CN110047048B - Phase recovery improved algorithm based on MSE (mean square error) optimization - Google Patents
Phase recovery improved algorithm based on MSE (mean square error) optimization Download PDFInfo
- Publication number
- CN110047048B CN110047048B CN201910309968.5A CN201910309968A CN110047048B CN 110047048 B CN110047048 B CN 110047048B CN 201910309968 A CN201910309968 A CN 201910309968A CN 110047048 B CN110047048 B CN 110047048B
- Authority
- CN
- China
- Prior art keywords
- fourier
- reconstruction result
- mean square
- phase recovery
- reconstruction
- 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.)
- Active
Links
- 238000011084 recovery Methods 0.000 title claims abstract description 40
- 238000005457 optimization Methods 0.000 title claims abstract description 12
- 238000011156 evaluation Methods 0.000 claims abstract description 4
- 238000012986 modification Methods 0.000 claims description 3
- 230000004048 modification Effects 0.000 claims description 3
- 230000007547 defect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000000034 method Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/77—Retouching; Inpainting; Scratch removal
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/40—Analysis of texture
- G06T7/41—Analysis of texture based on statistical description of texture
- G06T7/42—Analysis of texture based on statistical description of texture using transform domain methods
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Probability & Statistics with Applications (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Other Investigation Or Analysis Of Materials By Electrical Means (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
The invention discloses a phase recovery improvement algorithm based on MSE (mean square error) optimization, which comprises the following steps: a1, taking a random Fourier phase as an input, acquiring a reconstruction result of a phase recovery algorithm, and calculating a mean square error between a Fourier amplitude of the reconstruction result and an initial Fourier amplitude of a target; a2, repeating the step A1 by taking a plurality of different random Fourier phases as input, and obtaining a plurality of corresponding reconstruction results and a plurality of corresponding mean square errors; and A3, selecting the optimal reconstruction result from the plurality of reconstruction results as the final output by taking the mean square error as the evaluation index of the quality of the reconstruction result, and realizing the phase recovery of the target object. The invention can realize stable phase recovery.
Description
Technical Field
The invention relates to the field of computer vision and digital image processing, in particular to a phase recovery improvement algorithm based on MSE (mean square error) optimization.
Background
The traditional phase recovery algorithm is used for iteratively recovering the real Fourier phase of the target and reconstructing the spatial distribution of the target by taking a random phase as an initial guess of the Fourier phase of the target under the condition that the Fourier amplitude of the target is known but the Fourier phase is unknown. However, due to the randomness of initial phase guessing, the spatial distribution of the target reconstructed by the traditional phase recovery algorithm and the real fourier phase of the recovered target have great instability, so that accurate and effective spatial information of the target is difficult to acquire from a single phase recovery result, and the applicability of the traditional phase recovery algorithm is limited.
The above background disclosure is only for the purpose of assisting understanding of the inventive concept and technical solutions of the present invention, and does not necessarily belong to the prior art of the present patent application, and should not be used for evaluating the novelty and inventive step of the present application in the case that there is no clear evidence that the above content is disclosed before the filing date of the present patent application.
Disclosure of Invention
The invention mainly aims to overcome the defects of the prior art and provides a phase recovery improved algorithm based on MSE optimization, wherein the optimal reconstruction result is selected from the reconstruction results taking a plurality of different random phases as input by calculating the Mean Square Error (MSE) between the reconstruction result of the traditional phase recovery algorithm and the Fourier amplitude of a target object and taking the MSE as a judgment index of the goodness and badness of the reconstruction result so as to realize stable phase recovery and target object reconstruction.
In order to achieve the purpose, the invention provides the following technical scheme:
a phase recovery improvement algorithm based on MSE optimization, comprising the steps of:
a1, taking a random Fourier phase as an input, acquiring a reconstruction result of a phase recovery algorithm, and calculating a mean square error between a Fourier amplitude of the reconstruction result and an initial Fourier amplitude of a target;
a2, repeating the step A1 by taking a plurality of different random Fourier phases as input, and obtaining a plurality of corresponding reconstruction results and a plurality of corresponding mean square errors;
and A3, selecting the optimal reconstruction result from the plurality of reconstruction results as the final output by taking the mean square error as the evaluation index of the quality of the reconstruction result, and realizing the phase recovery of the target object.
The invention has the beneficial effects that: the method comprises the steps of taking a plurality of different random Fourier phases as input to reconstruct the spatial distribution of a target object, taking the mean square error between a reconstructed Fourier amplitude and an initial Fourier amplitude as a judgment index of the quality of a reconstruction result, selecting an optimal reconstruction result as algorithm output, and finally realizing stable phase recovery and target object reconstruction.
Further, in step a1, on the premise that the initial fourier amplitude of the object is known, a phase recovery algorithm is executed with a random fourier phase as the initial guess of the phase, and the spatial distribution of the object is iteratively reconstructed to obtain the reconstruction result.
Further, in step A1, the initial Fourier amplitude | F (O) | of the target is known, with a random Fourier phase θmAs an initial guess of the phase, the spatial distribution O of the object is iteratively reconstructedmObtaining a reconstruction result; and for the reconstruction result OmThe Fourier amplitude | F (O) of the reconstruction result is calculatedm) Mean square between | and initial Fourier amplitude of target | F (O) |Error of the measurement
Wherein (x, y) represents a two-dimensional coordinate distribution of a fourier domain of the object; H. v represents the maximum range of two-dimensional coordinates of the fourier domain of the object, respectively.
Further, in step A1, result O is reconstructedmFourier transform and then modulus are carried out to obtain Fourier amplitude | F (O) of a reconstruction resultm)|。
Further, in step a3, the selecting an optimal reconstruction result from the plurality of reconstruction results with the mean square error as a criterion of the quality of the reconstruction result includes: and selecting the reconstruction result with the minimum mean square error as a final output.
Further, in step a3, the mean square error is used as a criterion for evaluating the quality of the reconstruction result, and an optimal reconstruction result is selected from a plurality of reconstruction results, using the following formula:
where M is 1, 2.., M denotes the number of input random fourier phases.
Further, in step a1, the initial fourier amplitude of the object is substituted as a known signal into the phase recovery algorithm.
Drawings
Fig. 1 is a flow chart of the phase recovery improvement algorithm based on MSE optimization of the present invention.
Detailed Description
The invention is further described with reference to the following figures and detailed description of embodiments.
In order to achieve stable and effective reconstruction of spatial distribution and fourier phase of a target under the condition that the fourier amplitude of the target is known and the fourier phase of the target is unknown, the embodiment of the invention provides a phase recovery improvement algorithm based on MSE optimization, and referring to fig. 1, the phase recovery improvement algorithm comprises the following steps a 1-A3:
step A1, with random Fourier phase thetamAnd as input, acquiring a reconstruction result of the phase recovery algorithm, and calculating a mean square error between a Fourier amplitude of the reconstruction result and an initial Fourier amplitude of the target. The phase recovery improvement algorithm of the invention is mainly provided aiming at the defects of the traditional phase recovery algorithm, in the step A1, the initial Fourier amplitude | F (O) | of the target object is taken as a known signal to be substituted into the traditional phase recovery algorithm, and meanwhile, the random Fourier phase theta is usedmAs the initial guess of the phase of the traditional phase recovery algorithm, iteration is carried out to reconstruct the spatial distribution O of the target objectmAnd obtaining a reconstruction result. For the reconstruction result OmThe Fourier amplitude | F (O) of the reconstruction result is calculatedm) Mean square error between | and initial Fourier amplitude of target | F (O) |
Wherein the Fourier amplitude | F (O) of the reconstructed resultm) I is obtained by reconstructing the result OmAnd performing Fourier transform and then taking a module to obtain the product.
Wherein a Fourier amplitude | F (O) of the reconstruction result is calculatedm) Mean square error between | and initial Fourier amplitude of target | F (O) |
The formula of (1) is as follows:
wherein (x, y) represents a two-dimensional coordinate distribution of a fourier domain of the object; H. v represents the maximum range of two-dimensional coordinates of the fourier domain of the object, respectively.
Step a2, repeating step a1 with a plurality of different random fourier phases as inputs, resulting in a plurality of corresponding reconstruction results and a plurality of corresponding mean square errors. The phase recovery improvement algorithm of the present invention requires a mean square error based optimization from a plurality of reconstruction results, and therefore requires a plurality of different random fourier phases as inputs to obtain a plurality of reconstruction results and their mean square errors, thereby performing the optimization.
And A3, selecting the optimal reconstruction result from the plurality of reconstruction results as the final output by taking the mean square error as the evaluation index of the quality of the reconstruction result, and realizing the phase recovery of the target object. Specifically, the reconstruction result with the minimum mean square error is selected as the final output. The specific formula is as follows:
where M is 1, 2.., M denotes the number of input random fourier phases, OoptimalThe selected optimal reconstruction result is the final output of the improved algorithm for phase recovery of the invention, thereby realizing stable phase reconstruction. In some specific embodiments, the value of M may be, for example, 100, 150, 50, etc., without limitation.
The foregoing is a more detailed description of the invention in connection with specific preferred embodiments and it is not intended that the invention be limited to these specific details. For those skilled in the art to which the invention pertains, several equivalent substitutions or obvious modifications can be made without departing from the spirit of the invention, and all the properties or uses are considered to be within the scope of the invention.
Claims (6)
1. A phase recovery improvement algorithm based on MSE optimization is characterized by comprising the following steps:
a1, under the premise of knowing the initial Fourier amplitude of the object, inputting the random Fourier phase as the initial guess of the phase, executing a phase recovery algorithm, iteratively reconstructing the spatial distribution of the object, obtaining the reconstruction result of the phase recovery algorithm, and calculating the mean square error between the Fourier amplitude of the reconstruction result and the initial Fourier amplitude of the object;
a2, repeating the step A1 by taking a plurality of different random Fourier phases as input, and obtaining a plurality of corresponding reconstruction results and a plurality of corresponding mean square errors;
and A3, selecting the optimal reconstruction result from the plurality of reconstruction results as the final output by taking the mean square error as the evaluation index of the quality of the reconstruction result, and realizing the phase recovery of the target object.
2. The MSE-based phase retrieval improvement algorithm of claim 1, wherein in step a1, the initial fourier magnitude | f (o) | of the object is known, with a random fourier phase θmAs an initial guess of the phase, the spatial distribution O of the object is iteratively reconstructedmObtaining a reconstruction result; and for the reconstruction result OmThe Fourier amplitude | F (O) of the reconstruction result is calculatedm) Mean square error between | and initial Fourier amplitude of target | F (O) |
Wherein (x, y) represents a two-dimensional coordinate distribution of a fourier domain of the object; H. v represents the maximum range of two-dimensional coordinates of the fourier domain of the object, respectively.
3. The MSE-based phase recovery modification algorithm of claim 2, wherein in step a1, the reconstructed result OmFourier transform and then modulus are carried out to obtain Fourier amplitude | F (O) of a reconstruction resultm)|。
4. The phase recovery improvement algorithm based on MSE selection according to claim 2 or 3, wherein the mean square error is used as a criterion of the goodness or the badness of the reconstruction result in step a3, and the selecting the optimal reconstruction result from the plurality of reconstruction results comprises: and selecting the reconstruction result with the minimum mean square error as a final output.
5. The phase recovery improvement algorithm based on MSE optimization according to claim 4, wherein in step a3, the mean square error is used as a criterion of the goodness or the badness of the reconstruction result, and the optimal reconstruction result is selected from the plurality of reconstruction results, using the following formula:
where M is 1, 2.., M denotes the number of input random fourier phases.
6. The MSE-based phase recovery modification algorithm of claim 1, wherein in step a1, the initial fourier magnitude of the object is substituted into the phase recovery algorithm as a known signal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910309968.5A CN110047048B (en) | 2019-04-17 | 2019-04-17 | Phase recovery improved algorithm based on MSE (mean square error) optimization |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910309968.5A CN110047048B (en) | 2019-04-17 | 2019-04-17 | Phase recovery improved algorithm based on MSE (mean square error) optimization |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110047048A CN110047048A (en) | 2019-07-23 |
| CN110047048B true CN110047048B (en) | 2021-03-02 |
Family
ID=67277537
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910309968.5A Active CN110047048B (en) | 2019-04-17 | 2019-04-17 | Phase recovery improved algorithm based on MSE (mean square error) optimization |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110047048B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111325682B (en) * | 2020-01-21 | 2023-05-23 | 清华大学深圳国际研究生院 | Phase recovery improvement method and device for self-correlation signal with interference |
Family Cites Families (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102141627A (en) * | 2010-02-03 | 2011-08-03 | 中国科学院光电研究院 | Burst type navigation signal system and receiving method |
| US9772353B2 (en) * | 2013-06-04 | 2017-09-26 | Tektronix, Inc. | Equivalent-time sampling technique for non-coherently modulated signals |
| CN103745427B (en) * | 2014-01-16 | 2017-01-04 | 浙江传媒学院 | Image encryption method based on iteration amplitude-phase recovery algorithms |
| CN104808205B (en) * | 2015-05-06 | 2017-09-29 | 中国科学院电子学研究所 | Sparse microwave imaging method based on PhaseLift autofocus algorithms |
| CN106067156B (en) * | 2016-04-06 | 2019-07-09 | 浙江农林大学 | Digital image encryption method based on Phase Retrieve Algorithm and principle of interference |
| CN108550108B (en) * | 2017-09-28 | 2020-11-03 | 武汉大学 | Fourier laminated imaging image reconstruction method based on phase iteration minimization |
| CN108364309B (en) * | 2018-02-09 | 2020-09-01 | 清华大学深圳研究生院 | Space light field recovery method based on handheld light field camera |
| CN109060122B (en) * | 2018-07-05 | 2021-02-12 | 安徽大学 | Two-step phase recovery method, equipment and system based on single intensity measurement |
| CN109472842A (en) * | 2018-12-17 | 2019-03-15 | 长沙理工大学 | A kind of phase recovery imaging reconstruction method of no lens imaging |
-
2019
- 2019-04-17 CN CN201910309968.5A patent/CN110047048B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN110047048A (en) | 2019-07-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111062880B (en) | Underwater image real-time enhancement method based on condition generation countermeasure network | |
| CN107154064B (en) | Natural image compressed sensing method for reconstructing based on depth sparse coding | |
| CN105205788B (en) | A kind of denoising method for high-throughput gene sequencing image | |
| CN113538246B (en) | Remote sensing image super-resolution reconstruction method based on unsupervised multi-stage fusion network | |
| CN109215118B (en) | Incremental motion structure recovery optimization method based on image sequence | |
| CN109060122B (en) | Two-step phase recovery method, equipment and system based on single intensity measurement | |
| CN111899168A (en) | Remote sensing image super-resolution reconstruction method and system based on feature enhancement | |
| CN113238227B (en) | Improved least square phase unwrapping method and system combined with deep learning | |
| CN111079893B (en) | Acquisition method and device for generator network for interference fringe pattern filtering | |
| CN116563110A (en) | Blind image super-resolution reconstruction method based on Bicubic downsampling image space alignment | |
| CN114119987A (en) | Feature extraction and descriptor generation method and system based on convolutional neural network | |
| CN110047048B (en) | Phase recovery improved algorithm based on MSE (mean square error) optimization | |
| CN111080540B (en) | Training method of image restoration model and computer equipment | |
| CN102184451A (en) | Multi-wavelet fusion feature based neuron action feature extraction method | |
| CN113111706B (en) | SAR target feature unwrapping and identifying method for azimuth continuous deletion | |
| CN105931184B (en) | SAR image super-resolution method based on combined optimization | |
| CN114266850A (en) | Point cloud completion method based on continuous standardized stream and anti-refinement network | |
| CN103824286A (en) | Singular value decomposition-random sample consensus (SVD-RANSAC) sub-pixel phase correlation matching method | |
| CN117788284A (en) | DEM super-resolution reconstruction method and system based on dual-channel filter | |
| CN117408991A (en) | Image anomaly detection method and device based on reconstruction resistance and storage medium | |
| CN108510464B (en) | Compressed sensing network based on block observation and full-image reconstruction method | |
| CN114565772B (en) | Method and device for extracting set features, electronic equipment and storage medium | |
| CN114399547B (en) | Monocular SLAM robust initialization method based on multiframe | |
| CN109615584A (en) | A kind of SAR image sequence MAP super resolution ratio reconstruction method based on homography constraint | |
| CN105301655B (en) | A kind of common imaging gather line noise minimizing technology and device |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |