A multi-band extraction system
By using a consumer-grade CMOS sensor and a semi-reflective mirror to replace the CCD and beam splitter, the high cost and low adaptability of 3CCD cameras in deformation monitoring are solved, achieving low cost, high precision and high efficiency in deformation monitoring.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Utility models(China)
- Current Assignee / Owner
- KUNMING UNIV OF SCI & TECH
- Filing Date
- 2025-10-17
- Publication Date
- 2026-07-21
AI Technical Summary
Existing 3CCD cameras suffer from high manufacturing costs, low integration, high power consumption, complex structure, poor adaptability, and low data accuracy in deformation monitoring, resulting in large errors in deformation monitoring results.
A consumer-grade CMOS sensor is used to replace the professional CCD, a semi-reflective mirror is used to replace the beam splitter prism, and a bandpass filter is covered on the CMOS sensor. The beam is split by the semi-reflective mirror and the bandpass filter is used to improve the monochromaticity and stability of the beam.
It reduces deformation monitoring costs, improves system integration and layout adaptability, and ensures the accuracy of deformation monitoring and analysis results.
Smart Images
Figure CN224535032U_ABST
Abstract
Description
Technical Field
[0001] This utility model belongs to the field of optical imaging technology and relates to a multi-band extraction system. Background Technology
[0002] Deformation monitoring is the process of continuously or periodically detecting and observing deformable bodies using specialized instruments and methods to understand their deformation state. The data obtained from continuous or periodic detection of the deformation state of deformable bodies can serve as a basis for predicting relevant geological changes and the actual usage of easily deformable components such as load-bearing and structural parts. It can also serve as a basis for judging whether easily deformable components such as load-bearing and structural parts need to be replaced or maintained, which is of great significance.
[0003] The more accurate the baseline data acquired for deformation monitoring, the lower the error in subsequent analysis results, leading to more precise analysis of the deformation state of the deformed body. Because 3CCD technology has excellent color reproduction capabilities, it is currently commonly used for RGB band-dependent deformation monitoring. A 3CCD camera employs three charge-coupled devices (CCDs) to capture red, green, and blue primary colors of light respectively. Its working principle involves separating the incident light into red, green, and blue beams using a coated beam-splitting prism. These three beams are projected onto the three CCDs, resulting in images of the three primary colors. By analyzing the images of the three primary colors before and after deformation using appropriate software, the deformation status of the deformed body can be obtained. However, deformation monitoring using 3CCD cameras has several drawbacks: (1) The main function of CCD is to optimize the efficiency and integrity of charge transfer, which makes its manufacturing process quite special and requires customized production. It is not compatible with the mainstream digital logic chip manufacturing process. Therefore, the production of CCD requires investment in the establishment of a special production line and special maintenance and operation of the production line, resulting in a high manufacturing cost of CCD. (2) CCD chips essentially only have charge transport function, which requires a large number of expensive external chips for support. This increases the material cost of CCD chip manufacturing process, increases the complexity of circuit board design and packaging, and makes it difficult to achieve a high degree of integration. (3) The CCD consumes a lot of power during operation, which means it must be equipped with heat dissipation components (such as heat sinks or fans), which further increases its cost of use and its structural complexity, resulting in a large size and weight, which is not conducive to its practical use. (4) CCDs require more complex and expensive readouts and interfaces.
[0004] (5) As a relatively complex optical path system, the beam splitter works based on the principle of total internal reflection. Its function is highly dependent on the accuracy of its geometric structure. The beam splitter with high structural accuracy does not have an advantage in terms of manufacturing difficulty or cost control.
[0005] (6) The strong binding relationship between the function and structure of the beam splitter results in a large structural limitation on the shape of the beam splitter, which leads to its poor adaptability to specific spatial layout.
[0006] CMOS sensors (Complementary Metal-Oxide-Semiconductor) are core components of digital imaging technology and are widely used in various electronic products. They can be manufactured using the same or similar manufacturing processes as mainstream large-scale integrated circuits (such as CPUs and memory chips). Furthermore, their manufacturing process allows for the integration of photodiodes, amplifiers, analog-to-digital converters (ADCs), and image processing circuits onto a single chip. Therefore, compared to CCDs, CMOS sensors have advantages such as lower overall cost, simpler structure, higher integration, and lower power consumption. However, due to significant information loss in the demosaicing algorithm, CMOS sensors suffer from lower accuracy in acquiring basic data during deformation monitoring. Ultimately, this prevents them from achieving the monitoring accuracy of 3CCD technology, resulting in larger errors in the analysis results and significant deviations in the prediction or evaluation of deformation conditions.
[0007] Therefore, it is necessary to provide a multi-band extraction system that can reduce the overall cost of deformation monitoring, improve system integration and layout adaptability, and at the same time achieve accurate detection of deformation of deformable bodies, thereby providing an excellent foundation for the prediction or evaluation of deformation of deformable bodies. Utility Model Content
[0008] This invention reduces the overall cost of deformation monitoring by using a consumer-grade CMOS sensor instead of a professional CCD and a semi-reflective mirror instead of a precision beam splitter. At the same time, by covering the light receiving port of the CMOS sensor with a bandpass filter, a filtering effect is achieved, which makes the light beam received by the CMOS sensor have higher monochromaticity and stability, thereby ensuring the accuracy of subsequent image analysis.
[0009] To achieve the above objectives, this utility model is implemented through the following technical solution: This invention proposes a multi-band extraction system comprising a light source 1, a semi-reflective mirror 2, a CMOS sensor 3, and a bandpass filter 4. There are n+1 CMOS sensors 3 and n semi-reflective mirrors 2. Different colored coatings 6 are applied to different parts of the object under test 5. The light source 1 emits a light beam towards the object under test 5. The coatings 6 are excited by the light beam and reflect different colored beams. All different colored beams enter the first semi-reflective mirror 2, which splits the incoming beams into transmitted and reflected beams. The reflected beams enter the first CMOS sensor 3, and the transmitted beams sequentially enter the second to nth semi-reflective mirrors. 2. Each semi-reflective mirror 2 splits the light beam into a transmitted beam and a reflected beam. The reflected beam obtained by each semi-reflective mirror 2 enters the CMOS sensor 3 corresponding to the sequence of the semi-reflective mirror 2, until the transmitted beam enters the nth semi-reflective mirror. The transmitted beam obtained by the nth semi-reflective mirror enters the (n+1)th CMOS sensor 3, and the reflected beam obtained by the nth semi-reflective mirror enters the nth CMOS sensor 3. There are n+1 bandpass filters 4. One bandpass filter 4 allows one monochromatic light to pass through, and different bandpass filters 4 allow different colors of monochromatic light to pass through. The bandpass filters 4 cover the light receiving port of the CMOS sensor 3.
[0010] Preferably, both the incident surface and the exit surface of the semi-reflective lens 2 are planar.
[0011] Preferably, the acute angle between the incident surface of the semi-reflective lens 2 and the incident ray is 45°.
[0012] When there is a large height difference between different parts of the object to be measured, the object can be rotated to reduce the height difference between the different parts of the object to be measured, thereby reducing the optical path difference between the light reflected from the different parts of the object to be measured and the first half-reflective lens 2.
[0013] The coating 6 is formed by coating with fluorescent dyes or ordinary dyes.
[0014] When coating 6 is formed by coating with fluorescent dye, it is irradiated with a long-wave ultraviolet light source; when coating 6 is formed by coating with ordinary dye, it is irradiated with a normal white light source.
[0015] The beneficial effects of this utility model are: 1. This utility model significantly reduces the total system cost by using a consumer-grade CMOS sensor instead of a professional CCD, while maintaining the same specifications and quantity. Furthermore, by covering the light receiving port of the CMOS sensor with a bandpass filter, the system's imaging accuracy is improved, providing a more accurate basis for subsequent deformation analysis of deformable bodies, thereby enhancing the system's accuracy in monitoring deformation.
[0016] 2. This invention optimizes the spatial orientation of the semi-reflective mirror, giving the optical path better flexibility, which is beneficial for expanding the applicable scenarios of the system and improving the imaging accuracy of the system.
[0017] 3. By using a semi-reflective mirror instead of a beam splitter prism, this utility model reduces the limiting effect of its specific structure on its function, thereby improving the system's adaptability to different application spaces by using a wider variety of semi-reflective mirrors in terms of type and shape.
[0018] 4. This invention can achieve the same experiment and directly obtain the deformation analysis results of different test parts of the object, which is beneficial to improving experimental efficiency. Attached Figure Description
[0019] Figure 1 This is a schematic diagram of the overall structure of this utility model; Figure 2 This is a schematic diagram illustrating the specific working process of an embodiment of this utility model; Figure 3 This is a schematic diagram illustrating the DIC optical measurement principle of the measured part of the object under test according to this utility model. In the diagram, 1-light source, 2-half-reflective lens, 3-CMOS sensor, 4-bandpass filter, 5-object under test, 6-coating. Detailed Implementation
[0020] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments, but the scope of protection of the present invention is not limited to the content described.
[0021] like Figure 1-2As shown, the multi-band extraction system includes a light source 1, a semi-reflective mirror 2, a CMOS sensor 3, and a bandpass filter 4. There are n+1 CMOS sensors and n semi-reflective mirrors 2. Different colored coatings 6 are applied to different parts of the object under test 5. The light source 1 emits a light beam towards the object under test 5. The coatings 6 are excited by the light beam and reflect different colored beams. All different colored beams enter the first semi-reflective mirror 2, which splits the incoming beam into a transmitted beam and a reflected beam. The reflected beam enters the first CMOS sensor 3, and the transmitted beam sequentially enters the second to nth semi-reflective mirrors 2. Each semi-reflective mirror 2 splits the light beam into a transmitted beam and a reflected beam. The reflected beam obtained from the split beam of each semi-reflective mirror 2 enters the CMOS sensor 3 corresponding to the sequence of the semi-reflective mirror 2, until the transmitted beam enters the nth semi-reflective mirror. The transmitted beam obtained from the split beam of the nth semi-reflective mirror enters the (n+1)th CMOS sensor 3, and the reflected beam obtained from the split beam of the nth semi-reflective mirror enters the nth CMOS sensor 3. There are n+1 bandpass filters 4. One bandpass filter 4 allows one monochromatic light to pass through, and different bandpass filters 4 allow different colors of monochromatic light to pass through. The bandpass filters 4 cover the light receiving port of the CMOS sensor 3.
[0022] like Figure 2 As shown, this invention takes predictive deformation monitoring as an example. It uses a transparent test object to simulate the object under test, monitors the deformation of the transparent test object, and studies the mechanism of mechanical property changes based on the deformation monitoring results. This allows for the acquisition of the evolution mechanism of the deformation field of the transparent test object, and then the prediction of the actual performance change process of the object under test through this evolution mechanism. Since this experiment is predictive, a transparent test object needs to be fabricated to simulate the object under test. A transparent cuboid test object is fabricated using 3D printing technology to simulate the object under test. In this embodiment, the part of the transparent test object to be tested is shown in the figure. Figure 2 As shown, the three test areas are located on the left and right sides of the transparent test object and inside it, respectively, and are generally distributed from top to bottom. Fluorescent layers that can be excited by a light beam to produce red, green, and blue light are respectively set on the three test areas. The test area on the left side has a fluorescent layer that can be excited by a light beam to produce red light, the test area inside has a fluorescent layer that can be excited by a light beam to produce blue light, and the test area on the right side has a fluorescent layer that can be excited by a light beam to produce green light. Since the test areas on the left and right sides are located on the surface of the transparent test object, the fluorescent layers at the aforementioned two test areas can be attached to the transparent test object by coating or gluing after 3D printing. The fluorescent layer at the test area inside the transparent test object is directly embedded inside the transparent test object during the 3D printing process.
[0023] After the experiment begins, the light source is first turned on to illuminate the transparent test object. The light source emits a beam of light into the transparent test object. After the beam reaches the transparent test object, it comes into contact with the fluorescent layer at different test locations. The fluorescent layer is excited and reflects different colors of light. In this experiment, the beam comes into contact with the fluorescent layer on the left, exciting a red beam, so the test location on the left reflects red light. Similarly, the fluorescent layer inside the transparent test object is excited by the beam to produce a blue beam, and the test location inside reflects blue light. The fluorescent layer on the right is excited by the beam to produce a green beam. All three colors of light are reflected to the first semi-reflective lens. After entering the first semi-reflective lens, the three colors of light are evenly divided into transmitted and reflected beams. The reflected beam enters the first CMOS sensor. Because the light receiver of the first CMOS sensor is covered with a bandpass filter that allows green light to pass through, although all three colors of light have reflected beams, only the green light is reflected. The light beam can enter the first CMOS sensor. The transmitted beam of the three colors of light continues to travel and enters the second semi-reflective lens. Similarly, the second semi-reflective lens splits the three colors of light into transmitted beams and reflected beams. The light receiving port of the second CMOS sensor is covered with a bandpass filter that allows blue light to pass through. Only the reflected beam of blue light can pass through the bandpass filter and enter the second CMOS sensor. This experiment only monitors the deformation of three parts of the transparent test object. Therefore, only three fluorescent layers are set up. At the same time, only three CMOS sensors are needed to receive monochromatic light. Two semi-reflective lenses are needed to split the light. The transmitted beam obtained by the second semi-reflective lens continues to travel and enters the third CMOS sensor. The light receiving port of the third CMOS sensor is covered with a bandpass filter that allows red light to pass through. The red light passes through the bandpass filter and enters the third CMOS sensor. The bandpass filter can be covered at the light receiving port of the CMOS sensor by means of bonding, etc. At this point, the transparent test object simulates the undeformed state of the object under test and serves as a reference object. The three CMOS sensors transmit the received light signals to the computer. After receiving the signals, the computer uses software to create an image, which serves as the reference image. Then, a force can be applied to the transparent test object to cause it to deform. During this process, the system continues to receive signals in the same manner as described above and then transmits the signals to the computer. After the experiment is completed, the computer will generate several images of the transparent test object in different deformation states (deformed images). Furthermore, because the signals transmitted by different CMOS sensors are of different colors, the images generated by the different signals transmitted by different CMOS sensors will also be different. By comparing and analyzing the images of the transparent test object in different deformation states with the reference image using relevant software, the analysis results can be obtained, such as the degree of deformation and the direction of deformation.
[0024] If the object to be tested is opaque or the object simulating the object to be tested must be opaque, then applying the coating to the surface of the object to be tested or the object being tested will cause it to emit light of the corresponding color.
[0025] In this embodiment, DIC technology is used to analyze speckle images of a transparent object before and after deformation at different locations under force, received by three CMOS sensors. DIC is a non-contact optical measurement technology used to obtain full-field, high-precision deformation and even strain of an object's surface by analyzing images. Its core idea is to track the movement of the speckle pattern on the object's surface before and after deformation, thereby calculating displacement and internal deformation.
[0026] In this embodiment, three CMOS sensors receive signals from three measured parts of the object, thereby generating three reference images and three sets of deformed images. Each set of deformed images contains several individual deformed images, such as... Figure 3 As shown, the DIC optical measurement process is illustrated using one of the three measured parts of the object as an example; the process is the same for the other measured parts.
[0027] In the reference image corresponding to the measured area, select a sub-region of interest, such as... Figure 3 As shown, in this embodiment, an image is selected from the reference image. Centered A pixel-sized square is used as a sub-region in the reference image.
[0028] Let's redefine the following related function formulas: in, It is the grayscale function of the reference image. It is the grayscale function of the deformed image. The average grayscale value of a sub-region of the reference image is calculated using the formula... Calculated using the formula Calculations show that The average gray level of the sub-region of the deformed image. Let be the vector of deformation parameters to be determined. Let be a function of the deformation parameters to be determined. M denoted as the radius of the sub-region.
[0029] The maximum and minimum values of the aforementioned related functions are obtained by performing optimization using existing iterative algorithms (such as the Newton-Raphson method used in this embodiment).
[0030] The correlation function can be selected from existing first-order or second-order correlation functions based on the analysis method and the complexity of the deformation. At the same time, the maximum or minimum value of the correlation function is obtained by iterative algorithm after optimization based on the actual selected correlation function. In this embodiment, since the deformation complexity is relatively simple, a first-order correlation function is selected, and the minimum value of the above correlation function is found by iterative algorithm.
[0031] Then, by finding the maximum and minimum values of the correlation function through optimization, the sub-region in the deformed image that is closest to the sub-region in the reference image can be determined, such as... Figure 3 As shown, this embodiment is based on... The sub-region centered on.
[0032] Finally, the center point of a sub-region of the reference image is determined by the determined sub-region of the deformed image. displacement components u , v .
[0033] In actual deformation, the sub-region may undergo not only translation but also rotation, stretching, shearing, and other deformations. Therefore, the following shape function is used to describe the coordinate mapping relationship between any point in the sub-region of the reference image and the corresponding point in the sub-region of the deformed image.
[0034] and The desired ones are respectively x and y Displacement and deformation in the direction, and For point To the center of the reference image sub-region distance, u , v The center of the reference image sub-region is at x , where , is the displacement in the y-direction. = . The rate of change of horizontal displacement in the horizontal direction of the reference image sub-region The rate of change of horizontal displacement in the vertical direction of the reference image sub-region The rate of change of vertical displacement of a sub-region of the reference image in the horizontal direction, The rate of change of vertical displacement of a sub-region in the reference image. , , , It is actually a displacement field ( u , v ) for coordinates ( x ,y The partial derivatives of ) are all displacement gradients of the reference image sub-region.
[0035] During the experiment, because the right side of the transparent test object reflects green light, the first CMOS sensor receives a green light signal. The computer then generates an image based on the green light signal. Therefore, the deformation analysis result obtained from the image generated based on the green light signal corresponds to the actual deformation state of the right side of the transparent test object. Similarly, the results of analyzing images generated based on signals transmitted by different CMOS sensors can correspond to different test parts of the transparent test object.
[0036] Finally, it should be noted that the above preferred embodiments are only used to illustrate the technical solution of this utility model and not to limit it. Although the utility model has been described in detail through the above preferred embodiments, those skilled in the art should understand that various changes can be made to it in form and detail without departing from the scope defined by the claims of this utility model.
Claims
1. A multi-band extraction system, characterized in that: The multi-band extraction system includes a light source (1), a semi-reflective mirror (2), a CMOS sensor (3), and a bandpass filter (4). There are n+1 CMOS sensors and n semi-reflective mirrors (2). Different colored coatings (6) are applied to different parts of the object under test (5). The light source (1) emits a light beam to the object under test (5). The coatings (6) are excited by the light beam and reflect different colored light beams. The different colored light beams all enter the first semi-reflective mirror (2). The first semi-reflective mirror (2) splits the incoming light beam into a transmitted beam and a reflected beam. The reflected beam enters the first CMOS sensor (3), and the transmitted beam sequentially enters the second to the nth semi-reflective mirrors (2). Each half-reflective lens (2) divides the light beam into a transmitted beam and a reflected beam. The reflected beam obtained by the splitting of each half-reflective lens (2) enters the CMOS sensor (3) corresponding to the order of the half-reflective lenses (2), until the transmitted beam enters the nth half-reflective lens. The transmitted beam obtained by the splitting of the nth half-reflective lens enters the (n+1)th CMOS sensor (3), and the reflected beam obtained by the splitting of the nth half-reflective lens enters the nth CMOS sensor (3). There are n+1 bandpass filters (4). One bandpass filter (4) allows one monochromatic light to pass through, and different bandpass filters (4) allow different colors of monochromatic light to pass through. The bandpass filters (4) cover the light receiving port of the CMOS sensor (3).
2. The multi-band extraction system according to claim 1, characterized in that: The incident surface and the exit surface of the semi-reflective lens (2) are both planes.
3. The multi-band extraction system according to claim 2, characterized in that: The acute angle between the incident surface of the semi-reflective lens (2) and the incident ray is 45°.
4. The multi-band extraction system according to claim 1, characterized in that: The coating (6) is formed by coating with fluorescent dyes or ordinary dyes.