CN115330617A - Medical image self-adaptive enhancement method based on improved multivariate universe optimization algorithm - Google Patents
Medical image self-adaptive enhancement method based on improved multivariate universe optimization algorithm Download PDFInfo
- Publication number
- CN115330617A CN115330617A CN202210959451.2A CN202210959451A CN115330617A CN 115330617 A CN115330617 A CN 115330617A CN 202210959451 A CN202210959451 A CN 202210959451A CN 115330617 A CN115330617 A CN 115330617A
- Authority
- CN
- China
- Prior art keywords
- universe
- individual
- cosmic
- value
- image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
- 238000000034 method Methods 0.000 title claims abstract description 52
- 238000005457 optimization Methods 0.000 title claims abstract description 31
- JJWKPURADFRFRB-UHFFFAOYSA-N carbonyl sulfide Chemical compound O=C=S JJWKPURADFRFRB-UHFFFAOYSA-N 0.000 claims abstract description 105
- 238000010606 normalization Methods 0.000 claims abstract description 22
- 241000282461 Canis lupus Species 0.000 claims abstract description 19
- 230000008569 process Effects 0.000 claims description 34
- 230000007246 mechanism Effects 0.000 claims description 14
- 230000003044 adaptive effect Effects 0.000 claims description 9
- 239000013598 vector Substances 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 6
- 230000003247 decreasing effect Effects 0.000 claims description 6
- 238000012163 sequencing technique Methods 0.000 claims description 6
- 230000008859 change Effects 0.000 claims description 4
- 150000001875 compounds Chemical class 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000000739 chaotic effect Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 239000002775 capsule Substances 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/90—Dynamic range modification of images or parts thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Image Processing (AREA)
Abstract
The invention provides a medical image self-adaptive enhancement method based on an improved multivariate cosmic optimization algorithm, which adopts a non-complete Beta function to carry out enhancement processing on an image after normalization processing to obtain an enhanced image; performing inverse normalization processing to obtain an enhanced output image, and evaluating the image quality of the output image by taking the variance of the image as a fitness value; initializing a multi-element universe population, wherein the two-dimensional space position of each universe individual in the multi-element universe population corresponds to a parameter combination (alpha, beta) of the incomplete Beta function; adopting a multi-universe algorithm and a grayish wolf algorithm to update the position of the multi-universe population, and then continuously adopting a multi-universe optimization algorithm to output the optimal parameter combinations (alpha and Beta) of the incomplete Beta function; and substituting the optimal parameter combination (alpha, beta) into the incomplete Beta function to carry out image enhancement operation. The method has strong local exploration capability, and can quickly improve convergence speed and improve optimal solution precision.
Description
[ technical field ] A method for producing a semiconductor device
The invention relates to the technical field of image enhancement, in particular to a medical image self-adaptive enhancement method based on an improved multivariate cosmic optimization algorithm.
[ background ] A method for producing a semiconductor device
Image enhancement, which means purposefully emphasizing the overall or local characteristics of an image, turning an original unclear image into clear or emphasizing some interesting features, enlarging the difference between different object features in the image, suppressing the uninteresting features, improving the image quality and enriching the information content, enhancing the image interpretation and identification effects, and meeting the needs of some special analyses.
In the related technology, a multi-universe optimization algorithm is usually adopted for image enhancement, and the method has the advantages of few parameters, easiness in implementation and the like, is widely applied to solving the optimal problem, but has the problems of easiness in falling into local optimization, low convergence speed and the like. Therefore, there is a need to provide a medical image adaptive enhancement method based on an improved multivariate cosmic optimization algorithm to solve the above problems.
[ summary of the invention ]
The technical problem to be solved by the invention is to provide a medical image self-adaptive enhancement method based on an improved multivariate cosmic optimization algorithm, which has strong local exploration capability, can quickly improve convergence speed and improve optimal solution precision.
In order to achieve the purpose, the technical scheme of the invention is as follows:
a medical image self-adaptive enhancement method based on an improved multivariate cosmic optimization algorithm comprises the following steps:
s1: taking a medical image to be enhanced as an input image, carrying out normalization processing on the input image, and then adopting a non-complete Beta function to carry out enhancement processing on the image after the normalization processing to obtain an enhanced image;
s2: performing inverse normalization processing on the enhanced image to obtain an enhanced output image, and evaluating the image quality of the output image by taking the variance of the image as a fitness value;
s3: randomly initializing a multi-element universe population according to the upper and lower bounds of the variable of the incomplete Beta function, wherein the two-dimensional space position of each universe individual in the multi-element universe population corresponds to one parameter combination (alpha, beta) of the incomplete Beta function;
s4: generating a random number r 1E [0,1] for each cosmic individual, if r1 is less than 0.5, dividing the cosmic individual into a first cosmic population, and updating the position of the cosmic individual in the first cosmic population by adopting a multivariate cosmic optimization algorithm; if r1 is more than or equal to 0.5, dividing the universe individual into a second universe population, and updating the position of the universe individual in the second universe population by adopting a wolf algorithm; finishing the position updating of all the universe individuals in the multi-element universe population, and respectively calculating fitness values of the updated universe individuals;
s5: assigning each cosmic individual in the multivariate cosmic population a label K =0 for a plurality of iterations; during each iteration, if the fitness value of the universe individual is not updated, the mark K = K +1 is given again; if the fitness value of the universe individual is updated, the mark K =0 is newly given; setting a flag threshold K T If the mark of the universe individual reaches the mark threshold K T If no better fitness is obtained, the position of the universe individual is updated as follows:
in the formula, U (t + 1) represents the position of the space individual after update, U (t) represents the position of the space individual before update, and U d And L d For the upper and lower bounds of the range of positions of individuals in the universe, r8 is [0,1]Random vectors within a range;
s6: judging whether the updated position of the cosmic individual exceeds a gray value boundary or not, and if the position of the cosmic individual is smaller than a minimum boundary value, assigning the position of the cosmic individual to be the minimum boundary value; if the position of the universe individual is larger than the maximum boundary value, assigning the position of the universe individual as the maximum boundary value; if the gray value boundary is not exceeded, the actual position of the gray value boundary is reserved;
s7: if the fitness value of the updated cosmic individual is not smaller than the fitness value before updating, replacing the position of the updated cosmic individual with the position before updating; if the fitness value of the updated cosmic individual is smaller than the fitness value before updating, the position of the cosmic individual before updating is reserved; iterating for multiple times until the maximum iteration times is reached, outputting the position information of the cosmic individual with the maximum fitness value, and taking the position information as the optimal parameter combination (alpha and Beta) of the incomplete Beta function;
s8: and substituting the optimal parameter combination (alpha, beta) into the incomplete Beta function, adopting the incomplete Beta function to enhance the image after normalization processing, and carrying out reverse normalization on the enhanced image to obtain the enhanced optimal output image.
Preferably, the process of the "normalization processing" in step S1 is represented as:
wherein f' (x, y) represents a normalized grayscale value of the input image, f (x, y) represents an original grayscale value of the input image, and G max 、G min Respectively representing the maximum value and the minimum value of the original gray value of the input image;
the process of "enhancement processing" is represented as:
g′(x,y)＝F(f′(x,y))
where g' (x, y) represents the gray scale value of the enhanced image and F (-) represents an incomplete Beta function.
Preferably, the process of the "inverse normalization processing" in the step S2 is:
g″(x,y)＝(G′ max -G′ min )g′(x,y)+G′ min
wherein G ″ (x, y) represents a gradation value, G' max 、G′ min Respectively representing the maximum value and the minimum value, G ', of the gray value of the enhanced image' max ＝255、G′ min ＝0。
Preferably, the calculation process of the "fitness value" in step S2 is represented as:
in the formula, NI (X) represents a fitness value, M, N represents the length and width of the output image, respectively, N represents the total number of pixels of the output image, and N = M × N.
Preferably, in step S4, the process of updating the positions of the cosmic individuals in the first cosmic population is as follows:
s411: sequencing all universe individuals in the multi-element universe population according to the fitness：NI(U 1 )＞NI(U 2 )…＞NI(U i )＞…NI(U NU ) In the formula, NI (U) i ) Representing any universe individual U in the multi-element universe population U i Normalized fitness value of (a); selecting the cosmic individual with the largest fitness value as the current optimal cosmic individual, N U Representing the number of cosmic individuals in the multi-element cosmic population;
s412: performing a roulette mechanism to perform an adaptive search, the roulette mechanism being represented by:
wherein r2 is [0,1]A random number within the range of the random number,
j-dimensional position information indicating a k-th cosmic individual selected through the roulette mechanism,
j-dimensional position information indicating the ith cosmic individual selected by the roulette mechanism;
s413: under the condition of not considering the size of the expansion rate, the cosmic individuals can stimulate internal objects to move to the current optimal cosmic individuals in order to realize local change and improve the self expansion rate, so that the positions of the cosmic individuals are updated, and the moving process is represented as follows:
in the formula (I), the compound is shown in the specification,
position information of j-th dimension of i-th cosmic individual, D dimension degree, and X j J-th dimension position information, l, representing the current optimum universe bj 、u bj Respectively represent
R3, r4, r5 represent [0,1]Random numbers in the range, WEP represents the probability of existence of wormholes in the multi-element universe population, TDR represents the step length of movement of objects in the universe individual towards the current optimal universe individual;
and S414, entering an iterative loop, and updating the WEP and the TDR, wherein the updating process of the WEP and the TDR is as follows:
in the formula, WEP min The minimum probability of the existence of wormholes in the multi-element universe population is represented, the value is 0.2 max Representing the maximum probability of the existence of wormholes in the multi-element universe population, and the value is 1.0; t denotes the current number of iterations, T max The maximum number of iterations is indicated.
Preferably, in step S4, the process of updating the positions of the cosmic individuals in the second cosmic population is as follows:
s421: and sequencing all universe individuals in the multi-element universe population according to fitness: NI (U) 1 )＞NI(U 2 )…＞NI(U i )＞…NI(U NU ) In the formula, NI (U) i ) Representing any universe individual U in the multi-element universe population U i Normalized fitness value of, N U Representing the number of cosmic individuals in the multi-element cosmic population;
s422: obtaining the first three optimal fitness values NI (U) 1 )、NI(U 2 )、NI(U 3 ) And corresponding universe individual U 1 、U 2 、U 3 ；
S423: the location update process for any cosmic individual is represented as:
in the formula, U (t + 1) represents the updated position of any cosmic individual, and X 1 、X 2 、X 3 Respectively represent any universe individual in the universe individual U 1 、U 2 、U 3 Update location under guidance, t 2 Indicating the current number of iterations, X a (t)、X b (t)、X c (t) represents the universe unit U 1 、U 2 、U 3 The current position, U (t), represents the position of any cosmic individual before updating; a is an element of { A ∈ } 1 ,A 2 ,A 3 }、C∈{C 1 ,C 2 ,C 3 All represent coefficient vectors, and the calculation process is as follows:
A＝2α·r6-α
C＝2·r7
in the formula, T max Representing the maximum number of iterations, r6, r7 are [0,1]The random vector in the range, a, represents a non-linear decreasing factor.
Compared with the related art, the invention has the beneficial effects that: the gray wolf algorithm is combined with the multi-universe optimization algorithm, so that the capability of searching and finding the optimal solution of the algorithm is improved, and the stability of the algorithm is improved; the gray wolf algorithm is strong in stability and has strong local exploration capacity, can be fully coupled with a multi-universe optimization algorithm, and exerts respective advantages of the two algorithms, so that the convergence speed is rapidly increased, and the optimal solution precision is improved; meanwhile, a nonlinear degressive factor is adopted in the gray wolf algorithm, so that the global exploration capability of the gray wolf algorithm can be further improved.
[ detailed description ] embodiments
The following description of the present invention is provided to enable those skilled in the art to better understand the technical solutions in the embodiments of the present invention and to make the above objects, features and advantages of the present invention more comprehensible.
The invention provides a medical image self-adaptive enhancement method based on an improved multivariate cosmic optimization algorithm, which comprises the following steps of:
s1: taking a medical image to be enhanced as an input image, carrying out normalization processing on the input image, and then carrying out enhancement processing on the image after the normalization processing by adopting a non-complete Beta function to obtain an enhanced image.
The normalization process is as follows:
wherein f' (x, y) represents a normalized grayscale value of the input image, f (x, y) represents an original grayscale value of the input image, and G max 、G min Respectively representing the maximum value and the minimum value of the original gray value of the input image;
the process of "enhancement processing" is represented as:
g′(x,y)＝F(f′(x,y))
where g' (x, y) represents the gray scale value of the enhanced image and F (-) represents an incomplete Beta function.
The general expression of the incomplete Beta function is:
the Beta function is expressed as:
in the formula, α and β represent parameters of the incomplete Beta function.
Different nonlinear transformation curves can be obtained by adjusting the values of the alpha and the beta, and then the values of the optimal change parameters alpha and beta are dynamically determined by utilizing the global optimization searching capability of an intelligent optimization algorithm, so that the self-adaptive enhancement of the image is realized.
S2: and performing inverse normalization processing on the enhanced image to obtain an enhanced output image, and evaluating the image quality of the output image by taking the variance of the image as a fitness value.
The process of the "denormalization processing" is as follows:
g″(x,y)＝(G′ max -G′ min )g′(x,y)+G′ min
wherein G ″ (x, y) represents a gradation value, G' max 、G′ min Respectively representing the maximum value and the minimum value, G ', of the gray value of the enhanced image' max ＝255、G′ min ＝0。
The calculation process of the "fitness value" is expressed as:
in the formula, NI (X) represents a fitness value of the output image, M, N represents a length and a width of the output image, respectively, N represents the total number of pixels of the output image, and N = M × N.
S3: and randomly initializing a multi-element universe population according to the upper and lower bounds of the variable of the incomplete Beta function, wherein the two-dimensional space position of each universe individual in the multi-element universe population corresponds to one parameter combination (alpha, beta) of the incomplete Beta function.
During initialization, the position information of the universe individual is initialized by adopting a chaotic model, wherein the chaotic model is selected from one of Chebyshev, logistic and Sine. Obtaining a multi-element universe population after initialization
U i (i＝1,2,...,N U ) Represents the (i) th individual in the universe,
location information representing the ith universe individualIn the form of a capsule, the particles,
the position information of the jth dimension of the ith universe individual is represented, and D represents the dimension number of the problem and is used for representing the number of variables; n is a radical of U And representing the number of the universe individuals in the multi-element universe population, and representing the number of the candidate solutions.
S4: generating a random number r 1E [0,1] for each cosmic individual, if r1 is less than 0.5, dividing the cosmic individual into a first cosmic population, and updating the position of the cosmic individual in the first cosmic population by adopting a multivariate cosmic optimization algorithm; if r1 is more than or equal to 0.5, dividing the universe individual into a second universe population, and updating the position of the universe individual in the second universe population by adopting a wolf algorithm; and finishing the position updating of all the universe individuals in the multi-element universe population, and recalculating the fitness value of the updated universe individuals.
The multi-universe optimization algorithm simulates the movement behavior of the multi-universe population under the combined action of white holes, black holes and wormholes. White and black holes are used in the exploration phase, while wormholes are used in the mining phase. The position of a cosmic individual is a hypothetical concept, which is altered by the movement of objects within the individual. The multivariate universe optimization algorithm follows the following rules in the process of performing optimization: (1) If the expansion rate of a universe is higher, the probability of generating white holes is higher; if the expansion rate of a universe is relatively low, it is more likely to generate black holes; (2) generating a white hole universe to repel objects; the universe generating the black hole can absorb objects; (3) Regardless of the expansion rate, other universes may transmit objects to the current optimal universe through wormholes.
The multivariate universe optimization algorithm carries out cyclic iteration on the initial universe through the white hole/black hole orbits and the wormholes, the wormholes are always established between a certain universe and the current optimal universe, and the white hole/black hole orbits can be established between any two universes. Where the universe represents a feasible solution to the problem, objects in the universe represent components of the solution, and the inflation rate of the universe represents the fitness value of the solution.
The updating process of the first universe population position comprises the following steps:
s411: and sequencing all universe individuals in the multi-element universe population according to fitness: NI (U) 1 )＞NI(U 2 )…＞NI(U i )＞…NI(U NU ) In the formula, NI (U) i ) Represents any universe individual U in the multi-element universe population U i Normalized fitness value of (a); and selecting the cosmic individuals with the maximum fitness value as the current optimal cosmic individuals.
S412: and executing a roulette mechanism to realize self-adaptive search.
Objects in cosmic individuals in a multi-element cosmic population are transferred through a white/black hole trajectory following a roulette mechanism represented by:
wherein r2 is [0,1]A random number within the range of the random number,
j-dimensional position information indicating a k-th cosmic individual selected through the roulette mechanism,
and j-dimension position information representing the ith cosmic individual selected by the roulette mechanism.
The method is characterized in that the searching is carried out according to the size of the adaptability value, objects with low adaptability values can be conveyed through white holes or black holes more easily, objects with high adaptability values have higher possibility of possessing the white holes, and objects with low expansion rate have lower possibility of possessing the black holes. And a roulette mechanism is arranged to realize an adaptive search process.
S413, under the condition of not considering the size of the expansion rate, the cosmic individual can excite the internal object to move to the current optimal cosmic individual in order to realize local change and improve the expansion rate of the individual, so as to realize the updating of the position of the cosmic individual, wherein the moving process is represented as:
in the formula (I), the compound is shown in the specification,
position information of j-th dimension of i-th cosmic individual, D dimension degree, and X j J-th dimension position information, l, representing the current optimum universe bj 、u bj Respectively represent
R3, r4, r5 represent [0,1]Random numbers in the range, WEP represents the probability of the existence of wormholes in the multi-element cosmic population, and TDR represents the step length of the movement of objects in the cosmic individual towards the current optimal cosmic individual.
S413: and entering an iterative loop, and updating the WEP and the TDR.
The update process of WEP and TDR is as follows:
in the formula, WEP min The minimum probability of the existence of wormholes in the multi-element universe population is represented, the value is 0.2 max Representing the maximum probability of the existence of wormholes in the multi-element universe population, and the value is 1.0; t denotes the current number of iterations, T max The maximum number of iterations is indicated.
The grey wolf algorithm simulates a leader level and a hunting mechanism of the natural grey wolf, the social level of the grey wolf is simulated by using mathematical modeling, the optimal solution, the second optimal solution and the third optimal solution are used as a leader, the remaining candidate solutions are assumed to be followers, the followers follow the leader, and the leader guides the followers to hunt.
Specifically, the process of updating the positions of the cosmic individuals in the second cosmic population is as follows:
s421: sequencing all universe individuals in the multi-element universe population according to the fitness: NI (U) 1 )＞NI(U 2 )…＞NI(U i )＞…NI(U NU ) In the formula, NI (U) i ) Represents any universe individual U in the multi-element universe population U i Normalized fitness value of (a);
s422: obtaining the first three optimal fitness values NI (U) 1 )、NI(U 2 )、NI(U 3 ) And corresponding universe individual U 1 、U 2 、U 3 ；
S423: the location update process for any cosmic individual is represented as:
in the formula, U (t + 1) represents the updated position of any cosmic individual, and X 1 、X 2 、X 3 Respectively represent any universe individual in the universe individual U 1 、U 2 、U 3 Update location under guidance, t 2 Indicating the current number of iterations, X a (t)、X b (t)、X c (t) represents the universe unit U 1 、U 2 、U 3 The current position, U (t), represents the position of any cosmic individual before updating; a is in the form of { A ∈ } 1 ,A 2 ,A 3 }、C∈{C 1 ,C 2 ,C 3 All represent coefficient vectors, and the calculation process is as follows:
A＝2α·r6-α
C＝2·r7
in the formula, T max Representing the maximum number of iterations, r6, r7 are [0,1]The random vector in the range, a, represents a non-linear decreasing factor.
Compared with the traditional linear decrement, the nonlinear decrement factor can more effectively improve the global exploration capability of the wolf algorithm.
Compared with the traditional linear decreasing factor, the nonlinear decreasing factor can more effectively improve the global exploration capability of the wolf algorithm. Specifically, the method comprises the following steps: a plays a crucial role in the convergence speed and accuracy of the algorithm, the larger the value of a is, the stronger the global search capability is, the easier the local optimization is to get rid of, but the weaker the local development capability of the algorithm is, the convergence speed is reduced; conversely, the smaller the value of a, the greater the local growth ability and the faster the convergence rate, but the more likely it falls into local optima. Therefore, a affects the overall algorithm's capability. In the invention, a is decreased in a nonlinear manner from 2 to 0, so that nonlinear search under complex conditions can be comprehensively and accurately reflected, and the exploration capability of the algorithm is improved.
The random number r1 is set to divide individuals in the multi-element universe population, the wolf algorithm is introduced to be combined with the traditional multi-element universe algorithm, the exploration capability of the wolf algorithm on local parts is fully utilized, and the overall exploration and the capability of finding the most solutions of the algorithm can be greatly improved.
S5: assigning a label K =0 to each universe individual in the multi-element universe population for a plurality of iterations; during each iteration, if the fitness value of the universe individual is not updated, the mark K = K +1 is given again; if the fitness value of the universe individual is updated, the mark K =0 is newly given; setting a flag threshold K T If the mark of the universe individual reaches the mark threshold K T If no better fitness is obtained, the position of the universe individual is updated as follows:
in the formula, U (t + 1) represents the position of the space individual after update, U (t) represents the position of the space individual before update, and U d And L d Upper and lower bounds for the location range of universe individualsAnd r8 is [0,1]Random vectors within a range.
S6: judging whether the updated position of the cosmic individual exceeds a gray value boundary or not, and if the position of the cosmic individual is smaller than a minimum boundary value, assigning the position of the cosmic individual to be the minimum boundary value; if the position of the universe individual is larger than the maximum boundary value, the position of the universe individual is assigned as the maximum boundary value; if the gray value boundary is not exceeded, the actual position is retained.
S7: if the fitness value of the updated universe individual is not smaller than the fitness value before updating, replacing the position of the updated universe individual with the position before updating; if the fitness value of the updated cosmic individual is smaller than the fitness value before updating, the position of the cosmic individual before updating is reserved; and (4) iterating for multiple times until the maximum iteration times is reached, outputting the position information of the universe individual with the maximum fitness value, and taking the position information as the optimal parameter combination (alpha and Beta) of the incomplete Beta function.
S8: and substituting the optimal parameter combination (alpha, beta) into the incomplete Beta function, adopting the incomplete Beta function to enhance the image after normalization processing, and carrying out reverse normalization on the enhanced image to obtain the enhanced optimal output image.
Compared with the related art, the invention has the beneficial effects that: the gray wolf algorithm is combined with the multi-universe optimization algorithm, so that the capability of searching and finding the optimal solution of the algorithm is improved, and the stability of the algorithm is improved; the gray wolf algorithm is strong in stability and has strong local exploration capacity, can be fully coupled with a multi-universe optimization algorithm, and exerts respective advantages of the two algorithms, so that the convergence speed is rapidly increased, and the optimal solution precision is improved; meanwhile, a nonlinear degressive factor is adopted in the gray wolf algorithm, so that the global exploration capability of the gray wolf algorithm can be further improved.
The embodiments of the present invention have been described in detail, but the present invention is not limited to the described embodiments. Various changes, modifications, substitutions and alterations to these embodiments will occur to those skilled in the art without departing from the spirit and scope of the present invention.
Claims (6)
1. A medical image self-adaptive enhancement method based on an improved multivariate cosmic optimization algorithm is characterized by comprising the following steps:
s1: taking a medical image to be enhanced as an input image, carrying out normalization processing on the input image, and then carrying out enhancement processing on the image after the normalization processing by adopting a non-complete Beta function to obtain an enhanced image;
s2: performing inverse normalization processing on the enhanced image to obtain an enhanced output image, and evaluating the image quality of the output image by taking the variance of the image as a fitness value;
s3: randomly initializing a multi-element universe population according to the upper and lower bounds of the variable of the incomplete Beta function, wherein the two-dimensional space position of each universe individual in the multi-element universe population corresponds to one parameter combination (alpha, beta) of the incomplete Beta function;
s4: generating a random number r1 from each universe individual [0,1], if r1 is less than 0.5, dividing the universe individual into a first universe population, and updating the positions of the universe individuals in the first universe population by adopting a multivariate universe optimization algorithm; if r1 is more than or equal to 0.5, dividing the universe individual into a second universe population, and updating the position of the universe individual in the second universe population by adopting a wolf algorithm; finishing the position updating of all the universe individuals in the multi-element universe population, and respectively calculating fitness values of the updated universe individuals;
s5: assigning each cosmic individual in the multivariate cosmic population a label K =0 for a plurality of iterations; during each iteration, if the fitness value of the universe individual is not updated, the mark K = K +1 is given again; if the fitness value of the universe individual is updated, the mark K =0 is newly given; setting a marking threshold K T If the mark of the universe individual reaches the mark threshold K T If no better fitness is obtained, the position of the universe individual is updated as follows:
in the formula, U (t + 1) represents the position of the space individual after update, U (t) represents the position of the space individual before update, and U d And L d For the upper and lower bounds of the range of positions of individuals in the universe, r8 is [0,1]Random vectors within a range;
s6: judging whether the updated position of the cosmic individual exceeds a gray value boundary or not, and if the position of the cosmic individual is smaller than a minimum boundary value, assigning the position of the cosmic individual to be the minimum boundary value; if the position of the universe individual is larger than the maximum boundary value, the position of the universe individual is assigned as the maximum boundary value; if the gray value boundary is not exceeded, the actual position of the gray value boundary is reserved;
s7: if the fitness value of the updated cosmic individual is not smaller than the fitness value before updating, replacing the position of the updated cosmic individual with the position before updating; if the fitness value of the updated cosmic individual is smaller than the fitness value before updating, the position of the cosmic individual before updating is reserved; iterating for multiple times until the maximum iteration times is reached, outputting the position information of the universe individual with the maximum fitness value, and taking the position information as the optimal parameter combination (alpha and Beta) of the incomplete Beta function;
s8: and substituting the optimal parameter combination (alpha, beta) into the incomplete Beta function, adopting the incomplete Beta function to enhance the image after normalization processing, and carrying out reverse normalization on the enhanced image to obtain the enhanced optimal output image.
2. The medical image adaptive enhancement method based on the improved multivariate cosmic optimization algorithm according to claim 1, wherein the process of the normalization process in the step S1 is represented as follows:
wherein f' (x, y) represents a normalized grayscale value of the input image, f (x, y) represents an original grayscale value of the input image, and G max 、G min Respectively representing the maximum value and the minimum value of the original gray value of the input image;
the process of "enhancement processing" is represented as:
g′(x,y)＝F(f′(x,y))
where g' (x, y) represents the grayscale value of the enhanced image and F (·) represents a non-complete Beta function.
3. The medical image adaptive enhancement method based on the improved multivariate cosmic optimization algorithm according to claim 1, wherein the process of the "inverse normalization process" in the step S2 is as follows:
g″(x,y)＝(G′ max -G′ min )g′(x,y)+G′ min
wherein G ' (x, y) represents a gray scale value G ' of the output image ' max 、G′ min Respectively representing the maximum value and the minimum value, G ', of the gray value of the enhanced image' max ＝255、G′ min ＝0。
4. The medical image adaptive enhancement method based on the improved multivariate cosmic optimization algorithm according to claim 1, wherein the calculation process of the "fitness value" in the step S2 is represented as:
in the formula, NI (X) represents a fitness value, M, N represents the length and width of the output image, respectively, N represents the total number of pixels of the output image, and N = M × N.
5. The medical image adaptive enhancement method based on the improved multivariate cosmic optimization algorithm according to claim 1, wherein in the step S4, the updating process of the positions of the cosmic individuals in the first cosmic population is as follows:
s411: sequencing all universe individuals in the multi-element universe population according to the fitness: NI (U) 1 )＞NI(U 2 )…＞NI(U i )＞…NI(U NU ) In the formula, NI (U) i ) Representing any universe individual U in the multi-element universe population U i Normalized fitness value of (a); selecting the cosmic individual with the largest fitness value as the current optimal cosmic individual, N U Representing the number of cosmic individuals in the multi-element cosmic population;
s412: executing a roulette mechanism to perform an adaptive search, the roulette mechanism being represented by:
wherein r2 is [0,1]A random number within the range of the random number,
j-dimensional position information indicating a k-th cosmic individual selected through the roulette mechanism,
j-dimensional position information indicating the ith cosmic individual selected by the roulette mechanism;
s413: under the condition of not considering the size of the expansion rate, the cosmic individuals can stimulate internal objects to move to the current optimal cosmic individuals in order to realize local change and improve the self expansion rate, so that the positions of the cosmic individuals are updated, and the moving process is represented as follows:
in the formula (I), the compound is shown in the specification,
position information of j-th dimension of i-th cosmic individual, D dimension degree, and X j J-th dimension position information, l, representing the current optimum universe bj 、u bj Respectively represent
R3, r4, r5 represent [0,1]Random numbers in the range, WEP represents the probability of existence of wormholes in the multi-element universe population, TDR represents the step length of movement of objects in the universe individual towards the current optimal universe individual;
and S414, entering an iterative loop, and updating the WEP and the TDR, wherein the updating process of the WEP and the TDR is as follows:
in the formula, WEP min The minimum probability of the existence of the wormholes in the multi-element universe population is represented, the value is 0.2 max Representing the maximum probability of the existence of wormholes in the multi-element universe population, and the value is 1.0; t denotes the current number of iterations, T max The maximum number of iterations is indicated.
6. The medical image adaptive enhancement method based on the improved multivariate cosmic optimization algorithm according to claim 1, wherein in the step S4, the updating process of the positions of the cosmic individuals in the second cosmic population is as follows:
s421: sequencing all universe individuals in the multi-element universe population according to the fitness: NI (U) 1 )＞NI(U 2 )…＞NI(U i )＞…NI(U NU ) In the formula, NI (U) i ) Represents any universe individual U in the multi-element universe population U i Normalized fitness value of, N U Representing the number of cosmic individuals in the multi-element cosmic population;
s422: obtaining the first three optimal fitness values NI (U) 1 )、NI(U 2 )、NI(U 3 ) And corresponding universe individual U 1 、U 2 、U 3 ；
S423: the location update process for any cosmic individual is represented as:
in the formula, U (t + 1) represents the updated position of any cosmic individual, and X 1 、X 2 、X 3 Respectively represent any universe individual in the universe individual U 1 、U 2 、U 3 Update location under guidance, t 2 Indicating the current number of iterations, X a (t)、X b (t)、X c (t) represents the universe unit U 1 、U 2 、U 3 The current position, U (t), represents the position of any cosmic individual before updating; a is in the form of { A ∈ } 1 ,A 2 ,A 3 }、C∈{C 1 ,C 2 ,C 3 All represent coefficient vectors, and the calculation process is as follows:
A＝2α·r6-α
C＝2·r7
in the formula, T max Representing the maximum number of iterations, r6, r7 are [0,1]The random vector in the range, a, represents a non-linear decreasing factor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210959451.2A CN115330617A (en) | 2022-08-11 | 2022-08-11 | Medical image self-adaptive enhancement method based on improved multivariate universe optimization algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210959451.2A CN115330617A (en) | 2022-08-11 | 2022-08-11 | Medical image self-adaptive enhancement method based on improved multivariate universe optimization algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115330617A true CN115330617A (en) | 2022-11-11 |
Family
ID=83921935
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210959451.2A Withdrawn CN115330617A (en) | 2022-08-11 | 2022-08-11 | Medical image self-adaptive enhancement method based on improved multivariate universe optimization algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115330617A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117350929A (en) * | 2023-11-07 | 2024-01-05 | 国网浙江省电力有限公司双创中心 | Electric power image enhancement optimization method based on improved meta heuristic algorithm |
| CN117439190A (en) * | 2023-10-26 | 2024-01-23 | 华中科技大学 | Water, fire and wind system dispatching method, device, equipment and storage medium |
| CN117523060A (en) * | 2024-01-04 | 2024-02-06 | 虚拟现实(深圳)智能科技有限公司 | Image quality processing method, device, equipment and storage medium for metauniverse digital person |
-
2022
- 2022-08-11 CN CN202210959451.2A patent/CN115330617A/en not_active Withdrawn
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117439190A (en) * | 2023-10-26 | 2024-01-23 | 华中科技大学 | Water, fire and wind system dispatching method, device, equipment and storage medium |
| CN117439190B (en) * | 2023-10-26 | 2024-06-11 | 华中科技大学 | Water, fire and wind system dispatching method, device, equipment and storage medium |
| CN117350929A (en) * | 2023-11-07 | 2024-01-05 | 国网浙江省电力有限公司双创中心 | Electric power image enhancement optimization method based on improved meta heuristic algorithm |
| CN117523060A (en) * | 2024-01-04 | 2024-02-06 | 虚拟现实(深圳)智能科技有限公司 | Image quality processing method, device, equipment and storage medium for metauniverse digital person |
| CN117523060B (en) * | 2024-01-04 | 2024-05-17 | 虚拟现实(深圳)智能科技有限公司 | Image quality processing method, device, equipment and storage medium for metauniverse digital person |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115330617A (en) | Medical image self-adaptive enhancement method based on improved multivariate universe optimization algorithm | |
| US12112256B2 (en) | Loss-error-aware quantization of a low-bit neural network | |
| CN110288030B (en) | Image identification method, device and equipment based on lightweight network model | |
| CN108389211B (en) | Image segmentation method based on improved whale optimized fuzzy clustering | |
| CN109919183B (en) | Image identification method, device and equipment based on small samples and storage medium | |
| CN113963165B (en) | Small sample image classification method and system based on self-supervision learning | |
| Wismüller et al. | The deformable feature map—adaptive plasticity for function approximation | |
| CN109146061A (en) | The treating method and apparatus of neural network model | |
| Wiwatcharakoses et al. | A self-organizing incremental neural network for continual supervised learning | |
| CN108154235A (en) | A kind of image question and answer inference method, system and device | |
| CN112580728B (en) | Dynamic link prediction model robustness enhancement method based on reinforcement learning | |
| CN114332538A (en) | Image classification model training method, image classification method, device and storage medium | |
| CN112364980B (en) | Deep neural network training method based on reinforcement learning under weak supervision scene | |
| Zhai et al. | Fine-Grained Knowledge Selection and Restoration for Non-exemplar Class Incremental Learning | |
| CN113705724A (en) | Batch learning method of deep neural network based on self-adaptive L-BFGS algorithm | |
| US7933449B2 (en) | Pattern recognition method | |
| Wen et al. | A Batch Rival Penalized Expectation‐Maximization Algorithm for Gaussian Mixture Clustering with Automatic Model Selection | |
| Bansiwala et al. | Continual learning for food recognition using class incremental extreme and online clustering method: self-organizing incremental neural network | |
| CN114254618B (en) | Visual question-answering method, storage medium and device based on dynamic route attention mechanism | |
| CN113449817B (en) | Image classification implicit model acceleration training method based on phantom gradient | |
| CN113205159B (en) | Knowledge migration method, wireless network equipment individual identification method and system | |
| Borycki et al. | Hypernetwork approach to Bayesian maml | |
| US20220147790A1 (en) | Deep Polynomial Neural Networks | |
| Turajlic | Multi-swarm Particle Swarm Optimization with Chaotic Random Inertia Weight and Dynamic Learning Strategy for Multilevel Thresholding Image Segmentation | |
| CN111444959A (en) | Construction method of stack-type structure hierarchical classification model based on SVM |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20221111 |
|
| WW01 | Invention patent application withdrawn after publication |