CN103162843A - Zero shutter thermal infrared imager based on voice operated exchange (VOX) detector and use method thereof - Google Patents

Abstract

The invention relates to a zero shutter thermal infrared imager based on a voice operated exchange (VOX) detector. The zero shutter thermal infrared imager based on the VOX detector comprises a high-low temperature chamber and a machine core, wherein a monitor, a computer and a power source are respectively connected with the machine core. The use method of the thermal infrared imager includes the following steps: collecting original images; forecasting a background; revising a point 2 and a point 1; and performing image enhancement display processing. Compared with a usual shutter revision method, the zero shutter thermal infrared imager based on the VOX detector and the use method thereof can effectively reduce the size and the power consumption of an infrared machine core, are unlike a conventional method in which a shutter needs to be used to obtain uniform images at every time of revision, and use the original images as background images to revise, and therefore system design is simplified greatly, the size and the power consumption are reduced, and simultaneously due to zero shutter design, the problems of uneven images caused by heat emitting of the shutter and the like are avoided.

Description

A kind of based on the VOx detector without shutter thermal infrared imager and using method thereof

Technical field

The invention belongs to the infrared imagery technique field, be specifically related to a kind of based on the VOx detector without shutter thermal infrared imager and using method thereof.

Background technology

The developing history of non-refrigeration thermal imaging.The pick-up tube that infrared thermoelectrical material is made occurs in early days, and made thermal television.But sensitivity is too low, and the image streaking phenomenon is serious, is not therefore widely applied.Eight, the nineties once was applied in the industrial heat imaging.Although but this generation Thermal Imaging Common Modules is when obtaining widely applying, finding has very serious shortcoming to be: cost is high, working service and logistics support difficulty (refrigeration), poor reliability.So English, the military of U.S. two countries country have formulated the plan that develops non-refrigeration thermal imaging, purpose is to develop the thermal imaging that the user affords and affords to use.Plan is maintained secrecy, and starts from the late nineteen seventies in last century, until there was successively result to announce in 1992.The material for detector that adopts is mainly two kinds, i.e. pyroelectricity and bolometer.Adopt having of pyroelectricity material: U.S. Texas company developed the BST of 100 * 100 yuan, NETD0.5 ℃ in 1987.Nineteen ninety is assessed first imaging array of paying, and larger focal plane arrays (FPA) has reached desirable 0.3 ℃ of temperature resolution, and recording NETD is 0.08 ℃.Issued the array of 328 * 245 yuan in 1993, pixel centre distance is 48.5 μ m.System is tested the NETD that obtains when using f/1 optics less than 0.04 ℃.Britain has developed the PZT of 100 * 100 yuan in 1988.U.S. Loral company had developed the pyroelectric electric device of 192 * 128 yuan in 1991, and NETD is less than 0.07 ℃.What adopt bolometer is mainly U.S. Honeywell company, and the said firm is 80000 yuan of resistor-type bolometers of development in 1985, NETD0.3 ℃.Nineteen ninety adopts 336 * 240 element array of vanadium oxide to make first portable thermal imaging system, the NETD=39mK of acquisition (F=1).After this, no matter be pyroelectricity or bolometer, device and the thermal imaging of 320 * 240 series have occurred a lot.Since late nineteen nineties, the size reduction of detector to 30 about μ m, the array of 640 * 480 series occurred, and the NETD of product has reached below 50mK.This kind thermal imaging system has been demonstrated by present many companies.

The infrared focal plane array image-forming system is the trend of infrared imagery technique development, but the restriction due to technological level, between each probe unit response characteristic of infrared focal plane array (IRFPA), ubiquity the heterogeneity problem, it will cause the temperature resolution of infrared imaging system significantly to descend, satisfy the request for utilization of infrared imaging system so that it is difficult to, thereby the infrared focal plane array that uses in engineering almost all adopts Nonuniformity Correction without any exception.

For the corresponding bearing calibration that the heterogeneity problem of infrared focal plane array proposes, mainly be divided into two large classes: a class is based on the bearing calibration of calibration, as one point method, and two-point method etc.Such Method And Principle is succinct, and hardware is easy to realize and is integrated; Correction accuracy is high, can be used for the tolerance of scene temperature; Target is had no requirement, and is the main method that adopts in actual IRFPA subassembly product.But these class methods are subject to the correction error that the IRFPA response drift is brought;

The another kind of self-adapting correction method that is based on the scene class, as time domain high-pass filtering correction method, neural networks correction algorithm and constant statistics constraint correction method etc.These class methods can overcome the correction error that the IRFPA response drift is brought to a certain extent, do not require or only need simple calibration, renewal correction coefficient adaptive according to scene information, but during this class algorithm application, calculated amount is large, often need special parallel computer architecture to realize, be unfavorable for the realization, integrated and to the real-time processing of scene of system hardware.

The Central China University of Science and Technology easy newly-built wait the people in " infrared and laser engineering " the 33rd the 1st phase of volume in 2004 " two point calibrations of infrared focal plane array heterogeneity and according to " literary composition take the linear response model of Planck (Plank) radiation law and infrared acquisition unit for basic, two point calibration methods of the infrared focal plane asymmetric of in theory intactly having derived.The physical basis of two-point method proved theoretically in article, show if the response of IRFPA is stable, linear, the algorithm of two point calibrations does not have error, but in fact the first response of IRFPA detection is all nonlinear, and there is a problem of response drift, therefore, there is larger remainder error with two point calibration methods.

The deficiency of these bearing calibrations is all that remainder error is arranged, and remainder error is relevant with the temperature of detector, shutter, and it is exactly one deck noise that remainder error shows on video, affects performance and the application of thermal imaging system.

Analyze two large class asymmetric correction methods in the past, bearing calibration based on calibration, as two-point method, because there is response drift in IRFPA, actual timing need to utilize even reference source as shutter, periodically gather the extend blackbody image as a setting frame be used for calibration, otherwise can have larger remainder error after proofreading and correct; Based on the self-adapting correction method of scene class, as neural networks correction algorithm, during application, calculated amount is large, is unfavorable for the real-time processing of system, and the problems such as target fade-out and pseudomorphism occur toward the contact meeting after proofreading and correct.

Summary of the invention

The present invention propose a kind of based on the VOx detector without shutter thermal infrared imager and using method thereof.The method of developing on the basis of VOx detector and having realized proofreading and correct without shutter has greatly been eliminated remainder error, has made up weak point of the prior art.

Technical scheme of the present invention is achieved in that

A kind of thermal infrared imager without shutter based on the VOx detector comprises high-low temperature chamber, is provided with movement in described high-low temperature chamber, and described movement is connected with respectively monitor, computing machine and power supply.

The described thermal infrared imager without shutter based on the VOx detector also comprises camera lens, is arranged on described high-low temperature chamber.

The described using method without the shutter thermal infrared imager based on the VOx detector comprises the following steps:

1) original image acquisition step: program is write in infrared movement, switched on power, video and serial ports, be placed in high-low temperature chamber, the uniform planar in alignment box drops to lowest temperature T _L℃ degree is incubated 2 hours, then powers on to movement, sends acquisition to movement at once, and the automatic coefficient capture program in movement will start, and movement can gather the point under Current Temperatures at once, as the minimum temperature sample; Allow high-low temperature chamber slowly be warmed up to highest temperature T _H℃.In this process, movement can be according to the temperature that detects, at the sample image that gathers on one group of temperature spot under the detector different temperatures, and this image sets is existed on movement in FLASH, simultaneously the temperature value of record detector this moment.

2) background forecast step: having the image sets in FLASH more than using is the basis, utilizes prediction algorithm to obtain background image;

3) 1 and 2 point calibrations: take the temperature of current detector as input parameter, obtaining the background forecast image is ImgBK, recycles and 2 point calibration formula at 1 infrared image is proofreaied and correct:

Formula 1 Im hAdj (i, j)=(ImgU (i, j)-ImgBK (i, j)) * Nuc (i, j),

Wherein ImgAdj is the image after proofreading and correct, the horizontal and vertical pixel coordinate of (i, j) presentation video matrix; ImgU is original image, and ImgBK is the background image of prediction, and Nuc is 2 point calibration parameters.

Described original image acquisition step comprises:

Give thermal imaging system programming program, then the connecting communication cable is placed on thermal infrared imager in high-low temperature chamber, and setting high-low temperature chamber is lowest temperature T _L℃, being incubated 2 hours allows thermal imaging system fully lower the temperature, power on to thermal imaging system, send the control command of doing the shutter coefficient to thermal imaging system by communication cable, thermal imaging system begins detected temperatures, in the time of temperature variation n ℃, gathers present image, and exist in the FLASH of thermal imaging system inside, record simultaneously current accurate temperature value.Temperature stabilization and picture quality when guaranteeing that thermal imaging system gathers image must be controlled the programming rate of high-low temperature chamber, slowly are warmed up to highest temperature T _H℃;

For guaranteeing to collect the sample of maximum temperature point, at high temperature T _H℃ the time, be incubated 2 hours, the temperature that guarantees thermal imaging system itself rises to the highest, like this, thermal infrared imager has collected M and has organized image in whole temperature section, record as follows:

TEMP[]＝{T ₀，T ₁，T ₂，.....T _M-1}；

TEMP_ADDR[]＝{TA ₀，TA ₁，TA ₂，.....TA _M-1}；

TEMP[] be the temperature value of the correspondence of M group image;

TEMP_ADDR[] be that M organizes image FLASH memory address corresponding to image.

Described background forecast step comprises:

1) thinking or the thought of two point calibrations:

Im?gAdj(i，j)＝(ImgU(i，j)-ImgBK(i，j))*Nuc(i，j)

2) two point calibration coefficient Nuc (i, j) obtain unanimously with our two traditional point calibration methods, gain correction coefficient obtains: image B lackH and the BlackL of the homogeneous radiation body under the different radiation intensity of two width under same ambient temperature conditions;

Formula 2 $\mathrm{Nuc}(i,j)=\frac{\mathrm{Mean}\left(\mathrm{BlackH}\right)-\mathrm{Nean}\left(\mathrm{BlackL}\right)}{\mathrm{BlackH}(i,j)-\mathrm{Black}(i,j)}$

3) background image ImgBK (i, j) is the background of preserving in thermal imaging system, obtains new background image NImgBK in conjunction with actual image correction of obtaining;

Formula 3 NIm gBK=K*Im gBK+B

4) image adopts existing gain coefficient Nuc and biasing coefficient ImgBK to carry out two point calibrations and obtains image I m gAdj, in image I mgAdj, search gray scale difference value within facing the territory scope less than the pixel of certain threshold value pair, will satisfy all pixels of this threshold condition to search out; The computation process of adjusted coefficient K and B is: the output difference of neighbor:

Formula 4 is replaced partial parameters for simplifying expression formula:

E＝Nuc(a)(Im?gU(a)-K*Im?gBK(a)-B)

-Nuc(b)*(Im?gU(b)-K*Im?gBK(b)-B)

E＝Nuc(a)*Im?gU(a)-Nuc(b)*Im?gU(b)

-K*(Nuc(a)*Im?gBK(a)

-Nuc(b)*Im?gBK(b))

-B*(Gain(a)-Gain(b))

A1＝Nuc(a)*ImgU(a)-Nuc(b)*ImgU(b)

Formula 5A2=Nuc (a) * ImgBK (a)-Nuc (b) * ImgBK (b)

A3＝Nuc(a)*ImgBK(b)

Obtain respectively:

$K=\frac{\mathrm{\&Sigma;}(\mathrm{<figure-callout\; id="1"\; label="A"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">A</figure-callout>}1*A2)*\mathrm{\&Sigma;}{\mathrm{<figure-callout\; id="3"\; label="A"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">A</figure-callout>}3}^2-\mathrm{\&Sigma;}(A1*A3)*\mathrm{\&Sigma;}(A2*A3)}{\mathrm{\&Sigma;}{\mathrm{<figure-callout\; id="2"\; label="A"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">A</figure-callout>}2}^2*\mathrm{\&Sigma;}(A2*A3)*\mathrm{\&Sigma;}(A2*A3)}$

Formula 6 $B=\frac{\mathrm{\&Sigma;}(\mathrm{<figure-callout\; id="1"\; label="A"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">A</figure-callout>}1*A3)*\mathrm{\&Sigma;}{\mathrm{<figure-callout\; id="3"\; label="A"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">A</figure-callout>}3}^2-\mathrm{\&Sigma;}(A1*A2)*\mathrm{\&Sigma;}(A2*A3)}{\mathrm{\&Sigma;}{\mathrm{<figure-callout\; id="2"\; label="A"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">A</figure-callout>}2}^2*\mathrm{\&Sigma;}{\mathrm{<figure-callout\; id="3"\; label="A"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">A</figure-callout>}3}^2-\mathrm{\&Sigma;}(A2*A3)*\mathrm{\&Sigma;}(A2*A3)}$

Obtain real-time being biased to:

Formula 7 NIm gBK=M*Im gBK+C

Obtaining real-time NIm gAdj result is:

Formula 8NIm gAdj=(Im gU (i, j)-NIm gBK (i, j)) * Nuc (i, j)

Described background forecast step is switched based on many backgrounds, and the background correction is carried out in two background weightings.Above process is based on algorithm idea and the calculation procedure of single background modification method, but in the reality test, can find single background Shortcomings.When the focal plane temperature differed greatly with the background focal plane of preserving, this background can't obtain more suitable real-time background, therefore needed to adopt the background of a plurality of different focal planes temperature, switched background when the focal plane temperature changes.But can there be the problem of image flicker in so direct switching background, so considers to adopt through two revised backgrounds of background, and the weighting coefficient according to realtime graphic and two backgrounds of two background calculating obtains real-time background.

to gather 4 different focal planes temperature backgrounds as example, when the real time temperature scope between temperature 1 and temperature 2, background when background is with temperature 1 and temperature 2 in real time is weighted to be asked for, real time temperature is the closer to which background temperature, the weight of current background in background as a result is just large, when real-time focal plane temperature during near background temperature 2, at this moment the weight of temperature 2 is substantially near 1, when temperature continues to raise, real time temperature reaches between this degree temperature 2 and Ben Du temperature 3, namely adopt background 2 and background 3 to be weighted, because temperature is rising, when temperature during just greater than temperature 2, at this moment still take this degree temperature 2 as major weight.Like this, just the problem of image flicker can not occur when background switches, and prove in slave computer is transplanted.

Algorithm idea is the same with single background modification method of front, just adopted the background at two different focal planes temperature, calculates the weighting coefficient of two backgrounds according to realtime graphic and two backgrounds, obtains real-time background

Formula 9 Noffset=K1*Im gBK1+K2*Im gBK2+B

The computation process of weighting coefficient K1 and K2 and B is as follows:

With single background algorithm idea, according to two backgrounds when the selection of front focal plane temperature and this temperature contrast minimum, select one of them background and gain coefficient, obtain image NIm gAdj data, search is being faced gray scale difference value in the scope of territory less than the pixel of certain threshold value pair in image NIm gAdj, so the pixel that will satisfy this threshold condition to search out.

Adjusted coefficient K 1, the computation process of K2 and B is:

Formula 10

E＝Nuc(a)(ImgU(a)-K1*Im?gBK1(a)-K2*Im?gBK2(a)-C)-

Nuc(b)*(X(b)-K1*Im?gBK1(b)-K2*Im?gBK2(b)-C)

E＝Nuc(a)*X(a)-Nuc(b)*X(b)

-K1*(Nuc(a)*Im?gBK1(a)-Nuc(b)*Im?gBK1(b))

-K2*(Nuc(a)*Im?gBK2(a)*Nuc(b)*Im?gBK2(b))

-C*(Nuc(a)-Nuc(b))

For simplifying expression formula, partial parameters is replaced:

Formula 11

A1＝Nuc(a)*X(a)-Nuc(b)*X(b)

A2＝Nuc(a)*Im?gBK1(a)-Nuc(b)*Im?gBK1(b)

A3＝Nuc(a)*Im?gBK2(a)-Nuc(b)*Im?gBK2(b)

A4＝Nuc(a)-Nuc(b)

Order

Formula 12

$\mathrm{<figure-callout\; id="1"\; label="T"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">T</figure-callout>}1=\mathrm{\&Sigma;}{A2}^2-\frac{\mathrm{\&Sigma;}(A2*A4)*\mathrm{\&Sigma;}(A2*A4)}{\mathrm{\&Sigma;}{A4}^2}$

$\mathrm{<figure-callout\; id="2"\; label="T"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">T</figure-callout>}2=\mathrm{\&Sigma;}(A2*A3)-\frac{\mathrm{\&Sigma;}(A3*A4)*\mathrm{\&Sigma;}(A2*A4)}{\mathrm{\&Sigma;}{A4}^2}$

$\mathrm{<figure-callout\; id="2"\; label="T"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">T</figure-callout>}2=\mathrm{\&Sigma;}(A1*A2)-\frac{\mathrm{\&Sigma;}(A1*A4)*\mathrm{\&Sigma;}(A2*A4)}{\mathrm{\&Sigma;}{A4}^2}$

Order

Formula 13

$\mathrm{<figure-callout\; id="1"\; label="D"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">D</figure-callout>}1=\mathrm{\&Sigma;}(A2*A3)-\frac{\mathrm{\&Sigma;}(A2*A4)*\mathrm{\&Sigma;}(A3*A4)}{\mathrm{\&Sigma;}{A4}^2}$

$\mathrm{<figure-callout\; id="2"\; label="D"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">D</figure-callout>}2=\mathrm{\&Sigma;}{A3}^2-\frac{\mathrm{\&Sigma;}(A3*A4)*\mathrm{\&Sigma;}(A3*A4)}{\mathrm{\&Sigma;}{A4}^2}$

$\mathrm{<figure-callout\; id="3"\; label="D"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">D</figure-callout>}3=\mathrm{\&Sigma;}(A1*A3)-\frac{\mathrm{\&Sigma;}(A1*A4)*\mathrm{\&Sigma;}(A3*A4)}{\mathrm{\&Sigma;}{A4}^2}$

Obtain

Formula 14

$\mathrm{<figure-callout\; id="1"\; label="K"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">K</figure-callout>}1=\frac{\mathrm{<figure-callout\; id="3"\; label="T"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">T</figure-callout>}3*D2-\mathrm{<figure-callout\; id="3"\; label="D"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">D</figure-callout>}3*T2}{\mathrm{<figure-callout\; id="1"\; label="T"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">T</figure-callout>}1*D2-\mathrm{<figure-callout\; id="2"\; label="T"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">T</figure-callout>}2*\mathrm{<figure-callout\; id="1"\; label="D"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">D</figure-callout>}1}$

$\mathrm{<figure-callout\; id="2"\; label="K"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">K</figure-callout>}2=\frac{\mathrm{<figure-callout\; id="3"\; label="T"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">T</figure-callout>}3*D1-\mathrm{<figure-callout\; id="3"\; label="D"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">D</figure-callout>}3*T1}{\mathrm{<figure-callout\; id="2"\; label="T"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">T</figure-callout>}2*D1-\mathrm{<figure-callout\; id="1"\; label="T"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">T</figure-callout>}1*\mathrm{<figure-callout\; id="2"\; label="D"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">D</figure-callout>}2}$

$B=\frac{\mathrm{\&Sigma;}(A1*A4)}{\mathrm{\&Sigma;}{A4}^2}-\mathrm{<figure-callout\; id="1"\; label="K"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">K</figure-callout>}1*\frac{\mathrm{\&Sigma;}(A2*A4)}{\mathrm{\&Sigma;}{A4}^2}-\mathrm{<figure-callout\; id="2"\; label="K"\; filenames="HSA00000867135400012.png,HSA00000867135400013.png"\; state="\{\{state\}\}">K</figure-callout>}2*\frac{\mathrm{\&Sigma;}(A3*A4)}{\mathrm{\&Sigma;}{A4}^2}$

Obtain real-time being biased to:

Formula 15 Noffset=K1*Im gBK1+K2*Im gBK2+B

Obtaining realtime graphic NIm gAdj result is:

Formula 16NIm gAdj=(Im gU (i, j)-NIm gBK (i, j)) * Nuc (i, j).

Beneficial effect of the present invention is:

The present invention contrasts prior art and has following innovative point:

Remove the algorithm of shutter;

To go the algorithm application of shutter on the VOx movement;

Automatically do the apparatus and method of shutter coefficient.

The present invention contrasts prior art and has following remarkable advantage:

Improved the heterogeneity of image, the heterogeneity of having avoided structure and thermal environment to bring;

Compare with the thermal imaging system that shutter is arranged, reduced power consumption, reduced the complexity of system, improved signal to noise ratio (S/N ratio);

There is no shutter, improved the reliability of system, avoided some faults of shutter.

Description of drawings

In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, the below will do to introduce simply to the accompanying drawing of required use in embodiment or description of the Prior Art, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skills, under the prerequisite of not paying creative work, can also obtain according to these accompanying drawings other accompanying drawing.

Fig. 1 is the method flow diagram of the described thermal infrared imager of the embodiment of the present invention;

Fig. 2 be the described thermal infrared imager of the embodiment of the present invention do the structural representation of shutter coefficient without camera lens;

Fig. 3 be the described thermal infrared imager of the embodiment of the present invention do the structural representation of shutter coefficient with camera lens;

Fig. 4 is that common movement is to the uniform surface image;

Fig. 5 is that common movement is to target image;

Fig. 6 is to the uniform surface image without the shutter movement;

Fig. 7 is to target image without the shutter movement.

In figure:

1, high-low temperature chamber; 2, movement; 3, monitor; 4, computing machine; 5, power supply; 6, camera lens.

Embodiment

Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is clearly and completely described, obviously, described embodiment is only the present invention's part embodiment, rather than whole embodiment.Based on the embodiment in the present invention, those of ordinary skills belong to the scope of protection of the invention not making the every other embodiment that obtains under the creative work prerequisite.

Embodiment 1

As shown in Figure 2, the described thermal infrared imager without shutter based on the VOx detector of the embodiment of the present invention 1 comprises high-low temperature chamber 1, is provided with movement 2 in described high-low temperature chamber 1, and described movement 2 is connected with respectively monitor 3, computing machine 4 and power supply 5.

The described thermal infrared imager of employing embodiment 1 (as Fig. 2) is realized the collection without the shutter sample coefficient:

Give thermal imaging system programming program, then the connecting communication cable is placed on thermal infrared imager in high-low temperature chamber, and setting high-low temperature chamber is lowest temperature TL ℃, being incubated 2 hours allows thermal imaging system fully lower the temperature, power on to thermal imaging system, send the control command of doing the shutter coefficient by communication cable to thermal imaging system, thermal imaging system begins detected temperatures, in the time of temperature variation n ℃, gather present image, and exist in the FLASH of thermal imaging system inside, record simultaneously current accurate temperature value.Temperature stabilization and picture quality when guaranteeing that thermal imaging system gathers image must be controlled the programming rate of high-low temperature chamber, slowly are warmed up to highest temperature TH ℃,

For guaranteeing to collect the sample of maximum temperature point, when high-temperature T ℃, be incubated 2 hours, the temperature of assurance thermal imaging system itself rises to the highest, and like this, thermal infrared imager has collected M and has organized image in whole temperature section, record as follows:

TEMP[]＝{T0，T1，T2，……TM-1}；

TEMP_ADDR[]＝{TA0，TA1，TA2，……TAM-1}；

TEMP[] be the temperature value of the correspondence of M group image;

TEMP_ADDR[] be that M organizes image FLASH memory address corresponding to image.

After completing in the sample image collection of total temperature section, thermal imaging system just can not have under the condition of shutter, by obtaining correct images without the shutter algorithm.

Comparison diagram 4,5 and 6,7, that layer noise ratio on common movement is heavier, and be very light without noise on the shutter movement.

Embodiment 2

As shown in Figure 3, the described thermal infrared imager without shutter based on the VOx detector of the embodiment of the present invention 2 comprises high-low temperature chamber 1, is provided with movement 2 and camera lens 6 in described high-low temperature chamber 1, and described movement 2 is connected with respectively monitor 3, computing machine 4 and power supply 5.

The described thermal infrared imager of employing embodiment 2 (as Fig. 3) is realized the collection without the shutter sample coefficient: process is identical with embodiment 1.

The above is only preferred embodiment of the present invention, and is in order to limit the present invention, within the spirit and principles in the present invention not all, any modification of doing, is equal to replacement, improvement etc., within all should being included in protection scope of the present invention.

Claims (6)

1. thermal infrared imager without shutter based on the VOx detector, it is characterized in that: comprise high-low temperature chamber, be provided with movement in described high-low temperature chamber, described movement is connected with respectively monitor, computing machine and power supply.

2. the thermal infrared imager without shutter based on the VOx detector according to claim 1, is characterized in that: also comprise camera lens, be arranged on described high-low temperature chamber.

3. the described using method without the shutter thermal infrared imager based on the VOx detector of claim 1 or 2, is characterized in that, comprises the following steps:

1) original image acquisition step: program is write in infrared movement, switched on power, video, serial ports is placed in high-low temperature chamber, and the uniform planar in alignment box drops to lowest temperature T _L℃ degree is incubated 2 hours, then powers on to movement, sends acquisition to movement at once, and the automatic coefficient capture program in movement will start, and movement can gather point under Current Temperatures at once as the minimum temperature sample; Allow high-low temperature chamber slowly be warmed up to highest temperature T _H℃, in this process, movement can be according to the temperature that detects, at the sample image that gathers on one group of temperature spot under the detector different temperatures, and this image sets is existed in FLASH on movement, simultaneously the temperature value of record detector this moment;

2) background forecast step: having the image sets in FLASH more than using is the basis, utilizes prediction algorithm to obtain background image;

3) 1 and 2 point calibrations: be input parameter according to the temperature of current detector, obtaining the background forecast image is ImgBK, recycles and 2 point calibration formula at 1 infrared image is proofreaied and correct:

Formula 1 Im gAdj (i, j)=(Im gU (i, j)-Im gBK (i, j)) * Nuc (i, j),

Wherein ImgAdj is the image after proofreading and correct, the horizontal and vertical pixel coordinate of (i, j) presentation video matrix; ImgU is original image, and ImgBK is the background image of prediction, and Nuc is 2 point calibration parameters.

4. using method according to claim 3, is characterized in that, described original image acquisition step comprises:

Give thermal imaging system programming program, then the connecting communication cable is placed on thermal infrared imager in high-low temperature chamber, and setting high-low temperature chamber is lowest temperature T _L℃, being incubated 2 hours allows thermal imaging system fully lower the temperature, power on to thermal imaging system, send the control command of doing the shutter coefficient to thermal imaging system by communication cable, thermal imaging system begins detected temperatures, in the time of temperature variation n ℃, gathers present image, and exist in the FLASH of thermal imaging system inside, record simultaneously current accurate temperature value.Temperature stabilization and picture quality when guaranteeing that thermal imaging system gathers image must be controlled the programming rate of high-low temperature chamber, slowly are warmed up to highest temperature T _H℃;

For guaranteeing to collect the sample of maximum temperature point, at high temperature T _H℃ the time, be incubated 2 hours, the temperature that guarantees thermal imaging system itself rises to the highest, like this, thermal infrared imager has collected altogether M and has organized image in whole temperature section, record as follows:

TEMP[]＝{T ₀，T ₁，T ₂，.....T _M-1}；

TEMP_ADDR[]＝{TA ₀，TA ₁，TA ₂，......TA _M-1}；

TEMP[] be the temperature value of the correspondence of M group image;

TEMP_ADDR[] be that M organizes image FLASH memory address corresponding to image.

5. using method according to claim 3, is characterized in that, described background forecast step comprises:

1) thinking or the thought of two point calibrations:

Im?gAdj(i，j)＝(ImgU(i，j)-ImgBK(i，j))*Nuc(i，j)

2) two point calibration coefficient Nuc (i, j) obtain unanimously with our two traditional point calibration methods, gain correction coefficient obtains: image B lackH and the BlackL of the homogeneous radiation body under the different radiation intensity of two width under same ambient temperature conditions;

Formula 2 $\mathrm{Nuc}(i,j)=\frac{\mathrm{Mean}\left(\mathrm{BlackH}\right)-\mathrm{Nean}\left(\mathrm{BlackL}\right)}{\mathrm{BlackH}(i,j)-\mathrm{Black}(i,j)}$

3) background image ImgBK (i, j) is the background of preserving in thermal imaging system, obtains new background image NImgBK in conjunction with actual image correction of obtaining;

Formula 3 NIm gBK=K*Im gBK+B

4) after image adopts existing gain coefficient Nuc and biasing coefficient ImgBK to carry out two point calibrations, obtain image I m gAdj, in image I m gAdj, search gray scale difference value within facing the territory scope less than the pixel of certain threshold value pair, will satisfy all pixels of this threshold condition to search out; The computation process of adjusted coefficient K and B is:

The output difference of neighbor:

Formula 4 is replaced partial parameters for simplifying expression formula:

E＝Nuc(a)(Im?gU(a)-K*Im?gBK(a)-B)

-Nuc(b)*(Im?gU(b)-K*Im?gBK(b)-B)

E＝Nuc(a)*Im?gU(a)-Nuc(b)*Im?gU(b)

-K*(Nuc(a)*Im?gBK(a)

-Nuc(b)*Im?gBK(b))

-B*(Gain(a)-Gain(b))

A1＝Nuc(a)*ImgU(a)-Nuc(b)*ImgU(b)

Formula 5 A2=Nuc (a) * ImgBK (a)-Nuc (b) * ImgBK (b)

A3＝Nuc(a)*ImgBK(b)

Obtain respectively:

$K=\frac{\mathrm{\&Sigma;}(A1*A2)*\mathrm{\&Sigma;}{A3}^2-\mathrm{\&Sigma;}(A1*A3)*\mathrm{\&Sigma;}(A2*A3)}{\mathrm{\&Sigma;}{A2}^2*\mathrm{\&Sigma;}(A2*A3)*\mathrm{\&Sigma;}(A2*A3)}$

Formula 6 $B=\frac{\mathrm{\&Sigma;}(A1*A3)*\mathrm{\&Sigma;}{A3}^2-\mathrm{\&Sigma;}(A1*A2)*\mathrm{\&Sigma;}(A2*A3)}{\mathrm{\&Sigma;}{A2}^2*\mathrm{\&Sigma;}{A3}^2-\mathrm{\&Sigma;}(A2*A3)*\mathrm{\&Sigma;}(A2*A3)}$

Obtain real-time being biased to:

Formula 7 NIm gBK=M*Im gBK+C

Obtaining real-time NIm gAdj result is:

Formula 8 N Im gAdj=(Im gU (i, j)-NIm gBK (i, j)) * Nuc (i, j)

6. using method according to claim 3, is characterized in that, described background forecast step is switched based on many backgrounds and the background correction is carried out in two background weighting.Comprise: above formula is based on algorithm idea and the calculation procedure of single background modification method, but in the reality test, can find single background Shortcomings.When the focal plane temperature differed greatly with the background focal plane of preserving, this background can't obtain revising out more suitable real-time background, therefore needed to adopt the background of a plurality of different focal planes temperature.But switch background when the focal plane temperature changes, but thisly directly cut the problem that background can produce image flicker.Therefore to consider to adopt the background of two background modification methods.Weighting coefficient according to realtime graphic and two backgrounds of two background calculating obtains real-time background.

To gather 4 different focal planes temperature backgrounds as example.When the real time temperature scope was between temperature 1 and temperature 2, the background when background is by temperature 1 and temperature 2 in real time is weighted to be obtained.Real time temperature is the closer to which background temperature, and the weight that current background accounts in the background of obtaining is just large.During near background temperature 2, at this moment the weight of temperature 2 is substantially near temperature 1 when real-time focal plane temperature.When temperature continuation rising, real time temperature reaches between background temperature 2 and background temperature 3, adopts background temperature 2 and background temperature 3 to be weighted.Due to temperature rising rising, when temperature during just greater than background temperature 2, at this moment or take background temperature 2 as major weight.When at this moment not existing background to switch the problem of image flicker can not appear.This thought method is proof in slave computer is transplanted.

Algorithm idea is the same with single background modification method of front, just adopted the background at two different focal planes temperature, calculates the weighting coefficient of two backgrounds according to realtime graphic and two backgrounds, obtains real-time background

Formula 9 Noffset=K1*Im gBK1+K2*Im gBK2+B

The computation process of weighting coefficient K1 and K2 and B is as follows:

With single background algorithm idea, according to two backgrounds when the selection of front focal plane temperature and this temperature contrast minimum, select one of them background and gain coefficient, obtain image NIm gAdj data, search is being faced gray scale difference value in the scope of territory less than the pixel of certain threshold value pair in image NIm gAdj, will satisfy all pixels of this threshold condition to search out.

Adjusted coefficient K 1, the computation process of K2 and B is:

Formula 10

E＝Nuc(a)(ImgU(a)-K1*ImgBK1(a)-K2*ImgBK2(a)-C)-

Nuc(b)*(X(b)-K1*ImgBK1(b)-K2*ImgBK2(b)-C)

E＝Nuc(a)*X(a)-Nuc(b)*X(b)

-K1*(Nuc(a)*Im?gBK1(a)-Nuc(b)*Im?gBK1(b))

-K2*(Nuc(a)*Im?gBK2(a)*Nuc(b)*Im?gBK2(b))

-C*(Nuc(a)-Nuc(b))

For simplifying expression formula, partial parameters is replaced:

Formula 11

A1＝Nuc(a)*X(a)-Nuc(b)*X(b)

A2＝Nuc(a)*Im?gBK1(a)-Nuc(b)*Im?gBK1(b)

A3＝Nuc(a)*Im?gBK2(a)-Nuc(b)*Im?gBK2(b)

A4＝Nuc(a)-Nuc(b)

Order

Formula 12

$T1=\mathrm{\&Sigma;}{A2}^2-\frac{\mathrm{\&Sigma;}(A2*A4)*\mathrm{\&Sigma;}(A2*A4)}{\mathrm{\&Sigma;}{A4}^2}$

$T2=\mathrm{\&Sigma;}(A2*A3)-\frac{\mathrm{\&Sigma;}(A3*A4)*\mathrm{\&Sigma;}(A2*A4)}{\mathrm{\&Sigma;}{A4}^2}$

$T2=\mathrm{\&Sigma;}(A1*A2)-\frac{\mathrm{\&Sigma;}(A1*A4)*\mathrm{\&Sigma;}(A2*A4)}{\mathrm{\&Sigma;}{A4}^2}$

Order

Formula 13

$D1=\mathrm{\&Sigma;}(A2*A3)-\frac{\mathrm{\&Sigma;}(A2*A4)*\mathrm{\&Sigma;}(A3*A4)}{\mathrm{\&Sigma;}{A4}^2}$

$D2=\mathrm{\&Sigma;}{A3}^2-\frac{\mathrm{\&Sigma;}(A3*A4)*\mathrm{\&Sigma;}(A3*A4)}{\mathrm{\&Sigma;}{A4}^2}$

$D3=\mathrm{\&Sigma;}(A1*A3)-\frac{\mathrm{\&Sigma;}(A1*A4)*\mathrm{\&Sigma;}(A3*A4)}{\mathrm{\&Sigma;}{A4}^2}$

Obtain

Formula 14

$K1=\frac{T3*D2-D3*T2}{T1*D2-T2*D1}$

$K2=\frac{T3*D1-D3*T1}{T2*D1-T1*D2}$

$B=\frac{\mathrm{\&Sigma;}(A1*A4)}{\mathrm{\&Sigma;}{A4}^2}-K1*\frac{\mathrm{\&Sigma;}(A2*A4)}{\mathrm{\&Sigma;}{A4}^2}-K2*\frac{\mathrm{\&Sigma;}(A3*A4)}{\mathrm{\&Sigma;}{A4}^2}$

Obtain real-time being biased to:

Formula 15 Noffset=K1*Im gBK1+K2*Im gBK2+B

Obtaining realtime graphic NIm gAdj result is:

Formula 16 NIm gAdj=(ImgU (i, j)-NIm gBK (i, j)) * Nuc (i, j).

Images