] S1. Using a laser to generate incident light, the incident light automatically enters a spatial light modulator, and the spatial light modulator modulates the incident light according to a modulation signal to form an initial modulated light, and the initial modulated light automatically passes through a scattering medium to form an initial image;

[0041] S2, analyzing the intensity distribution of the initial image by calculating the histogram of the initial image, obtaining the maximum and minimum values ​​of the intensity distribution of the initial image, and calculating the contrast value of the initial image according to the maximum and minimum values ​​of the intensity distribution of the initial image, and finally defining the objective function according to the contrast value of the initial image;

[0042] S3, the control center introduces a genetic algorithm to optimize the initial modulation signal according to the objective function, obtains individuals of the genetic algorithm population by initializing the genetic algorithm population, and modulates the incident light using the modulation units corresponding to the individuals of the genetic algorithm population to obtain an optimized image, and uses the contrast value of the optimized image as the fitness value of the individual;

[0043] S4. Calculate the fitness values ​​of all individuals in the genetic algorithm population, and select individuals with high fitness values ​​as parents for crossover operation to generate new offspring individuals. Then perform mutation operation on the newly generated offspring individuals and add them to the population to replace individuals with low fitness values ​​until the maximum contrast value changes less than the contrast value threshold in several consecutive iterations. Then stop the iterative optimization, and form the modulation signal corresponding to the modulation units of the last batch of population individuals, which is determined to be the modulation signal that maximizes the difference between the bright area and the dark area in the initial image intensity distribution.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] In the quantum imaging system for penetrating scattering media based on light modulation and genetic algorithm, the genetic algorithm can search in the entire search space to avoid falling into the local optimal solution, making it more likely to find the global optimal modulation signal and improve the imaging quality. By optimizing the modulation signal through the genetic algorithm, the difference between the bright and dark areas in the initial image intensity distribution can be maximized, thereby improving the image contrast and making the image clearer. The optimized modulation signal enables the quantum imaging system to obtain better imaging effects when penetrating scattering media, reduce noise and distortion, and improve image quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 is an overall block diagram of the present invention;

[0047] Figure 2 It is the overall flow chart of the present invention.

[0048] The meaning of each number in the figure is:

[0049] 1. Generate measurement unit; 2. Optimize shaping unit; 3. Verify evaluation unit. DETAILED DESCRIPTION

[0050] 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.

[0051] Example 1

[0052] The present invention provides a quantum imaging system for penetrating scattering media based on light modulation and genetic algorithm. Figure 1 As shown, it includes a generation measurement unit 1, an optimization shaping unit 2, and a verification and evaluation unit 3;

[0053] The generating and measuring unit 1 sends the incident light generated by the laser to the spatial light modulator, and the spatial light modulator modulates the incident light according to the initial modulation signal sent by the control center to form an initial modulated light, and uses a detector to measure the initial modulated light to obtain the size and pixel intensity value of the initial image;

[0054] The spatial light modulator in the measurement unit 1 modulates the incident light according to the modulation signal sent by the control center to form the modulated light, specifically including:

[0055] In quantum imaging systems, the incident light needs to be modulated to obtain the required initial modulated light. By utilizing the birefringence characteristics of the liquid crystal spatial light modulator, the phase of the light can be effectively controlled to achieve light modulation.

[0056] The spatial light modulator is a liquid crystal spatial light modulator. The liquid crystal spatial light modulator uses the birefringence characteristics of liquid crystal molecules to control the phase of the incident light. When the arrangement of liquid crystal molecules changes under the action of different electric fields, their refractive index will also change accordingly, causing the phase of the incident light to change when passing through the liquid crystal layer, forming modulated light. Accurate phase modulation helps to better control the propagation and interference of light, thereby improving the quality and clarity of imaging, and providing a suitable basis for subsequent image measurement, analysis and optimization.

[0057] For example, for a specific pixel position, the modulation signal sent by the control center determines the orientation and electric field strength of the liquid crystal molecules at the pixel. According to the optical properties of the liquid crystal, the phase delay of the incident light passing through the pixel can be expressed as:

[0058]

[0059] Where (x, y) represents the pixel position, λ is the wavelength of the incident light, d is the thickness of the liquid crystal layer, Δn(x, y) is the change in the birefringence of the liquid crystal at the pixel position, and the control center achieves precise modulation of the incident light phase by changing Δn(x, y) of each pixel.

[0060] In quantum imaging systems, the initial image is often affected by the scattering medium, resulting in unclear differences between bright and dark areas and reduced imaging quality. By measuring and analyzing the initial image, obtaining its intensity distribution, and feeding back the initial modulation signal, the control center can use genetic algorithms to optimize the modulation signal to find the modulation signal that maximizes the difference between bright and dark areas, thereby improving imaging quality.

[0061] The optimization shaping unit 2 obtains the intensity distribution of the initial image according to the size and pixel intensity value of the initial image, and feeds back the initial modulation signal corresponding to the initial image to the control center. The control center introduces a genetic algorithm to iteratively update and optimize the initial modulation signal until a modulation signal is obtained that maximizes the difference between the bright area and the dark area in the intensity distribution of the initial image.

[0062] The steps to optimize the operation of shaping unit 2 are as follows:

[0063] Directly optimizing the intensity distribution of the initial image can more accurately find the direction to improve the imaging quality;

[0064] A1. According to the size and pixel value intensity of the initial image, the intensity distribution of the initial image is analyzed to obtain the maximum and minimum values ​​of the intensity distribution of the initial image, and the contrast value of the initial image is calculated according to the maximum and minimum values ​​of the intensity distribution of the initial image. Then, the objective function is defined according to the contrast value of the initial image. The objective function is a modulation signal that maximizes the difference between the bright area and the dark area in the intensity distribution of the initial image, and is fed back to the control center to maximize the difference between the bright area and the dark area in the intensity distribution of the initial image, thereby significantly improving the contrast of the image and making the image clearer and more recognizable.

[0065] A2. The control center introduces a genetic algorithm to optimize the initial modulation signal, that is, randomly extracts a group of initial modulation signals of spatial light modulators to form a genetic algorithm population. Each modulation unit of the initial modulation signal represents an individual of the genetic algorithm population. For the individuals in the genetic algorithm population, the incident light is modulated using its corresponding modulation unit to obtain optimized modulated light. Then, the optimized modulated light is passed through the scattering medium to form an optimized image, and the intensity distribution of the optimized image is measured using a detector. Finally, the contrast value of the optimized image is calculated based on the intensity distribution of the optimized image, and the contrast value of the optimized image is used as the fitness value of the individual. The genetic algorithm can automatically perform iterative update optimization, reduce manual intervention, and improve efficiency. The optimized modulation signal can better penetrate the scattering medium, reduce noise and distortion, and improve the overall quality of the image.

[0066] A3. Sort individuals according to fitness values, and select individuals with high fitness values ​​as parents for crossover operation to generate new offspring individuals. Then perform mutation operation on the newly generated offspring individuals and add them to the population to replace individuals with low fitness values ​​to form a new population. Perform the next iteration until the maximum contrast value changes less than the contrast value threshold in several consecutive iterations. Then stop the iterative optimization. The genetic algorithm can find the optimal solution in a larger search space and avoid falling into the local optimum. It can automatically adjust the modulation signal according to different initial images and scattering media conditions to obtain the best imaging effect.

[0067] The method of calculating the contrast value of the initial image according to the maximum value and the minimum value of the intensity distribution of the initial image in A1 specifically includes:

[0068] The histogram can intuitively display the distribution of pixels of different intensity levels in the image. By analyzing the histogram, the grayscale distribution characteristics of the image can be quickly obtained, including information such as the maximum and minimum values, thus providing a basis for calculating the contrast value.

[0069] Assume that the initial image has L different intensity levels, ranging from 0 to L-1. The size of the initial image measured by the detector is M×N, and the intensity value of each pixel is I i，j, where i = 1, 2, ..., M, j = 1, 2, ..., N, and the intensity value of each pixel is I i，j Substitute the formula to calculate the histogram H(k) of the initial image. The specific calculation formula is:

[0070]

[0071] Where k = 0, 1, ..., L-1, δ is the function of the histogram H(k) of the initial image. i，j -k=0,δ(I i，j -k)=1, otherwise it is 0. By traversing and counting the histogram, the maximum and minimum values ​​can be quickly calculated, and then the contrast value can be calculated. The calculation efficiency is high. The histogram H(k) of the initial image is traversed, and the first non-zero k value is found, which is the minimum value I min , find the last non-zero k value, which is the maximum value I max , the maximum and minimum values ​​calculated based on the histogram can accurately reflect the intensity range of the bright and dark areas in the image, so that the calculated contrast value can truly reflect the contrast of the image;

[0072] The maximum value of the initial image intensity distribution I max and minimum value I min Substitute the formula to calculate the contrast value C of the initial image, then the specific calculation formula is:

[0073]

[0074] The contrast value C of the initial image is used to measure the difference between the bright area and the dark area in the initial image, and maximizing this contrast value parameter is used as the objective function.

[0075] Genetic algorithm is a random search algorithm that simulates the natural evolution process. It can find the optimal solution in a complex search space. In quantum imaging systems, the optimization of the initial modulation signal is a complex problem. It is necessary to find the modulation signal that maximizes the difference between the bright and dark areas in the initial image intensity distribution among a large number of possible solutions. Genetic algorithm has global search capabilities and parallelism, and can effectively handle such complex optimization problems.

[0076] In A2, the control center introduces genetic algorithms to optimize the modulation signal, including:

[0077] The number of pixels of the spatial light modulator is recorded as A, where each pixel can control the phase of the incident light. The modulation signal emitted by the control center is now set to a vector of length A, where each element represents the phase value of a pixel, and the phase value range is set to [0, 2π]. According to the phase value range, B different vectors of length A are randomly selected as the initial population of the genetic algorithm. For the first individual in the initial population, its modulation signal is set to be the vector [a1, a2, ..., a A ], where a i It is represented as the phase value of the i-th pixel. The spatial light modulator adjusts the phase of the incident light according to these phase values ​​to obtain the optimized modulated light. The contrast value of the image is calculated according to the intensity distribution of the optimized modulated light and used as the fitness value of the individual. The specific calculation formula is:

[0078]

[0079] Among them, f represents the fitness value of the individual. The higher the contrast, that is, the greater the difference between the bright area and the dark area, the higher the fitness value.

[0080] The core idea of ​​the genetic algorithm is to find the optimal solution by simulating the natural evolution process. In this process, selection, crossover and mutation operations are key steps, which work together to realize the search and optimization of the solution space. By sorting individuals, selecting individuals with high fitness for crossover operations, and performing mutation operations on newly generated offspring individuals, the modulation signal can be gradually optimized to evolve in the direction of maximizing the difference between bright and dark areas in the initial image intensity distribution. After multiple iterations of optimization, a better modulation signal can be found, thereby improving the contrast of the initial image and maximizing the difference between bright and dark areas.

[0081] In A3, individuals are sorted according to their fitness values, and individuals with high fitness values ​​are selected as parents for crossover operations to generate new offspring individuals. Specifically, the method includes:

[0082] The fitness value f of each individual in the statistical genetic algorithm population i , sort each individual in the genetic algorithm population from large to small, find two individuals with high fitness values ​​as parents for crossover operation, and the crossover operation is performed as follows:

[0083] Assume that the modulation signals of the two parent individuals are vectors D = [d1, d2, ..., d A ] and the vector E = [e1, e2, ..., e A ], randomly select the crossover point as l, and the two offspring individuals generated after the crossover are D, = [d1, d2, ..., d l , d l+1 , ..., eA ] and E' = [e1, e2, ..., e l , e l+1 , ..., d A ], by exchanging some genes of two parent individuals after the crossover point, gene recombination and inheritance are achieved, the diversity of the population is increased, and it helps the algorithm explore more possibilities in the search space and avoid falling into the local optimal solution too early.

[0084] In A3, the newly generated offspring individuals are mutated and added to the population to replace individuals with low fitness values ​​to form a new population. Specifically, the following methods are used:

[0085] For the offspring individual vector D' and vector E', randomly select an element j in the vector for mutation, and randomly generate a new value j' from the phase value range [0, 2π] to replace the original element value, that is, update the element j in the individual vector D' to j', and perform the same operation on the vector E'. For the two mutated offspring individuals, use their corresponding modulation signals to modulate the incident light, let the modulated light pass through the scattering medium, and then use the detector to measure the intensity distribution, and calculate the image contrast value based on the intensity distribution as the fitness value of the two offspring individuals. Add the two mutated offspring individuals to the current population, traverse all individuals in the population, and sort the individuals according to the fitness value. If the fitness value of the newly added offspring individual is higher than that of the individual with the lowest fitness value, replace the individual with the lowest fitness value with the newly added offspring individual to form a new population, which increases the diversity of the population. By continuously mutating and updating the population, the algorithm can gradually find offspring individuals with higher fitness, thereby improving the quality of the final modulation signal, making the image contrast higher and the imaging quality better.

[0086] The iterative optimization process of the genetic algorithm is a process of constantly searching for the optimal solution. However, in practical applications, it is impossible to iterate infinitely due to the limitations of computing resources and time. Therefore, it is necessary to set a stop condition to determine when to stop the iteration and obtain a satisfactory solution. By comparing the difference between the maximum contrast value of the current iteration and the maximum contrast value of the previous iteration, it can be determined whether the algorithm has converged to a relatively stable state. If the absolute value of the difference is less than the preset contrast value threshold for multiple consecutive times, it means that the algorithm is difficult to optimize further. Stopping the iteration at this time can save computing resources.

[0087] In A3, the iterative optimization method is stopped until the maximum contrast value changes less than the contrast value threshold in several consecutive iterations, which specifically includes:

[0088] Set the contrast value threshold T. In each iteration, calculate the fitness values ​​of all individuals in the population and record the maximum contrast value C in this iteration.t After each iteration, the maximum contrast value C of the current iteration is t The maximum contrast value C with the previous iteration t-1 Compare and calculate the absolute value of the difference between them, dC, dC = |C t -C t-1 |, if the absolute value of the difference is less than the preset contrast value threshold T for 5 consecutive times, the stopping condition is considered to be met and the iterative optimization is stopped. Within a reasonable time, the algorithm can find a relatively optimal modulation signal to maximize the difference between the bright and dark areas in the initial image intensity distribution and improve the imaging quality.

[0089] The verification and evaluation unit 3 sends the modulation signal to the spatial light modulator through the control center. The spatial light modulator modulates the incident light according to the modulation signal to form modulated light. The modulated light passes through the scattering medium to form a final image. Finally, the imaging quality is evaluated by comparing the peak signal-to-noise ratio of the initial image and the final image.

[0090] In order to evaluate the imaging quality of the quantum imaging system after optimization and shaping, it is necessary to compare the initial image and the final image after modulated signal processing. The peak signal-to-noise ratio is a commonly used image quality evaluation indicator that can quantitatively reflect the difference between images.

[0091] The verification evaluation unit 3 evaluates the imaging quality by comparing the peak signal-to-noise ratio of the initial image and the final image, specifically including:

[0092] The size of the initial image and the final image obtained by the detector is M×N. The initial image is recorded as I and the final image is recorded as I'. The peak signal-to-noise ratio of the initial image and the final image is obtained by calculating the mean square error of the initial image and the final image. The calculation formula of the mean square error of the initial image and the final image is:

[0093]

[0094] Where i represents the row index of the initial image and the final image, j represents the column index of the initial image and the final image, i∈[0,M-1], j∈[0,N-1], and α represents the mean square error between the initial image and the final image;

[0095] Substitute the mean square error of the initial image and the final image into the formula to calculate the peak signal-to-noise ratio of the initial image and the final image. The specific calculation formula is:

[0096]

[0097] Among them, I maxIt is represented as the maximum value of the intensity distribution of the initial image. A reference image similar to the initial image is found from the image database. The peak signal-to-noise ratio β' of the initial image and the reference image is calculated by the above method. β' is used as the peak signal-to-noise ratio of the initial image. If β is greater than β', it means that the imaging quality of the final image is higher than that of the initial image.

[0098] In the present invention, a spatial light modulator is modulated by a modulation signal to obtain an initial modulated light, and the initial modulated light is passed through a scattering medium to form an initial image. The maximum and minimum values ​​of the intensity distribution of the initial image are obtained from the histogram of the initial image, and the contrast of the initial image is calculated based on the maximum and minimum values ​​of the intensity distribution of the initial image. Then, an objective function is defined based on the contrast value of the initial image. A control center introduces a genetic algorithm to optimize the initial modulation signal based on the objective function, finds a modulation signal that maximizes the difference between bright areas and dark areas in the intensity distribution of the initial image, and calculates the peak signal-to-noise ratio of a final image generated corresponding to the modulation signal to evaluate the imaging quality.

[0099] Example 2

[0100] The second object of the present invention is to provide a method for realizing any one of the above-mentioned quantum imaging systems based on light modulation and genetic algorithm for penetrating scattering media, comprising the following steps:

[0101] S1. Using a laser to generate incident light, the incident light automatically enters a spatial light modulator, and the spatial light modulator modulates the incident light according to a modulation signal to form an initial modulated light, and the initial modulated light automatically passes through a scattering medium to form an initial image;

[0102] S2, analyzing the intensity distribution of the initial image by calculating the histogram of the initial image, obtaining the maximum and minimum values ​​of the intensity distribution of the initial image, and calculating the contrast value of the initial image according to the maximum and minimum values ​​of the intensity distribution of the initial image, and finally defining the objective function according to the contrast value of the initial image;

[0103] S3, the control center introduces a genetic algorithm to optimize the initial modulation signal according to the objective function, obtains individuals of the genetic algorithm population by initializing the genetic algorithm population, and modulates the incident light using the modulation units corresponding to the individuals of the genetic algorithm population to obtain an optimized image, and uses the contrast value of the optimized image as the fitness value of the individual;

[0104] S4. Calculate the fitness values ​​of all individuals in the genetic algorithm population, and select individuals with high fitness values ​​as parents for crossover operation to generate new offspring individuals. Then perform mutation operation on the newly generated offspring individuals and add them to the population to replace individuals with low fitness values ​​until the maximum contrast value changes less than the contrast value threshold in several consecutive iterations. Then stop the iterative optimization, and form the modulation signal corresponding to the modulation units of the last batch of population individuals, which is determined to be the modulation signal that maximizes the difference between the bright area and the dark area in the initial image intensity distribution.

[0105] 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 by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A quantum imaging system for penetrating scattering media based on light modulation and genetic algorithm, characterized in that: It includes a generation measurement unit (1), an optimization shaping unit (2) and a verification and evaluation unit (3); The generating and measuring unit (1) sends the incident light generated by the laser to the spatial light modulator, and the spatial light modulator modulates the incident light according to the initial modulation signal sent by the control center to form initial modulated light, and uses a detector to measure the initial modulated light to obtain the size and pixel intensity value of the initial image; The optimization shaping unit (2) obtains the intensity distribution of the initial image according to the size and pixel intensity value of the initial image, and feeds back the initial modulation signal corresponding to the initial image to the control center, and the control center introduces a genetic algorithm to iteratively update and optimize the initial modulation signal until a modulation signal is obtained that maximizes the difference between the bright area and the dark area in the intensity distribution of the initial image; The verification and evaluation unit (3) sends a modulation signal to a spatial light modulator through a control center. The spatial light modulator modulates the incident light according to the modulation signal to form modulated light. The modulated light passes through a scattering medium to form a final image. Finally, the imaging quality is evaluated by analyzing the peak signal-to-noise ratio of the initial image and the final image.

2. The quantum imaging system for penetrating scattering media based on light modulation and genetic algorithm according to claim 1, characterized in that: The spatial light modulator in the generating and measuring unit (1) modulates the incident light according to the initial modulation signal sent by the control center to form the initial modulated light, specifically comprising: The spatial light modulator is a liquid crystal spatial light modulator, which uses the birefringence characteristics of liquid crystal molecules to control the phase of incident light. When the arrangement of the liquid crystal molecules changes under the action of different electric fields, their refractive index will also change accordingly, causing the phase of the incident light to change when passing through the liquid crystal layer, forming initial modulated light.

3. The quantum imaging system for penetrating scattering media based on light modulation and genetic algorithm according to claim 1, characterized in that: The steps of operating the optimization shaping unit (2) are specifically as follows: A1. Analyze the intensity distribution of the initial image according to the size and pixel value intensity of the initial image, obtain the maximum and minimum values ​​of the intensity distribution of the initial image, and calculate the contrast value of the initial image according to the maximum and minimum values ​​of the intensity distribution of the initial image, and then define the objective function according to the contrast value of the initial image, and feed it back to the control center, wherein the objective function is a modulation signal corresponding to the maximum contrast that maximizes the difference between the bright area and the dark area in the intensity distribution of the initial image; A2. The control center introduces a genetic algorithm to optimize the initial modulation signal, that is, randomly extracts a group of initial modulation signals of spatial light modulators to form a genetic algorithm population, each modulation unit in the spatial light modulator represents a genetic algorithm population individual, for the genetic algorithm population individual, uses its corresponding modulation unit to modulate the incident light to obtain optimized modulated light, then passes the optimized modulated light through the scattering medium to form an optimized image, and uses a detector to measure the intensity distribution of the optimized image, finally, calculates the contrast value of the optimized image according to the intensity distribution of the optimized image, and uses the contrast value of the optimized image as the fitness value for the genetic algorithm population individual; A3. Sort the genetic algorithm population individuals according to their fitness values, and select the genetic algorithm population individuals with high fitness values ​​as the parents for crossover operation to generate new offspring individuals, and then perform mutation operation on the new offspring individuals, add the mutated sub-band individuals to the population to replace the genetic algorithm population individuals with low fitness values, form a new population, and perform the next iteration until the maximum contrast value changes less than the contrast value threshold in several consecutive iterations, then stop the iterative optimization.

4. The quantum imaging system for penetrating scattering media based on light modulation and genetic algorithm according to claim 3, characterized in that: The method of calculating the contrast value of the initial image according to the maximum value and the minimum value of the intensity distribution of the initial image in S1 specifically includes: Assume that the initial image has L different intensity levels, the intensity levels are 0 to L-1, the size of the initial image is M×N, and the intensity value of each pixel is I i，j , where i=1, 2, ..., M, j=1, 2, ..., N, the intensity value of each pixel is substituted into the formula to calculate the histogram H(k) of the initial image. The specific calculation formula is: Where k = 0, 1, ..., L-1, δ is the function of the initial image histogram H(k), when I i，j -k=0,δ(I i，j -k)=1, otherwise 0, traverse the initial image histogram H(k), find the first non-zero k value as the minimum value I min , find the last non-zero k value as the maximum value I max , and the maximum value I of the initial image intensity distribution max and minimum value I min Substitute the formula to calculate the contrast value C of the initial image, then the specific calculation formula is: The contrast value C of the initial image is used to represent the difference between the bright area and the dark area in the initial image. The modulation signal corresponding to the maximum contrast value that maximizes the difference between the bright area and the dark area in the intensity distribution of the initial image is defined as the objective function.

5. The quantum imaging system for penetrating scattering media based on light modulation and genetic algorithm according to claim 4, characterized in that: The method in S2 where the control center introduces a genetic algorithm to optimize the initial modulation signal specifically includes: The number of pixels of the spatial light modulator is recorded as A, where each pixel can control the phase of the incident light. The initial modulation signal sent by the control center is set to a vector of length A, where each element represents the phase value of a pixel, and the phase value range is set to [0, 2π]. According to the phase value range, B different vectors of length A are randomly selected as the initial population of the genetic algorithm. For the first individual in the initial population, its modulation signal is set to the vector [a1, a2, ..., a A ], where a i It is represented as the phase value of the i-th pixel. The spatial light modulator adjusts the phase of the incident light according to the phase value to obtain the optimized modulated light. Then, the contrast value of the optimized image is calculated according to the intensity distribution of the optimized modulated light, and the contrast value of the optimized image is used as the fitness value of the individual. The specific calculation formula is: Among them, f represents the fitness value of the individual.

6. The quantum imaging system for penetrating scattering media based on light modulation and genetic algorithm according to claim 5, characterized in that: In S3, the individuals in the genetic algorithm population are sorted according to their fitness values, and individuals with high fitness values ​​are selected as parents to perform crossover operations to generate new offspring individuals, specifically including: The fitness value f of each individual in the statistical genetic algorithm population i , sort each individual in the genetic algorithm population from large to small, find two individuals with high fitness values ​​as parents for crossover operation, and the crossover operation is performed as follows: Assume that the modulation signals of the two parent individuals are vectors D = [d1, d2, ..., d A ] and the vector E = [e1, e2, ..., e A ], randomly select the crossover point l, and the two offspring individuals generated after the crossover are D'=[d1,d2,...,d l , d l+1 , ..., e A ] and E' = [e1, e2, ..., e l , e l+1 , ..., d A ].

7. The quantum imaging system for penetrating scattering media based on light modulation and genetic algorithm according to claim 6, characterized in that: In S3, the newly generated offspring individuals are mutated, and the mutated sub-band individuals are added to the population to replace the genetic algorithm population individuals with low fitness values ​​to form a new population, specifically including: For the offspring individual vector D' and vector E', randomly select an element j in the vector for mutation, and randomly generate a new value j' from the phase value range [0, 2π] to replace the original element value, that is, update the element j in the individual vector D' to j', and perform the same operation on the vector E'. For the two mutated offspring individuals, use their corresponding modulation signals to modulate the incident light, let the modulated light pass through the scattering medium, and then use the detector to measure the intensity distribution, and calculate the contrast value of the image based on the intensity distribution as the fitness value of the two offspring individuals. Add the two mutated offspring individuals to the current population, traverse all individuals in the population, and sort the individuals according to the fitness value. If the fitness value of the newly added offspring individual is higher than that of the individual with the lowest fitness value, replace the individual with the lowest fitness value with the newly added offspring individual to form a new population.

8. The quantum imaging system for penetrating scattering media based on light modulation and genetic algorithm according to claim 6, characterized in that: The method of stopping the iterative optimization in S3 until the maximum contrast value changes less than the contrast value threshold in several consecutive iterations specifically includes: Set the contrast value threshold T. In each iteration, calculate the fitness values ​​of all individuals in the population and record the maximum contrast value C in this iteration. t After each iteration, the maximum contrast value C of the current iteration is t The maximum contrast value C with the previous iteration t-1 Compare them and calculate the absolute value dC of the difference between them, where dC = |C t -C t-1 |, if the absolute value of the difference is less than the preset contrast value threshold T for 5 consecutive times, it is considered that the stopping condition is met, and the iterative optimization is stopped, and the modulation signal corresponding to the last batch of population individuals is determined as the modulation signal that maximizes the difference between the bright area and the dark area in the initial image intensity distribution.

9. The quantum imaging system for penetrating scattering media based on light modulation and genetic algorithm according to claim 1, characterized in that: The verification and evaluation unit (3) evaluates the imaging quality by comparing the peak signal-to-noise ratio of the initial image and the final image, specifically comprising: The size of the initial image and the final image obtained by the detector is M×N, the initial image is recorded as I, the final image is recorded as I', and the peak signal-to-noise ratio of the initial image and the final image is obtained by calculating the mean square error of the initial image and the final image. The calculation formula of the mean square error of the initial image and the final image is: Wherein, i represents the row index of the initial image and the final image, j represents the column index of the initial image and the final image, i∈[0,M-1], j∈[0,N-1], and α represents the mean square error between the initial image and the final image; Substitute the mean square error of the initial image and the final image into the formula to calculate the peak signal-to-noise ratio of the initial image and the final image. The specific calculation formula is: Among them, I max It is represented as the maximum value of the intensity distribution of the initial image. A reference image similar to the initial image is found from the image database. The peak signal-to-noise ratio β' of the initial image and the reference image is calculated by the above method. β' is used as the peak signal-to-noise ratio of the initial image. If β is greater than β', it means that the imaging quality of the final image is higher than that of the initial image.

10. A method for implementing a quantum imaging system for penetrating scattering media based on light modulation and genetic algorithm as claimed in any one of claims 1 to 9, characterized in that: The following steps are involved: S1. Using a laser to generate incident light, the incident light automatically enters a spatial light modulator, and the spatial light modulator modulates the incident light according to a modulation signal to form an initial modulated light, and the initial modulated light automatically passes through a scattering medium to form an initial image; S2, analyzing the intensity distribution of the initial image by calculating the histogram of the initial image, obtaining the maximum and minimum values ​​of the intensity distribution of the initial image, and calculating the contrast value of the initial image according to the maximum and minimum values ​​of the intensity distribution of the initial image, and finally defining the objective function according to the contrast value of the initial image; S3, the control center introduces a genetic algorithm to optimize the initial modulation signal according to the objective function, obtains individuals of the genetic algorithm population by initializing the genetic algorithm population, and modulates the incident light using the modulation units corresponding to the individuals of the genetic algorithm population to obtain an optimized image, and uses the contrast value of the optimized image as the fitness value of the individual; S4. Calculate the fitness values ​​of all individuals in the genetic algorithm population, and select individuals with high fitness values ​​as parents for crossover operation to generate new offspring individuals. Then perform mutation operation on the newly generated offspring individuals and add them to the population to replace individuals with low fitness values ​​until the maximum contrast value changes less than the contrast value threshold in several consecutive iterations. Then stop the iterative optimization, and form the modulation signal corresponding to the modulation units of the last batch of population individuals, which is determined to be the modulation signal that maximizes the difference between the bright area and the dark area in the initial image intensity distribution.

Images