TWI461650B - An image processing system and a method thereof - Google Patents

Description

Image processing system and method thereof

The present invention relates to an image processing system and method thereof, and more particularly to an image processing system and method thereof for use in an interference pattern.

In recent years, optical metrology has successfully developed some high-efficiency measurement techniques and is widely used in various deformation measurements, such as bone, IC packaging, or film deformation, among which, Electronic Speckle Pattern Interferometry (Electronic Speckle Pattern Interferometry; Referring to ESPI, it can measure the in-plane and out-of-plane vibration caused by pressure, temperature or stress, the fluctuation of the surface of the object to be tested, and the vibration mode of the object to be tested.

When the laser light illuminates the object to be tested with a rough surface, the uneven light-emitting surface causes the reflected light to interfere with each other due to scattering and form a complex light intensity distribution at the reflection, which is a speckle effect. In addition, when the surface of the object to be tested is displaced, the distribution of the spots also changes. Therefore, according to the interference principle of the speckle, the electron speckle interferometry uses the CCD to capture the image before and after the deformation of the object to be detected, and then interferes with the image superficial image to generate an interference phenomenon, which not only records the amplitude of the reflected light of the object to be tested, but also The phase data of the reflected light and the like are recorded.

Furthermore, electronic speckle interferometry is a global measurement technique, but the measurement results are often caused by noise, such as noise generated by electronic instruments, optical equipment or mechanical disturbances, resulting in reduced image quality, which is generated by electronic instruments. For background noise, in addition, lasers and lenses can reduce the noise generated by optical devices, but mechanical disturbances (generally due to air disturbances) are difficult to remove.

In the prior art, the subtraction method or the averaging method is used to filter out noise to improve image quality. However, the subtraction method can reduce the influence of background noise but it is difficult to eliminate the disturbance noise, and the averaging method can eliminate the disturbance noise but still remains. There is background noise. At present, a measurement technique that can eliminate or reduce background noise and disturbance noise has not been proposed.

To solve the above problems of the prior art, the present invention provides an image processing system and method thereof for reducing background noise and disturbance noise of an image.

The image processing method of the present invention is applied to an interference pattern, and the image processing method comprises the following steps: (1) using a photosensitive element to capture an image sequence of different exposure time for the interference pattern, so as to make a plurality of pixels of the photosensitive element Obtaining the light intensity of the interference pattern; (2) calculating an average value of the light intensity at each of the pixel points based on the exposure time; and (3) averaging the light intensity at each of the pixel points based on the number of sheets of the image sequence The value performs the operation of the standard deviation, and the background noise and the disturbance noise of the interference pattern are eliminated by using the result of the standard deviation. In addition, the step (4) may be performed to perform frequency scanning on the operation result of the standard deviation on each of the pixel points to obtain an image of the resonance frequency of the object to be tested and the interference pattern corresponding to the resonance frequency.

The image processing system of the present invention is applied to an interference pattern of a sample to be tested, the image processing system comprising: a photosensitive element for capturing an image sequence of different exposure time for the interference pattern, so as to have a plurality of pixels Obtaining a light intensity of the interference pattern; and a processor connecting the photosensitive element and obtaining a light intensity of the interference pattern, the processor comprising: a time domain averaging module, configured to calculate each pixel point based on the exposure time The average value of the light intensity; and the noise cancellation module is configured to perform a standard deviation operation on the average value of the light intensity at each pixel point based on the number of images of the image sequence, and use the standard deviation operation result Eliminate background noise and disturbance noise of the interference pattern.

In the image processing system and method of the present invention, the interference pattern may be irradiated by two dimming lights simultaneously to illuminate an object to be tested vibrating in one side, and the spots formed by the two dimming lights interfere with each other; or may be separately dimmed by two The object to be tested is irradiated with a reference surface and an external vibration, and the spots formed by the two dimming light interfere with each other to generate.

In addition, before the noise cancellation module has performed the standard deviation operation, the average light intensity at each pixel point still includes the disturbance noise, and the probability of occurrence of the disturbance noise can be estimated by using a Gaussian distribution function.

Compared with the prior art, the present invention applies the temporal standard deviation (TSTD) algorithm to the electronic speckle interferometry, and can simultaneously take advantage of the subtraction method and the averaging method, and can reduce the air disturbance noise and The effect of background noise on electronic speckle interferometry to obtain high-resolution interference patterns and improve image quality.

The embodiments of the present invention are described in the following specific embodiments, and those skilled in the art can easily understand other advantages and functions of the present invention by the disclosure of the present disclosure, and can also be implemented by other different embodiments. Or application.

FIG. 1A is a schematic diagram showing an optical architecture of a test object in a measurement surface vibration, and FIG. 1B is a schematic diagram showing an in-plane vibration of the object to be tested, which is obtained by the optical architecture shown in FIG. 1A. The interference pattern of the object to be tested shown in Fig. 1B captures an image sequence of different exposure times.

The object to be tested 10, for example, a piezoelectric piece, can be in an in-plane vibration state by the function wave generator 11 and the power amplifier 12, and as shown in FIG. 1B, the in-plane vibration includes displacements in the U direction and the V direction.

The beam emitter 13 emits the same dimming light, such as a 氦-氖 laser, which is split into two beams by the beam splitter 14 which are spatially filtered by the spatial filters 16a and 16b, respectively, and incident at equal angles. In the normal direction of the object 10 to be tested. The object to be tested 10 reflects the two beams of light so that the spots formed by the two beams interfere with each other to generate an interference pattern.

The photosensitive element 17 is placed in the normal direction of the object to be tested 10, and the image sequence of the different exposure time is captured for the interference pattern, so that the plurality of pixels of the photosensitive element 17 obtain the light intensity of the interference pattern, and the light intensity is obtained. The light intensity is transmitted to the processor 18 and the captured image is displayed on the display 19. In the present optical architecture, the processor 18 in the figure is a computer, but it can be replaced by any other electronic device having a computing function, and the image processing of the image by the processor 18 will be described later.

2A is a schematic diagram showing the optical architecture of the object to be tested for vibration outside the surface, and FIG. 2B is a schematic diagram showing the vibration of the surface of the object to be tested, which is obtained by the optical architecture shown in FIG. 2A. The interference pattern of the object to be tested shown in Fig. 2B captures an image sequence of different exposure times.

The out-of-plane vibration of the object to be tested is shown in Figure 2B and is in the W direction. The difference between the 2A and 1A is that after the dimming is reflected by the reflector 15 and the spatial filter 16 cancels the spatial noise, the beam splitter 14' splits the dimming into two beams, and one beam is irradiated to be tested. The object 10 is reflected back to the photosensitive element 17 (object light), and the other light is irradiated on the reference plate 20 as a reference light, and the spots formed by the object light and the reference light interfere with each other to generate an interference pattern. The photosensitive element 17 is placed in the normal direction of the object to be tested 10, and the image sequence of the different exposure time is captured for the interference pattern, so that the plurality of pixels of the photosensitive element 17 obtain the light intensity of the interference pattern, and the light intensity is obtained. The light intensity is transmitted to the processor 18 and the captured image is displayed on the display 19. In the present optical architecture, the processor 18 in the figure is a computer, but it can be replaced by any other electronic device having a computing function, and the image processing of the image by the processor 18 will be described later.

It should be noted that the image processing system and method of the present invention can perform image processing on an interference image of a test object of a vibration mode such as an in-plane vibration, an out-of-plane vibration, and an overall vibration, wherein the in-plane vibration is to be tested. The object vibrates parallel to its horizontal plane, and the out-of-plane vibration causes the object to be tested to vibrate perpendicular to its horizontal plane. The above 1A and 1B are merely illustrative optical architectures, and those of ordinary skill in the art should be able to change the optical architecture for the desired measured vibration modes.

Figure 3 is a block diagram of the image processing system of the present invention. The image processing system of the present invention basically includes a photosensitive element 31 and a processor 32.

The photosensitive element 31 can be, for example, a Charge-coupled Device (CCD) or a Complementary Metal-Oxide-Semiconductor (CMOS), for capturing an image sequence of different exposure time for the interference pattern, so that The plurality of pixels of the photosensitive element 31 acquire the light intensity of the interference pattern. In other words, the photosensitive element 31 takes a picture of the object to be measured at intervals (either fixed or non-fixed time period), wherein the object to be tested may be in-plane vibration, out-of-plane vibration or overall vibration state. Then, the plurality of pixels of the photosensitive element 31 acquire the light intensity of the interference pattern, and then transmit the charge signal of the light intensity obtained by each pixel to the processor 32.

The processor 32 is connected to the photosensitive element 31 and receives a charge signal of the intensity of the interference pattern obtained by each pixel. The processor 32 includes a time domain averaging module 321 and a noise cancellation module 322. The time domain averaging module 321 is configured to calculate an average value of the light intensity at each of the pixel points based on the exposure time. The noise cancellation module 322 is configured to perform a standard deviation operation on the average value of the light intensity at each pixel point based on the number of images of the image sequence. Therefore, the image processing system of the present invention can utilize the standard noise to eliminate the background noise and disturbance noise of the interference pattern to enhance the image quality.

Hereinafter, the image processing method of the present invention will be described with the flowchart shown in FIG. 4 in conjunction with an algorithm.

In step S301, the image sequence of the different exposure time is captured by the interference pattern generated by the photosensitive element to be measured, and the light intensity of the interference pattern is obtained by a plurality of pixels of the photosensitive element.

In step S302, an average light intensity ( I _i ( x , y )) at each of the pixel points is calculated based on the selected exposure time (i τ). In detail, after the interference of the two dimming lights, the light intensity formula of the interference pattern is: If we consider the noise caused by the disturbance, add ψ ( x , y , t ) to the formula, then the light intensity formula of the interference pattern is: Then, the light intensity time average of each pixel is: Where i denotes the ith image in the image sequence, (x, y) denotes a pixel point, J ₀ is a zero-order Bessel function, and Γ is a sensitivity factor. ( x , y ) represents the phase, and A ( x , y ) represents the amplitude. Moreover, the probability of disturbing the noise ψ ( x , y , t ) can be estimated by a Gaussian distribution function. Therefore, the average value of the light intensity time of each pixel is: Where I ₀ ( x , y )+ I _R ( x , y ) represents background noise, and ψ _i sin φ ( x , y ) represents disturbance noise.

In step S303, a standard deviation operation is performed on the average value of the light intensity at each of the pixel points based on the number of images ( n ) of the image sequence. The standard deviation formula of light intensity is: among them, It is a constant that does not contain time, that is, the disturbance noise converges to a constant when a large number of image sequences are used, thereby eliminating background noise and disturbing noise.

In addition, after the background noise and the disturbance noise are eliminated, the light intensity received by the object to be tested is calculated at a specific vibration frequency to perform frequency-sweeping on the object to be tested. It can be expressed by the following algorithms: Where L and W are the length and width of the object to be tested, respectively, and ω is a specific frequency.

In a specific implementation, first, for example, 50 image sequences are respectively taken at each specific frequency, then time domain average and standard deviation operations are performed, and an E ( ω ) operation is performed to establish a frequency scan curve, as shown in FIG. 5, and then Select the corresponding image. Referring to FIG. 5, it shows the normalization of the light intensity received by the object under test at each specific frequency, and shows the fitting curve, and the resonance frequency of the object to be tested is obtained from the turning of the optimum curve. For example, about 852 Hz, and below the resonance frequency, the image quality of the interference pattern of the object to be tested obtained by the photosensitive element is preferable.

Refer to Figures 6A, 6B, and 6C, which are images captured by the image processing method of the averaging method, the subtraction method, and the standard deviation method, respectively. The 6C picture significantly improves the image quality compared to the 6A and 6B pictures, and the background method still has background noise. The calculation formula of the subtraction method is: The result of the subtraction method still has disturbing noise | ψ _j |-| ψ _i |.

In summary, the image processing system and method thereof of the present invention mainly perform time averaging operation and standard deviation operation, that is, time domain standard deviation algorithm, on image sequences of different exposure times of interference patterns, to eliminate background noise and Air disturbances, which in turn improve image quality. In addition, the present invention can be applied to the design and development of electrodes, piezoelectric transducers, and high-frequency vibration elements of piezoelectric sheets in addition to measuring minute deformation of piezoelectric sheets.

The above-described embodiments are merely illustrative of the principles, features, and effects of the present invention, and are not intended to limit the scope of the present invention. Any person skilled in the art can recite the above without departing from the spirit and scope of the present invention. The embodiment is modified and changed. Any equivalent changes and modifications made by the disclosure of the present invention should still be covered by the following claims. Therefore, the scope of protection of the present invention should be as set forth in the scope of the claims described below.

10. . . Analyte

11. . . Function wave generator

12. . . Power amplifier

13. . . Beam emitter

14, 14’. . . Splitter

15. . . reflector

16, 16a, 16b. . . Spatial filter

17. . . Photosensitive element

18. . . processor

19. . . monitor

20. . . Reference board

31. . . Photosensitive element

32. . . processor

321. . . Time domain average module

322. . . Noise cancellation module

U, V, W. . . direction

S301~S303. . . step

Figure 1A is a schematic diagram showing the optical architecture of the object under test in the measurement plane; Figure 1B is a schematic diagram showing the in-plane vibration of the object to be tested: Figure 2A is a schematic diagram showing the measurement of the out-of-plane vibration FIG. 2B is a schematic diagram showing the out-of-plane vibration of the object to be tested; FIG. 3 is a block diagram of the image processing system of the present invention; FIG. 4 is an image of the present invention; A flowchart of a processing method; FIG. 5 is an explanatory diagram of an execution frequency scan of the image processing method of the present invention;

Figures 6A to 6C are images captured by an image processing method using an average algorithm, a subtraction algorithm, and a standard difference algorithm, respectively.

S301~S303. . . step

Claims (10)

An image processing method is applied to an interference pattern, the image processing method comprising the following steps: (1) using a photosensitive element to capture an image sequence of different exposure time for the interference pattern, so that the plurality of pixels of the photosensitive element acquire the image Interfering with the light intensity of the pattern; (2) calculating an average value of the light intensity at each pixel point based on the exposure time; and (3) averaging the light intensity at each of the pixel points based on the number of images of the image sequence The standard deviation operation is performed, and the background noise and the disturbance noise of the interference pattern are eliminated by using the operation result of the standard deviation. The image processing method according to claim 1, wherein before the step (1), the method further comprises: dimming the same object to be irradiated, so that the spots formed by the two dimming light interfere with each other to generate the The step of interfering with the pattern. The image processing method according to claim 2, wherein the object to be tested is in an in-plane vibration state. The image processing method of claim 1, wherein before the step (1), the method further comprises: dimming the two dimming lights to respectively irradiate a reference object and a test object to make the spot formed by the dimming The step of interfering with each other to produce the interference pattern. The image processing method according to claim 4, wherein the object to be tested is in an out-of-plane vibration state. The image processing method according to claim 1, wherein the object to be tested is in an overall vibration state. The image processing method according to claim 1, wherein in step (2), the average value of the light intensity at each of the pixels includes the disturbance noise. The image processing method according to claim 1, wherein in step (2), the probability of occurrence of the disturbance noise is estimated by using a probability distribution function. The image processing method of claim 1, further comprising the step (4) of performing a frequency sweep on the operation result of the standard deviation at each pixel to obtain a resonance frequency of the object to be tested and correspondingly An image of the interference pattern of the resonant frequency. An image processing system is applied to an interference pattern of a sample to be tested, the image processing system comprising: a photosensitive element for capturing an image sequence of different exposure time for the interference pattern, so as to obtain a plurality of pixel points a light intensity of the interference pattern; and a processor connecting the photosensitive element to obtain a light intensity of the interference pattern, the processor comprising: a time domain averaging module, configured to calculate each pixel point based on the exposure time An average value of the light intensity; and a noise cancellation module for performing a standard deviation operation on the average value of the light intensity at each pixel point based on the number of images of the image sequence, and eliminating the result by using the standard deviation Interference noise background noise and disturbance noise.