CN102891956A - Method for designing compression imaging system based on coded aperture lens array - Google Patents

Abstract

The invention is a design method of a compressed imaging system based on a coded aperture lens array. The invention includes a field diaphragm array, a coded aperture template array, a sub-aperture spliced lens array, a coded sampling template array and a detector image plane array. The field of view diaphragm array realizes the block processing of large scenes, and the coded aperture template array performs spatial light modulation on the light from the scene target. The front-end coded sampling template array performs spatial light sampling, and finally compresses the imaging on the detector image plane array. The compressed imaging system based on the coded aperture lens array can compress and sample the target scene synchronously, which can greatly reduce the sampling frequency of the image signal and the cost of data storage and transmission. The required storage space greatly reduces the workload of optical system calibration.

Description

Design Method of Compression Imaging System Based on Coded Aperture Lens Array

technical field

The invention relates to a design method of a compressed imaging system based on a coded aperture lens array, and a high-resolution imaging system applicable to different spectral bands belongs to the field of optical imaging.

Background technique

With the continuous increase of people's demand for image information, the signal bandwidth carrying image information is getting wider and wider. The image signal processing framework based on the Nyquist sampling theorem will inevitably require the sampling rate and processing speed of the imaging system to increase day by day. Higher resolution, denser sampling, acquisition and transmission of massive image signals make traditional imaging systems face huge challenges both in terms of hardware and algorithms. In order to deal with this problem, in the actual application process, people often add DSP to the sensor, and use the compression mechanism to reduce the cost of data storage, processing and transmission. However, on the other hand, the system needs to perform cumbersome image domain transformation, coefficient sorting, and codec work, which additionally increases the complexity and cost of the sensor. Unfortunately, the result of high-speed sampling is that more than 80% of non-important data is discarded. This traditional "high-speed sampling" and then "compression" imaging mode wastes a lot of sampling resources. If the synchronous compressed sampling of image information can be realized, the complete information of the original image can be carried while reducing the amount of data, and the complex data processing and transmission process of sampling first and then compressing in the traditional imaging method can be avoided. Therefore, the sampling frequency of the image signal and the cost of data storage and transmission are greatly reduced, and the hardware cost of the imaging system is significantly reduced.

Compressed imaging technology is a new scientific research direction developed rapidly on the basis of compressed sensing theory. In 2006, famous American scientists Candés and Donoho formally proposed the theory of compressed sensing based on related research. This theory breaks through the bottleneck of Nyquist sampling theorem, and believes that the sampling amount of a signal does not depend on the bandwidth of the signal, but on the internal structure of the signal. If the signal is sparse or sparse in a transform domain, then a measurement matrix that is uncorrelated with the transform basis and satisfies constrained isometry can be used to project the high-dimensional signal to a low-dimensional space. The original signal is reconstructed with high probability from a small number of projection measurements by solving the minimum 0-norm optimization problem. At present, the main units conducting research on compressed imaging technology include Rice University in the United States, University of Arizona, MIT, Duke University, and Swiss Federal Institute of Technology. In 2006, Rice University in the United States successfully developed a single-pixel digital camera. Its design principle is to project the imaging target onto the digital micromirror device through the optical path system for spatial light modulation, and the reflected light is focused by the lens to a single photodiode, and the voltage value at both ends of the photodiode is a measured value. Repeat this projection operation multiple times to obtain multiple observations. The original target image is recovered by using the minimum total variation image reconstruction algorithm. WL Chan et al. proposed a new terahertz imaging method based on the concept of a single-pixel camera, which overcomes the shortcomings of existing terahertz imaging systems and can provide higher processing speed and stronger detection capabilities. Baheti and Neifeld at the University of Arizona improved the optical path structure of the single-pixel camera developed by Rice University, making its optical structure more compact and its light energy utilization more efficient. Researchers at the University of Delaware have applied the idea of single-pixel imaging to an electron microscope system. However, the single-pixel compressed imaging system outputs compressed and sampled image signals in a serial working manner. It uses a digital micromirror device to perform spatial light modulation on the imaging target. It needs to sample and project multiple times to obtain the measurement values required to reconstruct the original image, so the system is time-consuming. Compressed imaging of moving scenes or video images has certain limitations.

MIT's freeman research group proposed a compression imaging method using random mirrors. The significant difference from the traditional imaging method is that the system is composed of a plane mirror, a random spliced mirror group and a detector. Since the light from every point on the object may be imaged on the detector through the random mirror, the arbitrary splicing of the mirror actually realizes the function of the random projection matrix. However, the system uses randomly spliced lenses to achieve random measurement of the target image, so there is a problem of difficult calibration of the projection matrix.

The Duke University research group proposes the use of coded apertures to achieve compressed imaging of target objects in a parallel fashion. However, since this work has only been carried out initially, no in-depth research has been carried out on the coding mode of the aperture, the size of the aperture, and the relationship between the coding aperture template and the recovery accuracy of the compressed image. In addition, the projection matrix calibration workload of the compressed imaging system is huge. Compressed imaging of large scene images has become a technical difficulty of the system.

Contents of the invention

The invention aims to overcome the problems of long imaging time and heavy calibration workload of projection matrix in the existing compressed imaging system. The compressed imaging system based on the coded aperture lens array uses the field diaphragm array to realize the block processing of the large scene, divides the entire optical system into several sub-optical systems with the same form and function, and the coded aperture template array performs spatial analysis of the scene objects. Light modulation, using the sub-aperture splicing lens array to realize the imaging of the optical system with a large field of view, and using the coded sampling template array to randomly sample the converging light generated by the system to realize the compressed imaging of a single exposure. The system is easy to process and detect, and can realize high-resolution imaging with a large field of view and a low-resolution detector, and can be applied to high-resolution optical imaging of different spectral bands.

The details of the present invention are shown in FIG. 1 , which consists of a field stop array 1 , a coded aperture template array 2 , a sub-aperture spliced lens array 3 , a coded sampling template array 4 and a detector image plane array 5 .

The field stop array 1 is located between the scene and the coded aperture template array 2. All sub-templates in the coded aperture template array 2 are coded aperture stops using a special coding method. The sub-aperture spliced lens array 3 is n×n sub-aperture The positive focal power sub-lens of aperture splicing, n is a positive real number, the structure of its sub-aperture splicing lens array is shown in Figure 2, the encoding mode of all sub-templates in the encoding sampling template array 4 is Gaussian random encoding, and the encoding sampling template The array 4 is located at the front end of the image plane 5 of the focal plane detector array.

The working principle of the present invention: In order to compress and image the scene target objects within a large field of view on the detector image plane array 5, the system adopts the imaging mode of the coded aperture lens array. Assuming that the distance from the target scene to the sub-aperture splicing array is d, the function of the field stop array 1 is to complete the equal division of the large scene, which is located between the scene and the coded aperture template array, and set it to the sub-aperture splicing lens array The distance is t, the coded aperture template array 2 is used to perform spatial light modulation on the scene target, and it is placed close to the lens spherical surface in the sub-aperture splicing array, and the scene target is irradiated to the sub- A converging image is generated on the aperture-stitched lens array 3 , and is finally imaged on the detector image plane array 5 through random sampling by the coded sampling template array 4 close to the detector surface. Among them, the optical system is composed of several sub-optical systems with the same form and function, and the determination of the specific parameters of the optical structure of each sub-optical system can be obtained from the knowledge of geometric optics, and the calculation and derivation can be carried out according to the following steps:

Suppose h is the half-width of the sub-field diaphragm, h _l is the half-width of the light aperture of the sub-lens, h _m is the vertical half-width of each block area, and L is the value of the sub-field diaphragm in the field diaphragm array Δ is the distance between the sub-lenses in the sub-aperture spliced lens array, t is the distance from the field stop array to the sub-aperture spliced array, d is the distance from the scene to the sub-aperture spliced array, and S is the scene to the field stop array distance.

First determine the relationship between the half-width h of each field stop in the field stop array, the half-width h _l of the clear aperture of each sub-lens in the sub-aperture spliced lens array, and the half-width h _m of the segmented scene.

As shown in Figure 4, the following formula can be obtained according to geometric knowledge:

ΔMNG is similar to ΔCDG to obtain:

$\frac{\mathrm{<span\; class="notranslate">\; h</span>}}{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}{{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}$

ΔGZC is similar to ΔGOB:

$\frac{\mathrm{<span\; class="notranslate">\; h</span>}}{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}{\mathrm{<span\; class="notranslate">\; d</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}$

And because: d ₁ +d ₂ =t

The formula can be obtained from the above three formulas:

${\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}}\mathrm{<span\; class="notranslate">\; S</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; hd</span>}}{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

Second, determine the spacing of the sub-field diaphragms. As shown in Figure 5, the derivation process is as follows:

PO=QZ=2h+L

Therefore, the distance between the sub-field diaphragms is: L=2h _m -2h (2)

Finally, the pitch of the sub-lenses in the sub-aperture stitching lens array is determined. As shown in Figure 6, the specific derivation process is as follows:

ΔMUV is similar to ΔFDV, and it can be obtained:

$\frac{\mathrm{<span\; class="notranslate">\; \&Delta;</span>}}{\mathrm{<span\; class="notranslate">\; L</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}}{\mathrm{<span\; class="notranslate">\; the\; y</span>}}$

ΔFDV is similar to ΔEBV, and it can be obtained:

$\frac{\mathrm{<span\; class="notranslate">\; L</span>}}{{\mathrm{<span\; class="notranslate">\; 4</span>}\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; the\; y</span>}}{\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; S</span>}}$

Among them, S=d-t.

From the above two formulas, the formula can be obtained:

$\mathrm{<span\; class="notranslate">\; \&Delta;</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; 4</span>}\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; L</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; S</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; L</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>$

From the above formulas (1), (2) and (3), it can be seen that if we select some of the parameters, the optical structure of the system can be completely determined.

The imaging system is arranged in an orderly manner according to the x, y, z right-hand space coordinate system, the z-axis direction is the direction of the optical axis, the y-axis is in the plane shown in Figure 2, the x-axis is perpendicular to the yz plane, and the yz coordinate plane is the meridian plane of the optical system . Field diaphragm array 1, coded aperture template array 2, sub-aperture spliced lens array 3, coded sampling template array 4 and detector image plane array 5, the optical axes of all subsystems are parallel to the optical axis of the system, and the light propagation Arranged sequentially in the direction, as shown in Figure 3.

Beneficial effects of the present invention: the compressed imaging system uses a coded aperture lens array to perform spatial light modulation on scene target objects, and simultaneously completes sampling and compression of image signals, thereby realizing compressed imaging with a large field of view and high resolution, thereby greatly reducing the The sampling frequency of image signals and the cost of data storage and transmission greatly reduce the amount of image data that needs to be acquired from the source. The system has a simple structure and is easy to process and detect, and is especially suitable for field, airborne or embedded high-resolution light imaging systems with limited hardware resources.

Description of drawings

Fig. 1 is a structural schematic diagram of the present invention.

In the figure, 1-field diaphragm array, 2-coded aperture template array, 3-sub-aperture spliced lens array, 4-coded sampling template array, 5-detector image plane array.

FIG. 2 is a schematic diagram of a sub-aperture stitched lens array used in the present invention.

Fig. 3 is a schematic diagram of a coordinate system used in the present invention.

Fig. 4 is a geometrical schematic diagram for calculating the half-width of the sub-field diaphragm, the half-width of the light aperture of the sub-lens and the half-width of the sub-block scene.

Fig. 5 is a schematic diagram of the geometry for calculating the distance between the diaphragms of the sub-fields of view.

Fig. 6 is a schematic diagram of the geometry for calculating the sub-lens spacing in the sub-aperture stitched lens array.

Detailed ways

The present invention will be described in further detail below in conjunction with specific implementation examples. Here, the exemplary embodiments and descriptions of the present invention are used to explain the present invention, but not to limit the present invention.

The present invention is implemented by the structure shown in Fig. 1, and its detailed content is as shown in Fig. 1, and system is by field stop array 1, code aperture template array 2, sub-aperture splicing lens array 3, code sampling template array 4 and detector The image plane array 5 constitutes.

As shown in Figure 1, a subsystem B in the scene is taken as an example for illustration. The scene light of subsystem B first passes through the corresponding sub-field diaphragm, which can effectively allow the scene light of area B to pass through, and block the light of the adjacent scene area from passing through, so as to achieve equalization of the entire scene. Block. The light passing through the sub-field diaphragm passes through the sub-coded aperture template corresponding to the sub-optical system to realize the spatial light modulation of the image information in the area, and the modulated light is irradiated on the corresponding sub-lens in the sub-aperture spliced lens array , wherein, all sub-lenses in the lens array include any of the following materials: acrylic resin, cycloolefin copolymer, polystyrene, Lucite resin, Ultem, Tyril, Merton, or polymethylpentene, etc. Finally, the image is imaged on a CCD detector or other array detectors through a corresponding sub-code sampling template to realize compressed imaging. Wherein, all the sub-coded aperture templates in the coded aperture template array 2 and all the sub-coded sampling templates in the coded sampling template array 4 can adopt grid-type distributed square or circular templates, and the size of the grid is between nanometers and micrometers Between, material can adopt glass material. The light transmittance of each grid on the sub-coding aperture template can adopt a cyclic coding method, and the light transmittance of each grid on the sub-coding sampling template can adopt a 0-1 pseudo-Gaussian random coding method.

The optical system is arranged in an orderly manner according to the x, y, z right-hand space coordinate system, the z-axis direction is the direction of the optical axis, the y-axis is in the 3 planes shown in the figure, the x-axis is perpendicular to the yz plane, and the yz coordinate plane is the meridian plane of the optical system .

The specific description above further elaborates the purpose, technical solution and beneficial effect of the invention. It should be understood that the above description is only a specific embodiment of the present invention and is not used to limit the protection of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims (5)

1. The design method of the compressed imaging system of the basic coded aperture lens array is characterized in that the field diaphragm array is used to limit the incident light so as to realize the block processing of the large scene, and the coded aperture template array with a special coded form is capable of processing the images from the scene. The target light is subjected to spatial light modulation, and the modulated light enters the sub-aperture splicing lens array to be converged. The converged light is then sampled spatially by the coded sampling template array located at the front of the focal plane, and finally compressed and imaged on the detector image plane array. 2. A method for designing a compressed imaging system based on a coded aperture lens array according to claim 1, further characterized in that the field stop array is used to block the scene, so that the entire optical system is divided into several Form sub-optical system, and then realize the splicing imaging of large scenes. 3. A kind of compression imaging system design method based on coded aperture lens array according to claim 1, it is also characterized in that, the special coding form of all coded aperture templates that coded aperture template array comprises all adopts cyclic coding or Top Leeds code. 4. A method for designing a compression imaging system based on a coded aperture lens array according to claim 1, further characterized in that the sub-aperture spliced lens array is n×n sub-lenses with positive refractive power, and n is a positive real number , where the sub-lenses in the sub-aperture spliced lens array all have the same specifications, and each sub-lens has a clear aperture half-width h _l , a segmented scene half-width h _m , and each field stop in the field stop array The half-width h of satisfies the relationship:

Wherein, t is the distance from the field stop array to the sub-aperture stitching array, d is the distance from the scene to the sub-aperture stitching array, and S is the distance from the scene to the field stop array. 5. A method for designing a compressed imaging system based on a coded aperture lens array according to claim 1, further characterized in that all coded sampling templates included in the coded sampling template array positioned at the front end of the focal plane adopt random Gaussian codes form.

Images