Multi-parameter ocular axial measurement and data processing system based on self-mixing vernier effect

Abstract

The application discloses a multi-parameter eye axis measurement and data processing system based on self-mixing vernier effect and relates to the field of optical fibers. A low-coherence light source is divided into a fiber reference light and a fiber measurement light through a first coupler; the fiber measurement light is output as spatial measurement light through a circulator and a collimator, scans different parts of an eyeball through a zoom lens, enters a fiber delay line through a circulator and a wavelength division multiplexer, is reflected and output from a second circulator; a high-coherence light source is reflected through a wavelength division multiplexer and a fiber delay line, self-mixing interference is generated at a phase shift grating, and a low-coherence peak pitch is calibrated; the reflected measurement light is output as low-coherence light from the circulator, interference is generated at a second coupler, and the low-coherence light is sent into a balanced photoelectric detector; spatial light reflected by the eyeball is focused after being reflected by a Hartmann micro-lens group, a Hartmann array diagram is collected, and diopter is calculated. The scheme uses optical fibers and self-mixing interference to calibrate measurement data, and the signal-to-noise ratio of the data is higher, and the data is more accurate.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the field of optical fiber technology, and particularly relate to a multi-parameter eye axis measurement and data processing system based on self-mixing vernier effect. BACKGROUND

[0002] At present, human health problems gradually become the focus of the whole society, among which ophthalmic problems are particularly valued. Most ophthalmic problems can be reflected from the eye axial parameters and the refractive power. The monitoring of the eye axial parameters and the refractive power can help ophthalmic experts to understand the changes of the eyeball structure in time and guide the treatment and management of related diseases. The eye axial parameters mainly include the axial length, the lens thickness, the corneal thickness and the anterior chamber depth, etc. The low coherence light interference technology, which combines the confocal microscope technology and the optical coherence tomography (OCT) technology with the advantages of white light coherence imaging, can realize the measurement of the micron level resolution in the biological tissue without causing secondary infection, and has gradually become the mainstream technology scheme for the non-contact and high-precision eye axial parameter measurement.

[0003] However, most of the related art eye axial biological measurement instruments adopt a spatial light path structure. The detection of the distance between the measured low coherence peaks needs to use a clock system to calibrate the position of the scanning delay line, which limits the utilization rate of the scanning light beam and the integration degree of the system, and affects the precision measurement of the signal-to-noise ratio and the sensitivity. Finally, the tested eyeball parameters cannot guarantee the high precision requirement. SUMMARY

[0004] Embodiments of the present application provide a multi-parameter eye axis measurement and data processing system based on self-mixing vernier effect to solve the problems. The system comprises:

[0005] a low coherence light source, a light beam emitted by the low coherence light source is divided into a fiber reference light and a fiber measurement light through a first coupler; the fiber measurement light passes through a first circulator and a collimator to output a spatial measurement light, the spatial measurement light focuses a scanning light beam at different parts of the eyeball through a zoom lens; the fiber measurement light passes through a second circulator and a wavelength division multiplexer to enter a fiber delay line, and is output from the reflection end of the second circulator after being reflected by the fiber delay line;

[0006] a high coherence light source, a light beam emitted by the high coherence light source passes through a doped fiber with a phase shift grating, and is reflected by a wavelength division multiplexer and a fiber delay line to generate self-mixing interference at the phase shift grating, so as to calibrate the low coherence peak distance; the spatial measurement light is reflected by the eyeball and then output from the first low coherence light through the collimator and the reflection end of the first circulator; the sample measurement light is output from the second low coherence light through the reflection end of the second circulator; and the two beams of low coherence light generate interference through a second coupler;

[0007] The Hartmann micro-lens group is used to focus the spatial light reflected by the eyeball on a CCD photosensitive surface, and the diopter is calculated by the Hartmann array diagram collected by the Hartmann array.

[0008] The balanced photodetector receives two interference signals output by the second coupler after interference and performs photoelectric conversion, adjusts a filter window according to a frequency value of a current signal after photoelectric conversion, and obtains eyeball data according to low-coherence peak spacing measurement after window filtering.

[0009] Specifically, the balanced photodetector obtains an original current signal after photoelectric conversion, and converts a time domain into a frequency domain through fast Fourier transform (FFT) conversion.

[0010] A band-pass FIR filter is designed and a window is set according to a frequency range of the interference signal and a band-pass frequency; and the center frequency of the band-pass filtering is the maximum frequency value identified in the frequency domain signal;

[0011] The envelope is extracted and the position is determined through Hilbert transform.

[0012] The Hilbert envelope is subjected to Fourier transform to obtain envelope spectrum data.

[0013] A low-pass filter is set for the envelope spectrum data according to a frequency range of the interference signal, secondary filtering is performed, and a filtering combination is determined according to a sum of signal-to-noise ratios; the minimum frequency value in the frequency domain signal is identified during the secondary filtering, and the low-pass filter is set according to the minimum frequency value.

[0014] Specifically, the filtering combination is determined according to the sum of signal-to-noise ratios, including:

[0015] The interference signals before and after the secondary filtering are obtained, and the signal-to-noise ratios are calculated according to adjacent interference peaks.

[0016] The sum of the signal-to-noise ratios calculated under different setting parameters before and after the secondary filtering is sorted in descending order, and the filtering combination with the largest sum of signal-to-noise ratios is selected as a target filtering strategy in a period.

[0017] Specifically, the real signal of the interference signal after the first filtering is , the Hilbert transform virtual signal is represented as follows:

[0018]

[0019] The analytical signal constructed according to the real signal and the virtual signal is represented as:

[0020]

[0021] wherein the analytical signal The length of the module represents the envelope of the original signal.

[0022] Specifically, the signal intensity received by the balanced photodetector I is expressed as follows:

[0023]

[0024]

[0025]

[0026] wherein and respectively represent the center wavelength and the spectral full width at half maximum of the low-coherence light source, represents the energy of the received interference signal I ; represents the DC components of the two low-coherence electrical signals I1 and I2, represents the change amount after interference; represents the change amount of the reference light phase;

[0027] In the process of measuring the signal intensity I , the reference light phase is changed by changing the internal posture of the fiber delay line , and the axial parameter is inversely deduced based on the adjustment posture size.

[0028] Specifically, the interference signal generated by the high-coherence light source presents periodic changes, and the low-coherence peak distance D value is determined by comparing the interference peaks of the interference signals generated by the low-coherence light source, and is expressed as follows:

[0029]

[0030] wherein is the wavelength of the high-coherence laser, is the refractive index of air at the wavelength is the number of complete interference fringes, is the phase difference of the incomplete interference fringes, is the overall phase difference.

[0031] Specifically, the high-coherence light source, the phase-shift grating, the doped fiber, the wavelength division multiplexer and the fiber delay line form a three-mirror cavity model, the phase-shift grating is equivalent to the first mirror and the second mirror of the three-mirror cavity model, and the fiber delay line is equivalent to the third mirror. The adjustment posture of the fiber delay line is determined according to the self-mixing interference formula, and is used as a signal ruler for low-coherence interference, and is expressed as follows:

[0032]

[0033] ​ Represents the group refractive index. Indicates the line width expansion factor. This represents the cavity length of the equivalent resonant cavity of the phase-shifting grating. Coupling efficiency, and These represent optical frequency and critical frequency, respectively. Represents the speed of light. This represents the equivalent external cavity round-trip delay between the phase-shifting grating and the fiber delay line.

[0034] Specifically, a two-dimensional galvanometer is set in the spatial optical path between the zoom lens and the eyeball. Spatial light is focused onto different areas of the eyeball through the two-dimensional galvanometer and reflects eyeball data from different areas.

[0035] A beam splitter and a reflector are set in the spatial optical path between the zoom lens and the two-dimensional galvanometer to introduce part of the light beam reflected by the eyeball into the Hartmann microlens optical path. The refractive power of the eyeball is calculated by the Hartmann array diagram acquired by the Hartmann optical path.

[0036] Specifically, when the wavefront to be measured is received by the microlens array, it is divided into countless sub-wavefronts by the microlens array, and then converged on the CCD to form a light spot array pattern containing the information of the wavefront to be measured; the position offset of the light spot is calculated based on the light spot array pattern, and the diopter and divergence data are calculated.

[0037] The process of calculating the position offset of the light spot is based on constructing five sets of two-dimensional surfaces using Zernike polynomials. The partial derivatives of the surfaces are used to determine the position offset of each sampling point. The surface polynomials and partial derivatives are expressed as follows:

[0038]

[0039] Obtain at least one set of light spots offset Based on the coordinates of the reference point Invert the coefficients of the five polynomials; offset It is expressed as follows:

[0040]

[0041] The partial derivatives at the sampling points are expressed as follows:

[0042]

[0043] Among them Represents the wavefront function. and Indicates the offset value. and For the corresponding derivative, Let n be the focal length of the lens, and n is 5.

[0044] Specifically, based on the partial derivatives at the sampling points and the correspondence between matrices, the polynomial coefficient identity is derived as follows:

[0045]

[0046] The Zernike coefficients corresponding to the wavefront phase difference are obtained using the inverse matrix, as follows:

[0047]

[0048] Through five coefficients The refractive power and divergence are calculated and expressed as follows:

[0049]

[0050]

[0051] Indicates the diopter value. This represents the divergence value.

[0052] The beneficial effects of the technical solutions provided in this application include at least the following:

[0053] This scheme utilizes time-domain low-coherence interferometry to measure various parameters along the axial direction of the eyeball. The time-domain low-coherence interferometry measurement system consists of a laser, a reference arm, a measuring arm, a delay device, and a detector. The measuring light, carrying the information to be measured, undergoes low-coherence optical interference with the reference light to obtain interference peaks on various surfaces of the eye tissue. Based on the time-domain low-coherence interferometry fiber optic structure system, interference signals are collected. However, the resulting interference signals, without processing, are typically difficult to use for high-precision biological measurements. Therefore, signal filtering is usually required. The filtering technique involves a series of operations, including Fourier transform and FIR bandpass filtering of the original time-domain data, as well as Hilbert transform and low-pass filtering of the pre-processed signal. A cyclic process involving frequency and filter order is introduced, combined with signal-to-noise ratio calculations to select the optimal filtering combination. This ultimately achieves short-time processing of the obtained interference signal with a high signal-to-noise ratio. The processing process and results are visualized for easy observation of signal changes. Finally, combined with high-coherence interferometry, using the high-coherence interference signal as a scale, precise calibration of the positional information of the eye axial direction can be achieved. Attached Figure Description

[0054] Figure 1 This is a schematic diagram of the structure of a multi-parameter axial length measurement and data processing system based on the self-mixing vernier effect provided in one embodiment of the application;

[0055] Figure 2 This is a theoretical schematic diagram of a three-mirror cavity model constructed using laser feedback interferometry.

[0056] Figure 3 is another application embodiment provided by the high and low coherence measurement combined with filter design of the structure schematic diagram of high-resolution measurement system of axial length;

[0057] Figure 4 Exemplary schematic diagram of Hartmann array diagram is shown;

[0058] Figure 5 Listed are the schematic diagram of original data time domain diagram and original data frequency domain diagram;

[0059] Figure 6 A possible FIR band-pass filter diagram is shown;

[0060] Figure 7 is listed a possible Hilbert envelope diagram;

[0061] Figure 8 is the frequency domain diagram of envelope after data processing;

[0062] Figure 9 A possible filter effect diagram is shown. DETAILED DESCRIPTION

[0063] In order to make the purpose, technical scheme and advantages of the present application more clear, the embodiments of the present application will be further described in detail below with reference to the drawings.

[0064] In this paper, "a plurality of" refers to two or more. "And / or", which describes the association between the associated objects, means that there can be three relationships, for example, A and / or B, which can mean: A exists alone, A and B exist simultaneously, and B exists alone. The character " / " generally represents that the associated objects before and after are a "or" relationship.

[0065] Figure 1 is the structure schematic diagram of the multi-parameter axial measurement and data processing system based on self-mixing vernier effect provided by the embodiments of the present application. The whole contains a low coherence interferometric measurement system part composed of a low coherence light source, a circulator and a coupler, a high coherence auxiliary measurement system part composed of a high coherence light source and a fiber delay line, etc. In addition, it also includes a balanced photodetector for measuring interference data, a filter processing algorithm after photoelectric conversion, etc. After the balanced photodetector measurement and filter processing, the relevant eyeball data can be obtained.

[0066] The low-coherence light source is responsible for emitting a corresponding wavelength beam, the output of which is connected to the first coupler through an optical fiber, the first coupler is divided into two paths, and is connected to the first and second circulators through optical fibers respectively. The direct output end of the first circulator is connected to the collimator through the optical fiber, and the light goes out through the collimator and is converted into spatial light transmission. The direct output end of the second circulator is connected to the wavelength division multiplexer (i.e. wavelength division multiplexer 2) through an optical fiber, and the output end of the wavelength division multiplexer is connected to the optical fiber delay line. The spatial measurement light is reflected by the eyeball and then output from the first low-coherence light through the collimator and the reflection end of the first circulator, and the sample measurement light is output from the second low-coherence light through the reflection end of the second circulator; the two beams of low-coherence light interfere through the second coupler, and are detected by the balanced photodetector after interference.

[0067] The optical path is described as follows: the light beam emitted by the low-coherence light source is divided into fiber reference light and fiber measurement light through the first coupler. The fiber measurement light is output as spatial measurement light through the first circulator and the collimator, and the spatial measurement light is focused on different parts of the eyeball through the zoom lens. The fiber measurement light enters the optical fiber delay line through the second circulator and the wavelength division multiplexer, and is output from the reflection end of the second circulator after reflection.

[0068] In some embodiments, the low-coherence light source emits light at 850 nm, which is divided into two beams that can interfere with each other after passing through a 95:5 coupler 1. The polarization directions of the two beams are consistent and the frequencies are the same. Subsequently, the two beams are reflected by different reflecting surfaces and return to the coupler 3 to interfere. Within the interference distance, the light intensity of the superimposed wave is maximum when the phase difference of the two light waves is 0 or an integer multiple. By changing the position of the reference arm to adjust the phase of the reference light, the maximum interference light intensity is generated with the light reflected by different eye tissues in the measurement arm. Thus, the movement distance of the reference arm can be used to deduce the eyeball parameters such as the axial length, cornea and pre-optic segment. The reference arm and the measurement arm mentioned here respectively refer to the output path part of the two circulators. In Figure 1 , the reference arm is the optical fiber delay line after the output of the circulator 2; and the measurement arm is the spatial optical path device group after the collimating lens and the eyeball part.

[0069] Based on the consideration of measurement accuracy, the system uses a low-coherence light source with a certain bandwidth. The signal received on the detector is the superposition of the interference of all wavelengths within the bandwidth. Therefore, the signal intensity I received by the balanced photodetector is the superposition of the interference intensity of each wavelength, which can be represented as:

[0070]

[0071]

[0072]

[0073] wherein and respectively represent the center wavelength and the spectral full width at half maximum of the low-coherence light source, represents the energy of the received interference signal I ; represents the sum of the direct current components of the two low-coherence electrical signals, represents the change amount after interference; represents the change amount of the phase of the reference light.

[0074] Because the eyeball parameters are inversely deduced by the moving distance of the reference arm, that is, the internal posture of the fiber delay line is changed in the process of measuring the signal strength I , the purpose is to change the phase of the reference light , and finally the eyeball parameters are inversely deduced based on the adjusted posture data and the measured related parameter size.

[0075] In order to meet the requirements of low-coherence measurement accuracy and scanning speed, we use a fiber delay line to realize high-precision and high-frequency scanning of the reference arm. In addition, in order to obtain accurate phase change amount, the system introduces high-coherence interference technology to assist measurement. The application selects a fiber laser as a high-coherence light source, and a phase-shift grating doped fiber is connected to the high-coherence light source. The doped fiber is sent to a wavelength division multiplexer 2 and then input to the fiber delay line. That is, the fiber delay line is a common part of the high-coherence and low-coherence light paths. The fiber delay line reflection will interfere with the input beam, and the phase-shift grating is located at the interference position. The interference generated at the phase-shift grating can calibrate the low-coherence peak spacing based on the vernier effect.

[0076] Specifically, when the reference arm is scanned, the low-coherence interference light path can detect the interference peaks of different optical surfaces, and the high-coherence signal strength will present periodic changes. By comparing the two kinds of interference signals, the phase difference between the low-coherence peak signals can be obtained. By comparing the interference peaks generated by the low-coherence light source, the adjacent low-coherence peak spacing D value can be determined, which is represented as follows:

[0077]

[0078] wherein is the wavelength of the high-coherence laser, is the refractive index of air at the wavelength of the high-coherence laser, is the number of complete interference fringes, is the phase difference of the incomplete interference fringes, is the overall phase difference.

[0079] In view of the fact that high coherence technology is used in some experiments, the interference vernier effect and the precision are limited, therefore, in order to further improve the integration and sensitivity and reduce the complexity of the optical path, the self-mixing effect is selected to generate interference. That is, a coherent light source, a phase-shifted grating, a doped fiber, a wavelength division multiplexer and a fiber delay line are combined to form a three-mirror cavity model, the phase-shifted grating is equivalent to the first and second mirrors of the three-mirror cavity model, and the fiber delay line is equivalent to the third mirror. The adjustment posture of the fiber delay line is determined according to the self-mixing interference formula , as a low coherence interference signal scale.

[0080] Figure 2 is a theoretical schematic diagram of a three-mirror cavity model formed by using laser feedback interference technology, M1, M2 and M3 respectively represent three mirrors, after the laser Laser is emitted, the beam is emitted through M2, after being reflected at M3 Target, it returns to the position between M1 and M2 again, and self-mixing interference is generated with the incident beam at this position. represents the effective refractive index of the phase-shifted grating intracavity, represents the effective refractive index of the external cavity.

[0081] Considering that the intensity reflectivity of the output mirror of the laser, that is, the second mirror M2, is , wherein R2 represents the reflectivity of the mirror. The third mirror M3 affects the second mirror in that it changes the reflectivity and phase of the second mirror. The reflectivity of the second mirror after the influence of the third mirror of the external cavity can be represented as:

[0082]

[0083] Here represents the coupling efficiency, represents the optical frequency, and represents the round-trip delay of the external cavity. At the same time, the stability condition of the laser resonant cavity still needs to be met, that is, is a positive real number. Since the laser resonant cavity still needs to be in a relatively stable state, the modulus of the imaginary part of is a small amount. When the above relatively stable condition is met, there is , and it is generally defined that .

[0084] When , can be obtained . Since the laser still maintains stable output, the phase change of the laser inside the round trip is 2π, so it can be obtained:

[0085]

[0086] , wherein is the eigen wave number of the laser resonator, is the cavity length of the laser resonator, is the effective refractive index of the intra-cavity, is an integer. Since the effective refractive index of the intra-cavity is , we have:

[0087]

[0088] The influence of the light field change on the group refractive index is considered. The group refractive index is , and the line width broadening factor is replaced to obtain the adjusted attitude (displacement information) as a signal ruler of low coherence interference, which is expressed as follows:

[0089]

[0090] represents the group refractive index, represents the line width broadening factor, represents the cavity length of the equivalent resonator of the phase-shifted grating, coupling efficiency, and represent the optical frequency and the critical frequency, respectively, represents the speed of light, represents the equivalent external cavity round trip delay between the phase-shifted grating and the fiber delay line.

[0091] In some embodiments, the sample arm of the spatial light part can measure the eyeball parameters according to the actual situation. Since the distance from the cornea to the retina is more than 20 mm, the galvanometer scanning can only scan the image near the sample imaging point. If the detection light cannot be well focused on the layer of the tissue structure to be measured in the eye, the light signal intensity reflected by the sample arm will be reduced, and then the obtained coherent light signal intensity will be reduced, thereby affecting the detection signal-to-noise ratio. Therefore, all imaging data cannot be obtained by one-time scanning. For example Figure 1 ​To achieve high-contrast and high-resolution images of the fundus and anterior segment, this application additionally incorporates a two-dimensional galvanometer in the spatial optical path between the zoom lens and the eyeball. Spatial light is focused onto different regions of the eyeball by the two-dimensional galvanometer, reflecting data from these regions. The two-dimensional galvanometer includes mutually perpendicular X-mirrors and Y-mirrors. Light passes through the X-mirror to complete one cycle of X-Scan scanning, is reflected to the Y-mirror, and undergoes another cycle of deflection to achieve multiple X-Scan scans. This focuses the light beam onto the lens, glial membrane, fundus, and anterior segment interface of the eye tissue under test, performing a two-dimensional scan to achieve scanning of multiple measurement planes, thereby improving the signal-to-noise ratio. During focusing, the system divides the measurement area into three parts: the first part includes the retina, the second part includes the posterior surface of the lens, and the third part includes the anterior surface of the lens and the cornea. By pre-calculating the focusing position of the light and changing the lens group position, zoom scanning can be achieved.

[0092] It should be noted that the ocular parameters mentioned here do not include ocular refractive power and aberration, which need to be measured and calculated using a Hartmann lens system. Assuming that ocular refractive power and aberration are not considered, and the system is specifically designed for measuring axial length, it can be further simplified to... Figure 3 The diagram shows a high-resolution eye axis measurement system designed with high and low coherence measurements combined with filtering.

[0093] The high-coherence interferometry system for measuring axial length using a combination of high and low coherence measurements and filtering also includes a low-coherence interferometry system and a high-coherence auxiliary measurement system. The difference lies in the elimination of the Hartmann lens group and the two-dimensional galvanometer. In other words, it directly adjusts the parameters to measure the axial length data and generates normal reflection and interference.

[0094] Because the first low-coherence light and the second low-coherence light have equal power, the two interference signals are input to a balanced photodetector via positive and negative inputs, respectively, for differential amplification and comparison processing. The converted two current signals are 180 degrees out of phase. The differential amplification formula is as follows:

[0095]

[0096] Indicates differential amplification output. and These represent two power signals respectively. G represents the responsivity of a photodiode at a given wavelength, where G is the gain coefficient and M is the optical gain factor.

[0097] Due to current signal and There is a phase difference of π, after differential processing, the direct current signal is eliminated, and the amplitude of the alternating current signal in the output signal is doubled. In addition, due to the use of differential amplification in balanced detection, the common mode noise signal in the two detectors can be greatly suppressed, and the signal-to-noise ratio of the output signal of the multi-parameter ophthalmic biological measurement system can be improved.

[0098] It is particularly pointed out that considering the eyeball tissue characteristics, the transmittance of light with a wavelength of 850 nm in water is very high, and the power loss of light with a wavelength of 850 nm in water caused by absorption is small, and the main component in human eye tissue is water, and the water content in the vitreous body is more than 90%. Therefore, we selected 850 nm wavelength as the light source of the low coherence interference system. The high coherence light source of the present application is composed of 980 nm pump and wavelength division multiplexer (i.e. wavelength division multiplexer 1), and the wavelength division multiplexer 1 is connected between the pump and the phase shift grating on the optical fiber, and the wavelength division multiplexer 1 is connected to the photodetector, and the self-mixing interference signal is detected by the photodetector to determine the signal scale. It should be noted that the 980 nm pump will be excited to 1550 nm laser by the doped fiber, and since two different wavelength signals exist, the 980 nm pump will be matched with the wavelength division multiplexer as a whole light source device, i.e. high coherence light source. The wavelength division multiplexer 2 connected to the optical fiber delay line is used to cooperate with the wavelength division multiplexer 1 to allow the two different wavelength beams to transmit without interference.

[0099] In summary, the system is divided into two parts: time domain low coherence measurement and high coherence interference auxiliary measurement, and is connected by all-fiber structure. This design not only saves the work of adjusting the spatial light path, but also helps to reduce the size of the system and facilitate integration. At the same time, polarization maintaining fiber is selected for connection in the interference light path part, which avoids the change of the polarization of the transmitted light caused by external environmental factors such as temperature, pressure or bending, ensures the consistency of the polarization states of the two interference lights, and further improves the signal-to-noise ratio of the system.

[0100] Further, considering that Figure 1 The system needs to calculate the diopter and dispersion information, so a beam splitter and a mirror need to be arranged in the spatial light path between the zoom lens and the two-dimensional galvanometer to introduce the part of the light beam reflected by the eyeball into the Hartmann microlens light path, and the Hartmann array diagram collected by the Hartmann light path is used to calculate the diopter of the eyeball. The following are the steps of collecting the Hartmann array diagram and calculating the diopter parameters:

[0101] S1, when the wavefront to be measured is received by the microlens array, it is divided into countless sub-wavefronts by the microlens array, and converged into a spot array diagram containing the wavefront information to be measured on the CCD;

[0102] The Hartmann wavefront imaging optical path is composed of a plurality of microlenses arranged in order and closely with the same parameters. When the wavefront to be measured is received by the microlens array, the microlens array divides the wavefront to be measured into an infinite number of sub-wavefronts, and finally converges into an array of spot arrays containing wavefront information to be measured on the CCD.

[0103] S2, calculating the spot position offset based on the spot array image;

[0104] The process of calculating the spot position offset is based on the construction of five groups of two-dimensional curved surfaces based on Zernike polynomials to back-calculate the curvature and divergence data. The surface partial derivative is determined as the position offset of each sampling point in the embodiment of the application, and the polynomials and partial derivatives constructed based on the Zernike multi-curved surface are as follows:

[0105]

[0106] In the above formula, the polynomial is determined according to the selected number and Zernike polynomial, and the application does not make too much explanation. The offset of at least one group of light points is obtained , and the coefficient values of the five polynomials are inverted according to the reference point coordinates . The offset is as follows:

[0107]

[0108] The sampling point partial derivative is as follows:

[0109]

[0110] Among them, represents the wavefront function, and represent the offset value, and are the corresponding derivatives, is the focal length of the lens, and n is 5 in the embodiment, indicating 5 curved surface polynomials.

[0111] S3, calculating the refractive power and divergence data based on the spot position offset.

[0112] Further, when the above relationship is obtained, the polynomial coefficient identity is derived according to the corresponding relationship between the sampling point partial derivative and the matrix, and is as follows:

[0113]

[0114] The Zernike coefficients corresponding to the wavefront difference are obtained by using the inverse matrix, and are as follows:

[0115]

[0116] By five coefficients The diopter and the divergence are calculated and expressed as follows:

[0117]

[0118]

[0119] The diopter value is expressed as: The divergence value is expressed as:

[0120] Figure 4 The schematic diagram of the Hartmann array diagram is exemplarily shown, the light beam is focused on the CCD photosensitive surface, and the Hartmann array diagram is obtained by the data acquisition unit of the CCD, and the accurate diopter (S) and divergence (C) data can be obtained by comparing the standard Hartmann array diagram (S=0C=0). Figure 4 The left side in the middle is the standard Hartmann array diagram, and the middle and right sides are the Hartmann array diagrams under different degrees of diopter and divergence.

[0121] Based on the above high and low coherence measurement and combined with the self-mixing vernier effect and the Hartmann microlens group, the measurement of the eye axis and the diopter can be realized, that is, a multi-parameter eye axis measurement system based on the self-mixing vernier effect is composed, and the system diagram and Figure 1 The structures are the same.

[0122] Considering that the light path of the high and low coherence light source is converted and reflected for many times, and finally the electrical signal converted by the balanced detector will have more interference. In order to ensure the accuracy of data processing and detection, the embodiment selects to filter the data. After comparing various filtering methods, the application finally designs a filtering method for filtering frequency self-adjusting of interference signal processing. The following is the data processing step:

[0123] S1, the original current signal is obtained after the balanced photoelectric detector is photoelectrically converted, and the time domain is converted into a frequency domain signal by FFT transformation;

[0124] The mathematical model of Fourier transform is as follows:

[0125]

[0126] Wherein is the Fourier transform result of the signal , indicating the frequency domain signal (spectrum); is the time domain signal, and w is the angular frequency; the inverse transformation formula of Fourier transform is:

[0127]

[0128] This indicates that the frequency domain signal can be restored to the original time domain signal by inverse Fourier transform.Figure 5 The schematic diagram of the original data time domain graph and the original data frequency domain graph is listed.

[0129] It should be noted that there will be a symmetric peak in the Fourier transform (due to the symmetry of complex conjugate symmetry, the Fourier transform will appear symmetric peaks at positive and negative frequencies), and the essential reason is that the mathematical properties of real-valued signals lead to the conjugate symmetry of the spectrum. This phenomenon is very important in theory and practical signal processing (such as spectrum analysis, filter design, etc.), and provides important information about the signal components. And the negative frequency part will be symmetric to the positive frequency part in matlab, which will appear two peaks in the figure below, and the right peak is actually the negative frequency symmetric peak.

[0130] S2, according to the frequency range of the interference signal, a bandpass FIR filter is designed in combination with the bandpass frequency, and a window is set;

[0131] According to the frequency range of the interference signal, a bandpass FIR filter is designed in combination with the bandpass frequency, and the bandpass FIR filter mainly includes bandpass filtering, filter order and window function, and the most important bandpass filtering includes some important parameters:

[0132] Lower cutoff frequency ( ): that is, the cutoff frequency of the low-pass filter, indicating the lowest frequency that can pass;

[0133] Upper cutoff frequency ( ): that is, the cutoff frequency of the high-pass filter, indicating the highest frequency that can pass;

[0134] Bandwidth (BW): refers to the frequency bandwidth that the bandpass filter can pass, which is usually calculated as:

[0135]

[0136] Center frequency ( ): is the center point of the frequency passed by the bandpass filter, which is usually calculated as:

[0137]

[0138] In this embodiment, the center frequency of the periodic identification bandpass filter is The maximum frequency value identified in the frequency domain signal, and then the And .

[0139] The frequency response of the bandpass FIR filter can be represented by a frequency response curve, which generally presents a bell-shaped curve (i.e. Gaussian or Lorentz curve), which means that the signal near the center frequency is transmitted most effectively, and the signal outside the cutoff frequency is gradually attenuated. Its mathematical model is:

[0140]

[0141] where is the frequency response of a bandpass filter.

[0142] In the algorithm implementation, the frequency is normalized by the target frequency and the ratio of the bandpass upper and lower cutoff frequencies to the sampling rate, and then a window function is used to implement the preliminary bandpass filtering. Meanwhile, a loop of the bandpass upper and lower cutoff frequencies and the filter order is introduced, and the loop range and its step size are adjusted to find the most suitable FIR filter combination through the loop system. Figure 6 A possible FIR bandpass filter diagram is shown.

[0143] S3, extract the envelope and determine the position by Hilbert transform;

[0144] Hilbert transform to extract the envelope and determine the position has significant advantages, especially in processing non-stationary signals, signals with large noise and complex signals, it not only can accurately extract the instantaneous characteristics of the signal, but also does not need additional preprocessing steps, and is not affected by the influence of artificial threshold setting. Therefore, the envelope of the signal after preliminary filtering is extracted by Hilbert transform.

[0145] The real signal of the filtered interference signal is set as , and the Hilbert transform virtual signal is expressed as follows:

[0146]

[0147] In essence, the signal is shifted by 90 degrees, and a complex signal is constructed with the real part and the imaginary part . The complex signal constructed according to the real signal and the imaginary signal is expressed as:

[0148]

[0149] The modulus of the complex signal represents the envelope of the original signal, and its calculation formula is:

[0150]

[0151] Figure 7 is a possible Hilbert envelope diagram listed, and such envelope signal usually represents the amplitude modulation of the signal, i.e. the amplitude change of the signal over time.

[0152] S4, Fourier transform is performed on the extracted Hilbert envelope to obtain envelope spectrum data;

[0153] The Fourier transform is performed on the Hilbert envelope to obtain a frequency domain image of the envelope as Figure 8 , and a desired interference signal frequency range is found.

[0154] S5, a low-pass filter is set for the envelope spectrum data according to the interference signal frequency range, secondary filtering is performed, and a filter combination is determined according to the sum of signal-to-noise ratios.

[0155] The process mainly obtains interference signals before and after twice filtering, calculates signal-to-noise ratios according to adjacent interference peaks. Then, the sum of signal-to-noise ratios calculated under different setting parameters before and after twice filtering is sorted in descending order, and the filter combination with the largest sum of signal-to-noise ratios is selected as the target filter strategy in the cycle. It specifically includes the upper and lower cutoff frequencies of the band-pass filter, the filter order, and the low-pass frequency of the Hilbert envelope filter.

[0156] For the determination method of the sum of signal-to-noise ratios, the present application calculates the signal-to-noise ratios of two interference peaks respectively, and then obtains the maximum value of the sum of the two items as the determination output. Because if only one interference peak is used as the basis for judgment, the frequency selection of the other interference peak will be ignored, resulting in distortion of some signals. Such a determination method can simultaneously consider the frequency part of two peaks to meet the filtering requirements.

[0157] Based on the above design idea, the present application uses the filter strategy for data processing, saves calculation time on the basis of not affecting the filtering effect, improves the step size to reduce the number of cycles. Figure 9 In the shown filter effect diagram, the calculation time is reduced from 35s to about 1.5s, while the sum of signal-to-noise ratios can still maintain about 65dB, and the signal-to-noise ratio of a single peak can maintain about 32dB. The red marked position is the front surface peak of the glass, and the green marked position is the back surface peak of the glass. Through the high signal-to-noise ratio measurement of the structure and the processing of the filter, an ocular axial high-resolution measurement system combining high and low coherence measurement and filter design is finally realized.

[0158] In summary, in the present scheme, the parameters of the eyeball axis are measured by time-domain low-coherence interference, and the time-domain low-coherence interference measurement system is composed of a laser, a reference arm, a measurement arm, a delay device (optical fiber delay line), and a detector. The interference peaks of the surfaces of the eye tissue are obtained by low-coherence light interference between the measurement light carrying the information to be measured and the reference light. Based on the time-domain low-coherence interference fiber structure system, the interference signal is obtained, and the obtained interference signal is usually difficult to realize high-precision biological measurement without processing. Therefore, the signal usually needs to be filtered and processed. In the filtering technology, a series of operations such as Fourier transform of the original time-domain data, FIR band-pass filtering, Hilbert transform of the preliminary processed signal, low-pass filtering, and the like are adopted, and the cycle process of the frequency and the filter order is introduced. The best filter combination is selected by combining the calculation of the signal-to-noise ratio, and finally the short-time and high signal-to-noise ratio processing of the obtained interference signal is realized. At the same time, the processing process and the result are visualized, which is convenient for observing the change of the processed signal. Finally, combined with the high-coherence interference technology, the high-coherence interference signal is used as a ruler, and the accurate calibration of the position information of the axial tissue of the eye can be realized.

[0159] The embodiments are only an explanation of the present application, and are not a limitation of the present application. Those skilled in the art can make modifications to the embodiments without creative contribution after reading the present specification, and the modifications are protected by the patent law as long as they are within the scope of the claims of the present application.

Claims

1. A multi-parameter axial length measurement and data processing system based on the self-mixing vernier effect, characterized in that, The system includes: A low-coherence light source emits a beam that is split into fiber reference light and fiber measurement light by a first coupler. The fiber measurement light is output as spatial measurement light by a first circulator and collimator. The spatial measurement light is focused onto different parts of the eyeball by a zoom lens. The fiber measurement light enters the fiber delay line by a second circulator and wavelength division multiplexer, and is output from the reflection end of the second circulator after being reflected by the fiber delay line. A high-coherence light source emits a beam that passes through an erbium-doped fiber connected to a phase-shifting grating. After reflection by a wavelength division multiplexer and a fiber delay line, self-mixing interference is generated at the phase-shifting grating to calibrate the low-coherence peak spacing. The spatial measurement light is reflected by the eyeball and then passes through a collimator and is output as a first low-coherence light from the reflection end of the first circulator. The sample measurement light is output as a second low-coherence light from the reflection end of the second circulator. The two low-coherence beams interfere through a second coupler. The Hartmann microlens system focuses spatial light reflected from the eye onto the CCD photosensitive surface after entering the Hartmann microlens system, and calculates the refractive power by acquiring the Hartmann array image. A balanced photodetector receives the two interference signals output from the second coupler and performs photoelectric conversion. After photoelectric conversion, the filtering window is adjusted according to the frequency value of the current signal. After window filtering, eye data is obtained by measuring the low coherence peak spacing. Specifically, after photoelectric conversion, the original current signal is obtained, and the time domain is converted to the frequency domain by Fast Fourier Transform (FFT). Design a bandpass FIR filter and set its window based on the frequency range of the interference signal and the bandpass frequency; the center frequency of the bandpass filter. The maximum frequency value identified in the frequency domain signal; The envelope was extracted and the location determined using the Hilbert transform. The extracted Hilbert envelope is subjected to Fourier transform to obtain the envelope spectrum data; A low-pass filter is set for the envelope spectrum data according to the frequency range of the interference signal, and a second filtering is performed to obtain the interference signal before and after the two filterings. The signal-to-noise ratio is calculated based on the adjacent interference peaks. During the second filtering, the minimum frequency value in the frequency domain signal is identified, and the low-pass filter is set according to the minimum frequency value. The sum of the signal-to-noise ratios calculated under different settings before and after the two filtering operations are sorted in descending order, and the filtering combination with the largest sum of signal-to-noise ratios is selected as the target filtering strategy within the cycle.

2. The multi-parameter axial length measurement and data processing system based on the self-mixing vernier effect according to claim 1, characterized in that, The real signal of the interference signal after one filtering is set as follows: Hilbert transform of imaginary signals It is expressed as follows: The analytic signal constructed from real and imaginary signals is represented as follows: analytic signal The modulus represents the envelope of the original signal.

3. The multi-parameter axial length measurement and data processing system based on the self-mixing vernier effect according to claim 1 or 2, characterized in that, The signal strength received by the balanced photodetector I It is expressed as follows: Among them and These represent the center wavelength and full width at half maximum (FWHM) of a low-coherence light source, respectively. Indicates receiving interference signals I Energy; This represents the sum of the DC components of two low-coherence electrical signals. Indicates the amount of change after interference; This indicates the amount of phase change of the reference light; Measuring signal strength I The process changes the internal orientation of the fiber delay line and alters the phase of the reference light. The axial length parameters are inferred by adjusting the size of the posture.

4. The multi-parameter axial length measurement and data processing system based on the self-mixing vernier effect according to claim 1 or 2, characterized in that, The interference signal generated by a highly coherent light source exhibits periodic variations. The peak spacing D of the low-coherence signal is determined by comparing it with the interference peak of the interference signal generated by a low-coherence light source, as shown below: Among them The wavelength of highly coherent laser, For air at wavelength The refractive index below, For the complete number of interference fringes, The phase difference of the incomplete interference fringes. This represents the overall phase difference.

5. The multi-parameter axial length measurement and data processing system based on the self-mixing vernier effect according to claim 4, characterized in that, The high-coherence light source, phase-shifting grating, erbium-doped fiber, wavelength division multiplexer, and fiber delay line constitute a three-mirror cavity model. The phase-shifting grating is equivalent to the first and second reflecting mirrors of the three-mirror cavity model, and the fiber delay line is equivalent to the third reflecting mirror. The adjustment posture of the fiber delay line is determined according to the self-generating hybrid interference formula. As a signal scale for low-coherence interference, it is represented as follows: Represents the group refractive index. Indicates the line width expansion factor. This represents the cavity length of the equivalent resonant cavity of the phase-shifting grating. Coupling efficiency, and These represent optical frequency and critical frequency, respectively. Represents the speed of light. This represents the equivalent external cavity round-trip delay between the phase-shifting grating and the fiber delay line.

6. The multi-parameter axial length measurement and data processing system based on the self-mixing vernier effect according to claim 1, characterized in that, A two-dimensional galvanometer is set in the spatial optical path between the zoom lens and the eyeball. Spatial light is focused onto different areas of the eyeball by the two-dimensional galvanometer and reflects eyeball data from different areas. A beam splitter and a reflector are set in the spatial optical path between the zoom lens and the two-dimensional galvanometer to introduce part of the light beam reflected by the eyeball into the Hartmann microlens optical path. The refractive power of the eyeball is calculated by the Hartmann array diagram acquired by the Hartmann optical path.

7. The multi-parameter axial length measurement and data processing system based on the self-mixing vernier effect according to claim 6, characterized in that, When the wavefront to be measured is received by the microlens array, it is divided into countless sub-wavefronts by the microlens array, and then converged on the CCD to form a light spot array pattern containing the information of the wavefront to be measured; the position offset of the light spot is calculated based on the light spot array pattern, and the diopter and divergence data are calculated. The process of calculating the position offset of the light spot is based on constructing five sets of two-dimensional surfaces using Zernike polynomials. The partial derivatives of the surfaces are used to determine the position offset of each sampling point. The surface polynomials and partial derivatives are expressed as follows: Obtain at least one set of light spots offset Based on the coordinates of the reference point Invert the coefficients of the five polynomials; offset It is expressed as follows: The partial derivatives at the sampling points are expressed as follows: Among them Represents the wavefront function. and Indicates the offset value. and For the corresponding derivative, Let n be the focal length of the lens, and n is 5.

8. The multi-parameter axial length measurement and data processing system based on the self-mixing vernier effect according to claim 7, characterized in that, Based on the partial derivatives at sampling points and the correspondence between matrices, the polynomial coefficient identity is expressed as follows: The Zernike coefficients corresponding to the wavefront phase difference are obtained using the inverse matrix, as follows: Through five coefficients The refractive power and divergence are calculated and expressed as follows: in, It is a coefficient matrix. It is the gradient matrix. It is the wavefront slope matrix; Indicates the diopter value. This represents the divergence value.

Images