CN102821236A

CN102821236A - Compression imaging system based on separated two-dimensional compression sensing principle

Info

Publication number: CN102821236A
Application number: CN2012102443503A
Authority: CN
Inventors: 王为
Original assignee: Tianjin Normal University
Current assignee: TIANJIN JIANYI TECHNOLOGY Co Ltd
Priority date: 2012-07-16
Filing date: 2012-07-16
Publication date: 2012-12-12
Anticipated expiration: 2032-07-16
Also published as: CN102821236B

Abstract

The invention discloses a compressed imaging system based on a separated two-dimensional compressed sensing theory. The imaging system adopts a sequential structure, and the image is imaged on the digital micromirror device DMD through the focusing lens, and the micromirror angle of the DMD is controlled by the constructed random row measurement matrix to process the row information of the image; the obtained row information is passed through the microlens array Focusing on the linear array CMOS image sensor, the constructed random column measurement matrix controls the CMOS sensor pixels to read and process the column information of the image. So far, an imaging process is realized, and the compressed information of the image is obtained through a certain number of imaging processes to complete the entire process. Compressed imaging. The system has fast imaging speed and strong anti-interference ability, and can provide a new way for harsh environment imaging and encrypted imaging.

Description

Compression imaging system based on separate type two dimensional compaction perception theory

The application obtains the young fund (12JCQNJC00400) in Tianjin; The subsidy of Tianjin institution of higher education science and technology development fund planning item (20100716).

Technical field

The present invention relates to a kind of compression imaging system, but belong to imaging and the ciphered compressed imaging field that is applicable to adverse circumstances based on the compressed sensing theory.

Background technology

The compressed sensing theory is pointed out as long as signal has sparse property or compressibility; Promptly can through one with the incoherent random measurement matrix of sparse transform-based with on higher-dimension signal projection to a lower dimensional space, can from these a small amount of projections, reconstruct primary signal through finding the solution an optimization problem then.Under this theoretical frame, sample frequency no longer is decided by the bandwidth of signal, and depends on structure and the content of information in signal; And the compressed sensing theory will be sampled and combined with compression, the contradiction between effectively having solved sampling during classical signals is handled and having compressed.Image information has characteristics such as data volume is big, bandwidth is wide, in order to reduce the channel pressure of images, usually image information is carried out compressed encoding, to reduce the data volume of presentation video.Existing image processing based on compressed sensing belongs to digital compression perception field more, promptly utilizes the compressed sensing theory to handle to the image digitization information of having gathered; Do not realize the advantage of image information compression (that is compression imaging technique) in performance and the sampling process simultaneously.

Existing minority mainly adopts sequential organization and parallel organization based on the theoretical compression imaging system of compressed sensing, but the former is longer for the required Measuring Time of large-size image, and it is big to measure noise, poor practicability.Though latter's image taking speed is fast, imaging system need be demarcated in advance, its process more complicated, and amount of calculation is bigger; And for realizing the large-size images imaging, its imaging system structure is bigger, is difficult to practicability.

Summary of the invention

The present invention is directed to the problem and the active demand of compressed sensing theory existence in being compressed into the picture realization of prior art; Give full play to the advantage that realizes the image information compression in the theoretical sampling process of compressed sensing simultaneously, a kind of compression imaging system based on separate type two dimensional compaction perception theory is provided.

The technical scheme that the present invention takes for the technical problem that exists in the solution known technology is:

Compression imaging system based on separate type two dimensional compaction perception theory of the present invention, this imaging system comprise condenser lens, DMD DMD (Digital Micromirror Device), microlens array and linear array CMOS (the Complementary Metal Oxide Semiconductor) imageing sensor that is linked in sequence and is provided with; Target image projects on the DMD DMD via condenser lens, and the micro mirror angle of STOCHASTIC CONTROL DMD DMD realizes the row processing of image information; By microlens array reflective information is focused on the linear array cmos image sensor, the pixel of STOCHASTIC CONTROL linear array cmos image sensor reads the row of realizing image information and handles again.

This imaging system is provided with control circuit, and control circuit is made up of image-forming module drive circuit and control module circuit; The image-forming module drive circuit comprises the optics control section, and DMD control section that is connected with the optics control section and CMOS control section; The control module circuit comprises FPGA timing sequence generating and the control section that is connected with the CMOS control section with the DMD control section, DSP calculation process part that is connected with control section with the FPGA timing sequence generating and the demonstration storage area that partly is connected with the DSP calculation process.

Advantage and good effect that the present invention has are:

The present invention is theoretical according to compressed sensing, and is theoretical in conjunction with tensor product, is separate type two dimensional compaction perception theory with one dimension compressed sensing theoretical transformation, for the compression imaging system of practicality and high efficiency is provided fundamental basis.Theoretical based on two dimensional separation formula compressed sensing, reduced to produce and the required memory headroom of storage random measurement matrix, improved the practicality of imaging system, can be used for the picture that is compressed into of large-size images.The present invention adopts sequential organization; Realize the capable Information Compression of image by DMD DMD; The column information of being realized image by the linear array cmos image sensor compresses, and can reduce the required number of times of imaging process, improves the imaging system signal to noise ratio; Can realize the picture that is compressed into of highly effective, can be used for adverse circumstances imagings, be encrypted to and special imaging field such as look like.This system imaging speed is fast, antijamming capability strong, can form images, be encrypted to picture for adverse circumstances provides new approach.

Description of drawings

Fig. 1 is separate type two dimensional compaction perception principle figure of the present invention;

Fig. 2 is the picture system optics structural representation that is compressed into of the present invention;

Fig. 3 is the structural representation that is compressed into the picture system, control circuit of the present invention.

Critical piece symbol description among the figure:

1. condenser lens 2. DMDs

3. microlens array 4. line scan image sensors.

Embodiment

Followingly the present invention is carried out detailed explanation with reference to accompanying drawing and embodiment.

Compression imaging system based on separate type two dimensional compaction perception theory of the present invention; Theoretical according to compressed sensing; Theoretical in conjunction with tensor product, be separate type two dimensional compaction perception theory with one dimension compressed sensing theoretical transformation, for the compression imaging system of practicality and high efficiency is provided fundamental basis.Fig. 1 has described the principle of separate type two dimensional compaction perception; If original image two dimension expression-form is

Figure DEST_PATH_249016DEST_PATH_IMAGE001

; Random measurement matrix

Figure DEST_PATH_878712DEST_PATH_IMAGE002

can be expressed as

Figure DEST_PATH_494019DEST_PATH_IMAGE003

through tensor product; Then theoretical according to CS, image accidental projection value G is:

Figure DEST_PATH_892771DEST_PATH_IMAGE004

In the formula

Figure DEST_PATH_260298DEST_PATH_IMAGE005

,

Figure DEST_PATH_377290DEST_PATH_IMAGE006

,

Figure DEST_PATH_590096DEST_PATH_IMAGE007

are respectively random row survey bureau matrix and random column is measured matrix.Can know from separate type two dimensional compaction perception theory; The memory space of random measurement matrix becomes

Figure DEST_PATH_378853DEST_PATH_IMAGE009

from traditional C theoretical

Figure DEST_PATH_106004DEST_PATH_IMAGE008

; For large-size images; With the treating capacity that greatly alleviates imaging system, make imaging system practicality and high efficiency more.

Fig. 2 is the picture system optics structural representation that is compressed into of the present invention; According to principle shown in Figure 1, the compression imaging system of a kind of sequential organization of design.As shown in Figure 2, imaging system of the present invention comprises condenser lens 1, DMD 2, microlens array 3 and the array image sensor 4 that is linked in sequence and is provided with.Imaging system adopts sequential organization, and target image images on the DMD DMD through condenser lens, is measured the micro mirror angle of matrix control DMD by the random row of structure the capable information of image is handled; The capable information of obtaining is focused on the linear array cmos image sensor through microlens array; Measuring matrix control cmos sensor pixel by the random column of structure reads the column information of image is handled; So far realize Polaroid process; Obtain the compressed information of image through the imaging process of certain number of times, accomplish the whole picture that is compressed into.

Fig. 3 is the control circuit structural representation of compression imaging system of the present invention, and is as shown in Figure 3, and control circuit is made up of image-forming module drive circuit and control module circuit; The image-forming module drive circuit comprises the optics control section, and DMD control section that is connected with the optics control section and CMOS control section; The control module circuit comprises FPGA timing sequence generating and the control section that is connected with the CMOS control section with the DMD control section; DSP calculation process part that is connected with control section with the FPGA timing sequence generating and the demonstration storage area that partly is connected with the DSP calculation process, and read-only memory RAM.The control module circuit mainly realizes compressing the storage of imaging system random measurement matrix and the functions such as generation of control timing.

Claims

1. A compressed imaging system based on separate two-dimensional compressed sensing theory, characterized in that: the imaging system includes sequentially connected focusing lens (1), digital micromirror device (2), microlens array (3) and Array image sensor (4); the target image is projected onto the digital micromirror device (2) through the focusing lens (1), and the micromirror angle of the digital micromirror device (2) is randomly controlled to realize row processing of image information; then the microlens The array (3) focuses reflection information on the line array image sensor (4), and randomly controls the pixel reading of the line array image sensor (4) to realize column processing of image information.

2. A compressed imaging system based on separate two-dimensional compressed sensing theory, characterized in that: the imaging system is provided with a control circuit, the control circuit is composed of an imaging module drive circuit and a control module circuit; the imaging module drive circuit includes an optical control part , and the DMD control part and the CMOS control part connected with the optical control part; the control module circuit includes an FPGA timing generation and control part connected with the DMD control part and the CMOS control part, and a DSP operation processing part connected with the FPGA timing generation and control part and a display storage part connected with the DSP operation processing part.