A system and method for effective region segmentation of four-beam lateral shearing interferograms based on FFT and global thresholding

Abstract

The application discloses a four-wave transverse shearing interferogram effective area segmentation system and method based on FFT and global threshold, and belongs to the technical field of optical interference measurement and wavefront sensing. The method comprises an image acquisition module, a preprocessing module, an FFT low-frequency extraction module, an automatic threshold segmentation module and a morphological optimization module. The application carries out smooth processing on the background brightness distribution of the interferogram through FFT low-pass filtering, and realizes automatic segmentation of the effective area in combination with Otsu global threshold, effectively avoids technical defects such as center area inversion blackening, edge fracture and excessive noise that are prone to occur in traditional area segmentation methods, can completely extract the four-wave overlapping area, has the advantages of stable segmentation, strong anti-interference capability, no need for manual parameter adjustment, good real-time performance and the like, and can be widely applied to various four-wave transverse shearing interferometer wavefront detection systems.

Description

Technical Field

[0001] This invention relates to the field of optical interferometry and wavefront sensing technology, specifically to a system and method for effective region segmentation of four-wave transverse shearing interferograms based on FFT and global threshold. Background Technology

[0002] The four-wave transverse shearing interferometer, with its outstanding advantages such as common-path structure, no need for reference light, and strong vibration resistance, has been widely used in many fields such as transient wavefront measurement, optical component inspection, and laser beam quality analysis. Accurate segmentation of the effective interference region is a crucial step in achieving wavefront slope demodulation, wavefront reconstruction, and aberration analysis; its segmentation accuracy directly affects the overall performance of subsequent wavefront detection.

[0003] Currently, there are many methods for segmenting single-frame interferogram regions, but most of them are designed based on the idea of ​​stripe image processing. When applied to the segmentation of four-wave transverse sheared interferograms, they have obvious defects: Gaussian background subtraction relies on large-scale filtering, which easily leads to blurred edges of the effective region, incomplete segmentation of the central bright area, and weak noise suppression ability; local thresholding methods such as Niblack and Sauvola rely on the local gray-level mean and variance for threshold determination, which is sensitive to the polarity of the central bright area and is prone to problems such as effective region reversal, central region being completely black, and edge breakage, and cannot correctly extract the four-wave overlapping area; morphological top-hat transformation can only enhance fine stripe features and cannot retain large areas of central bright area, which easily leads to the loss of the four-wave effective region; traditional global thresholding methods do not eliminate the influence of background brightness gradient, have poor threshold stability, and are easily affected by factors such as illumination changes and noise interference, resulting in unstable segmentation results.

[0004] Compared to typical single-frame interference fringes, the fringes generated by a four-wave transverse shearing interferometer are formed by the superposition of shearing interference fringes from four beams of light. Their distribution pattern is more complex, and traditional segmentation methods struggle to achieve stable, complete, and accurate segmentation of the effective region of the four-wave transverse shearing interferogram, thus affecting the accuracy of subsequent wavefront detection.

[0005] To address the shortcomings of existing technologies, this invention proposes a segmentation method based on FFT low-frequency extraction and automatic global thresholding. This method achieves stable, complete, and polarity-inverted segmentation of the effective region from a technical structural perspective, thus overcoming the technical bottlenecks of existing methods. Summary of the Invention

[0006] To address the technical problems of existing segmentation methods, such as polarity reversal, loss of central region, large noise interference, and incomplete segmentation, this invention provides an effective region segmentation system and method for four-wave transverse shearing interferograms based on FFT and global threshold, which meets the engineering application requirements of practical wavefront detection.

[0007] A four-wave transverse shearing interferogram effective region segmentation system based on FFT and global thresholding includes an image acquisition module, a preprocessing module, an FFT low-frequency extraction module, an automatic thresholding segmentation module, and a morphological optimization module. The FFT low-frequency extraction module performs a two-dimensional FFT transformation on the interferogram, filters out high-frequency components in the image using a low-pass mask, and outputs a smooth background brightness image. The automatic thresholding segmentation module uses the Otsu global thresholding algorithm to binarize the smooth background brightness image to obtain an effective region mask. The morphological optimization module performs noise reduction, hole filling, and edge smoothing on the effective region mask.

[0008] A method for effective region segmentation of four-wave transverse shearing interferograms based on FFT and global thresholding includes the following steps:

[0009] Step 1: The image acquisition module acquires four-wave transverse shearing interference fringe patterns using image acquisition devices such as a CCD camera and transmits them to the preprocessing module.

[0010] Step 2: The preprocessing module performs preliminary noise reduction on the original interference fringe pattern to remove random noise from the image and provide high-quality image data for subsequent FFT processing.

[0011] Step 3: The FFT low-frequency extraction module performs a two-dimensional FFT transformation on the preprocessed interference fringe pattern to generate the image spectrum. The high-frequency fringe components and noise are filtered out by the central low-pass mask, and only the low-frequency brightness contour is retained.

[0012] Step 4: Perform an inverse FFT transform on the filtered low-frequency spectrum to obtain a smooth background brightness map;

[0013] Step 5: The automatic threshold segmentation module uses the Otsu algorithm to perform global threshold binarization on the smooth background brightness map to obtain the initial effective region mask;

[0014] Step 6: The morphological optimization module sequentially performs small-area noise removal, hole filling, and closing operations on the initial effective region mask to smooth the region edges and finally output the effective region, completing the segmentation process.

[0015] The core improvement of this invention lies in the fact that by extracting the low-frequency brightness contour of the interferogram through FFT, the influence of the background brightness gradient is eliminated, and the global threshold can stably segment the effective area of ​​the center highlight. This completely avoids the problems of polarity reversal and center area blackening that occur in traditional methods, and ensures the complete extraction of the four-wave overlap area.

[0016] The present invention has the following beneficial effects:

[0017] 1. No polarity reversal and intact central region: Compared with the traditional local thresholding method, this method does not have the problem of polarity reversal where the central bright area is misjudged as black, and can completely preserve the central bright area formed by the overlap of four waves;

[0018] 2. Strong anti-interference capability: FFT low-pass filtering effectively suppresses noise interference, while the global threshold segmentation method is not affected by local brightness changes, and the segmentation effect is minimally affected by external factors such as lighting and noise;

[0019] 3. Fully automatic without manual parameter adjustment: The entire segmentation process does not require manual setting of thresholds and other parameters, and can adapt to four-wave transverse shearing interferograms with different brightness and contrast, with strong adaptability;

[0020] 4. Simple structure and easy to implement in engineering: The method has a simple process and moderate computational load. It can be embedded in FPGA, DSP or embedded system, and can meet the engineering application requirements of real-time wavefront detection.

[0021] 5. The FFT low-frequency extraction, Otsu global thresholding, and morphological optimization employed in this invention are general techniques in the field of interferogram processing. However, existing methods have not addressed the unique challenges, such as the difficulty in extracting the central four-wave overlap region of a four-wave transverse shearing interferogram. This invention, by specifically combining FFT background extraction with global thresholding, achieves stable, complete, and inversion-free segmentation of the effective region of a four-wave transverse shearing interferogram, demonstrating outstanding technical innovation and specialization. Attached Figure Description

[0022] Figure 1 To provide an overall structural diagram for the invention;

[0023] Figure 2 This is the original four-wave transverse shearing interferogram;

[0024] Figure 3 The spectrum of four-wave transverse shearing interference fringes;

[0025] Figure 4 This is a brightness map of the interference fringes background.

[0026] Figure 5 The background light intensity segmented by the method of the present invention;

[0027] Figure 6 The four-wave shearing interference fringes are segmented by the method of the present invention. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other. To achieve the above objectives, this invention adopts the following technical solution.

[0029] This invention relates to a four-wave transverse shearing interferogram effective region segmentation system based on FFT and global thresholding, comprising an image acquisition module, a preprocessing module, an FFT low-frequency extraction module, an automatic thresholding segmentation module, and a morphological optimization module. The FFT low-frequency extraction module performs a two-dimensional FFT transformation on the interferogram, filters out high-frequency components in the image using a low-pass mask, and outputs a smooth background brightness image. The automatic thresholding segmentation module uses the Otsu global thresholding algorithm to binarize the smooth background brightness image to obtain an effective region mask. The morphological optimization module performs noise reduction, hole filling, and edge smoothing on the effective region mask.

[0030] The FFT low-frequency extraction module is implemented through a center low-pass filter, which retains only the low-frequency brightness contours of the image, effectively removes high-frequency stripe components and noise, and eliminates the influence of background brightness gradient.

[0031] The automatic threshold segmentation module does not rely on local statistical information of the image, does not produce effective region polarity reversal, and can completely preserve the central four-wave overlapping region.

[0032] The specific processing steps of the morphological optimization module include: small-area noise removal, hole filling within the region, and edge smoothing using closing operations, further improving the integrity and edge smoothness of the effective region.

[0033] The method is a fully automated processing structure that does not require manual setting of thresholds and other related parameters, and can adapt to four-wave transverse shearing interferograms with different brightness and contrast.

[0034] The method can be integrated into various four-wave transverse shearing interferometry wavefront detection devices, such as wavefront sensors, beam quality analyzers, and optical surface inspection systems.

[0035] In this invention, the expression for the wavefront of the light being measured is defined as follows:

[0036] (1)

[0037] in, The axes are y and y, where A represents the amplitude distribution of the measured light and W represents the phase distribution of the measured light. For wave number, λ is the wavelength.

[0038] The four-wave transverse shearing interference is formed by the coherent superposition of four misaligned shearing wavefronts generated by grating diffraction of the measured wavefront. The expressions for the four wavefronts are as follows:

[0039] (2)

[0040] (3)

[0041] (4)

[0042] (5)

[0043] s represents the shearing amount between each pair of the four diffracted beams in the horizontal and vertical directions.

[0044] The total light field is the superposition of four beams of light, that is: The expression for the total light field intensity is:

[0045] (6)

[0046] in Represents background light intensity.

[0047] (7)

[0048] In equation (6), Represents the grating period. Let be the distance between the grating and the CCD camera. From equation (1), it can be seen that the four-wave transverse shearing interference fringes are determined by the background light intensity. The background light intensity is formed by the superposition of interference fringes formed by the pairwise coherence of the four first-order diffracted beams. Furthermore, the greater the overlap of the four beams (i.e., the four-wave shear region), the higher the background light intensity. The brighter the light, the more effective the region segmentation in this invention.

[0049] like Figure 2 As shown, this is the original four-wave transverse shearing interferogram. A two-dimensional Fourier transform of the interferometric fringe image yields the following spectral distribution: Figure 3 As shown, background light intensity is extracted through low-pass filtering. By analyzing the corresponding spectral components and performing an inverse Fourier transform on them, a smooth background intensity map can be obtained. ,like Figure 4 As shown. The background light intensity map is segmented using the segmentation method of this invention. Effective region segmentation is performed, and the final background light intensity segmentation result is as follows: Figure 5 As shown, the interference fringe segmentation results are as follows: Figure 6 As shown, this method can completely and accurately extract the effective region of four overlapping waves.

[0050] The image acquisition module, preprocessing module, FFT low-frequency extraction module, automatic threshold segmentation module, and morphological optimization module work together in sequence to form a complete and effective region segmentation system, ensuring the stability and reliability of the segmentation process from the hardware structure level.

[0051] The specific workflow of this invention, which provides a four-wave transverse shearing interferogram effective region segmentation method based on FFT and global threshold, is as follows: Figure 1 As shown:

[0052] Step 1: The image acquisition module acquires four-wave transverse shearing interference fringe patterns using image acquisition devices such as CCD cameras, and transmits them to the preprocessing module (the interference fringe expression is shown in equation (6)).

[0053] Step 2: The preprocessing module performs preliminary noise reduction on the original interference fringe pattern to remove random noise from the image and provide high-quality image data for subsequent FFT processing.

[0054] Step 3: The FFT low-frequency extraction module performs a two-dimensional FFT transformation on the preprocessed interference fringe pattern to generate the image spectrum as shown in equation (8):

[0055] (8)

[0056] in, Represents Fourier transform, , , , These are the spectra of the interference fringes along the x-direction, y-direction, x+y-direction (+45° direction), and xy-direction (-45° direction), respectively. , These are the spectral coordinates in the horizontal and vertical directions, respectively.

[0057] ;

[0058] ;

[0059] ;

[0060] ;

[0061] , , , They are respectively , , , The negative exponential function;

[0062] High-frequency stripe components and noise are filtered out using a central low-pass mask, retaining only the low-frequency brightness profile, such as... Figure 3 As shown; ;

[0063] Step 4: Perform an inverse FFT transform on the filtered low-frequency spectrum to obtain a smooth background brightness map;

[0064] ;

[0065] in, This represents the inverse Fourier transform.

[0066] Step 5: The automatic thresholding module uses the Otsu algorithm to perform global thresholding binarization on the smooth background brightness map to obtain the initial effective region mask; the Otsu algorithm is the image histogram. Figure 2 Value-based region segmentation utilizes the bimodal distribution characteristics of the image grayscale histogram and automatically selects the optimal threshold based on maximizing the inter-class variance, enabling high-precision and robust region segmentation of smooth background images.

[0067] Step 6: The morphological optimization module performs small-area noise removal, hole filling, and closing operations on the initial effective region mask in sequence to smooth the region edges and finally output a complete and accurate effective region, thus completing the segmentation process.

Claims

1. A four-wave transverse shearing interferogram effective region segmentation system based on FFT and global threshold, characterized in that, The system includes an image acquisition module, a preprocessing module, an FFT low-frequency extraction module, an automatic thresholding segmentation module, and a morphological optimization module. The FFT low-frequency extraction module performs a two-dimensional FFT transformation on the interferogram, filters out high-frequency components in the image using a low-pass mask, and outputs a smooth background brightness image. The automatic thresholding segmentation module uses the Otsu global thresholding algorithm to binarize the smooth background brightness image to obtain an effective region mask. The morphological optimization module performs noise reduction, hole filling, and edge smoothing on the effective region mask.

2. The effective region segmentation system for four-wave transverse shearing interferograms based on FFT and global thresholding according to claim 1, characterized in that, The FFT low-frequency extraction module is implemented through a center low-pass filter, which retains only the low-frequency brightness contours of the image, effectively removes high-frequency stripe components and noise, and eliminates the influence of background brightness gradient.

3. The effective region segmentation system for four-wave transverse shearing interferograms based on FFT and global thresholding according to claim 1, characterized in that, The specific processing steps of the morphological optimization module include: small-area noise removal, hole filling within the region, and edge smoothing using closing operations, further improving the integrity and edge smoothness of the effective region.

4. A method for effective region segmentation of four-wave transverse shearing interferograms based on FFT and global thresholding, applied to the system described in any one of claims 1-3, characterized in that, Includes the following steps: Step 1: The image acquisition module acquires four-wave transverse shearing interference fringe patterns using image acquisition devices such as a CCD camera and transmits them to the preprocessing module. Step 2: The preprocessing module performs preliminary noise reduction on the original interference fringe pattern to remove random noise from the image and provide high-quality image data for subsequent FFT processing. Step 3: The FFT low-frequency extraction module performs a two-dimensional FFT transformation on the preprocessed interference fringe pattern to generate the image spectrum. The high-frequency fringe components and noise are filtered out by the central low-pass mask, and only the low-frequency brightness contour is retained. Step 4: Perform an inverse FFT transform on the filtered low-frequency spectrum to obtain a smooth background brightness map; Step 5: The automatic threshold segmentation module uses the Otsu algorithm to perform global threshold binarization on the smooth background brightness map to obtain the initial effective region mask; Step 6: The morphological optimization module sequentially performs small-area noise removal, hole filling, and closing operations on the initial effective region mask to smooth the region edges and finally output the effective region, completing the segmentation process.

5. The method for effective region segmentation of four-wave transverse shearing interferograms based on FFT and global threshold as described in claim 4, characterized in that, The four-wave transverse shearing interferogram is specifically as follows: Let the expression for the wavefront of the measured light be: （1） in, The axes are y and y, where A represents the amplitude distribution of the measured light and W represents the phase distribution of the measured light. For wave number, Wavelength; The four-wave transverse shearing interference is formed by the coherent superposition of four misaligned shearing wavefronts generated by grating diffraction of the measured wavefront. The expressions for the four wavefronts are as follows: （2） （3） （4） （5） s represents the shearing amount between each pair of the four diffracted beams in the horizontal and vertical directions; the total optical field is the superposition of the four beams, i.e.: The expression for the total light field intensity is: （6） in, Represents background light intensity; （7） In equation (6), Represents the grating period. This is the distance between the grating and the CCD camera.

6. The method for effective region segmentation of four-wave transverse shearing interferograms based on FFT and global thresholding according to claim 4, characterized in that, The generated image spectrum in step 3 is shown in equation (8): （8） in, Represents Fourier transform, , , , These are the spectra of the interference fringes along the x-direction, y-direction, x+y-direction (+45° direction), and xy-direction (-45° direction), respectively. , These are the spectral coordinates in the horizontal and vertical directions, respectively; ； ； ； ； , , , They are respectively , , , It is a negative exponential function.

7. The method for effective region segmentation of four-wave transverse shearing interferograms based on FFT and global thresholding according to claim 6, characterized in that, Further filtering of high-frequency stripe components and noise using a center low-pass mask, retaining only the low-frequency brightness profile: .

8. The method for effective region segmentation of four-wave transverse shearing interferograms based on FFT and global threshold as described in claim 7, characterized in that, In step 4, an inverse FFT transform is performed on the filtered low-frequency spectrum to obtain a smoothed background brightness map; the formula is: ； in, This represents the inverse Fourier transform.

9. The method for effective region segmentation of a four-wave transverse shearing interferogram based on FFT and global threshold as described in claim 4, characterized in that, The Otsu algorithm is an image histogram binarization region segmentation method. It utilizes the bimodal distribution characteristics of the image grayscale histogram and automatically selects the optimal threshold based on maximizing the inter-class variance to achieve region segmentation in smooth background images.

10. The method for effective region segmentation of a four-wave transverse shearing interferogram based on FFT and global threshold as described in claim 4, characterized in that, The method is integrated into various four-wave transverse shearing interferometry wavefront detection devices, including wavefront sensors, beam quality analyzers, and optical surface inspection systems.

Images