Optical Imaging - Science method

I can suggest the following approach.

1. Cells detection using instance segmentation or just detection. There are a lot of approaches.

2. Track cells. The simplest approach is SORT track algorithm.

3. Analyze a cell (each track)

3.1 Align cell crops using circle detection, build "video" for each cell

3.2 Calculate optical flow

3.3 Convert the optical flow from cartesian to circular coordinates

3.4 The <mean> amplitude of result vector field is rotation speed.

Probably, you will need to crop this field or do some additional transformation in another steps.

Thank you

As nanowires are narrow, completely intransparent objects, physical effects like diffraction, dispersion and chromatic abberation of the microscope lenses, as well as reflection of ambient light should be taken into account.

Seonggeun Hong

I think 5.7c is more suitable.

In work from attached file there are CCT diagrams for AISI 4140, so you can make a rough estimation of what structures have been received in your case.

Once again - it is neseccary to know values of cooling rate and what kind of thermal treatment was carry out.

Dear Mahyar Radak ,

You can create a contour plot of the data in a grayscale image using imcontour . This function is similar to the contour function in MATLAB®, but it automatically sets up the axes so their orientation and aspect ratio match the image. To label the levels of the contours, use the clabel function.

https://www.mathworks.com/help/images/contour-plot-of-image-data.html#:~:text=You%20can%20create%20a%20contour,contours%2C%20use%20the%20clabel%20function.

Regards,

Shafagat

I think your premise is flawed. There isn’t going to be “an answer” because tailoring this parameter and trading it against other properties you might like is the crux of coating design and the answer might be anything over a wide range depending on what was designed under what set of constraints. For example, your example of a mirror being 4% is at best a rule of thumb and most often completely wrong. Over a fair range of wavelengths bare i coated aluminum happens to be around 4% absorptive. However Silver is only 2% absorptive in that range. Bare Gold may be terribly absorptive at shorter wavelengths. At longer wavelengths it doesn’t reach the 98% of silver, but over much of the IR and aluminum become terribly absorptive and gold is the best. More importantly, mirrors are rarely uncoated and a dielectric coating can raise the reflectivity of metallic mirrors above 99%. See for example Edmunds “ultrafast” enhanced aluminum coating.

And that is just metallic coatings. Metal is useful over a wide wavelength range when you don’t know what a mirror is going to be used for. However, if you know the wavelength (and what acceptance angle you need, and other constraints) you can use a pure dielectric stack. Dielectric mirrors can be made very close to 100% reflective. What’s more, very little light is absorbed, so what little doesn’t reflect transmits.

That brings us to beam splitters. It is not at all difficult to make a dielectric coating where essentially no light is absorbed. It is all either transmitted or reflected. Adding the reflected to the transmitted should yield just about 100% every time. When you found placed where they appeared to add up to higher than 100%, that is just experimental error or round off error, but they probably do add up to almost 100%

Tora Båtvik You are correct, the Level 1O falls between L1R and L1G. I didn't noticed that there was updated version of the KOMPSAT-3 Image Data Manual. The V 2.1 provides more detail about KOMPSAT-3 data and pre-processing operations applied to compensate for sensor radiometric and geometric artifacts.

According to the Manual V 2.1 "Level 1O is a processing level so called ortho ready. It is the product corrected for geometric distortions and projected to UTM. Processing for Level 1O includes all radiometric corrections and sensor corrections applied to Level 1R processing. Optical distortions are corrected and terrain effects are corrected using average height over the region...".

You may access the KOMPSAT-3 Image Data Manual V 2.1 from here;

In addition, three Enhanced Processing Levels for KOMPSAT 3 image data: Level 1R DZ, Level 1ODZ and Level 1G DZ are also offered.

Please note that , initial sensor-recorded signals are calibrated to radiance values using gains and offsets and then 'rescaled' to DNs. DN commonly used to describe pixel values that have not yet been calibrated into physically meaningful units. The three levels for KOMPSAT-3 image data: Level 1R, Level 1O and Level 1G underwent 'basic processing'. As of L1R, the Level 1O Product is corrected for radiometric and sensor distortions. The difference of relative radiometric response between detectors is corrected and internal detector geometry and mis-registrations between detectors are corrected when applicable.

So, it is safe to conclude that your data has been stored as DN (as 14 bits/pixel) after applying aforementioned radiometric, sensor and geometric corrections. Now you are supposed to convert the these DNs to TOA Radiance and then TOA reflectance following instructions provided on Pg. 17 of the manual. You may use ENVI to do this job for you.

I agree with William Thoma that it should be able to pick up the Cherenkov signal. You may need longer exposure (5 min) and/or pixel binning. Make sure to remove any optical filters as the Cherenkov signal is broadband.

Dear Quilang Ye

Although I don't know the details of your SLM, often ghosts can appear when the incident polarization is not orientated properly with respect to the direction of the liquid crystal's alignment. If this is your problem, one way to solve it is to use a polarizer to clean up your incident polarization, a waveplate to rotate the transmitted polarization to the proper orientation for the SLM, placing the reflective beam splitter between the waveplate and the SLM to couple a portion of the modulated light out. Also, be aware that the coatings on beamsplitters (both on the reflective side and the antireflective side) usually have different reflection coefficients for s and p polarizations.

Hi Stefan,

I'm the CEO of Cellvis, thank you for your interest in our products.

You can view the plate tech spec and dimension of the 96 well plate you mentioned at:

https://www.cellvis.com/_96-well-glass-bottom-plate-with-high-performance-number-1.5-cover-glass_/product_detail.php?product_id=50#tech-spec

Whole plate flatness variation is within 0.1mm.

Coverglass is suitable for tissue culture, however usually cell attachment is weaker compared with tissue culture treated polystyrene dish/plate. For some cells coating is needed.

We have 96 well plate with #1.5P coverslip. It has optical characteristics similar to glass, with cell attachment comparable to tissue culture treated polystyrene dish/plate:

https://www.cellvis.com/_96-well-plate-with--number-1.5-glass-like-polymer-coverslip-bottom-tissue-culture-treated-for-better-cell-attachment-than-cover-glass_/product_detail.php?product_id=69

I thank this is very important work .

I hope for you a nice day

Qusay Kh. Al_Dulamey

Isn't that a question of satisfying the sampling condition? If you imagine the speckle pattern is formed from a range of plain waves (angular), then the highest frequency component is resulted from the interference between the two outermost, in angular term, plain wave components. This will set the limit of the sampling spacing you need to achieve, taking into account if you are dealing with intensity rather than amplitude.

Here a 2019 paper about this:

Dear William,

Thank you for your response.

Actually the light source that we simulate is a point source light that has spectral width of 50 nm with peak around 400 nm and the detector is photo multiplier tube.

Bests

Suggestions?

Highly depending on the original image/scene. Knowing nothing about that, nothing substantial can be proposed.

Do you want to move that position as time goes on to accommodate changing plant size, or one position for the season? Changing positions allows more accuracy at each size, especially if you have some known objects in the mix to standardize from... In our case, we are looking for pathogens, etc, so moving the cameras makes sense as we strive go image the areas we need to evaluate... this then requires a one size fits one system, which is a hassle to set up compared to a one size fits all...

ImageJ is a Java-based image processing program with a lot of useful plug-ins, where the one or the other might fulfill your requirements. However, you did not mention what exactly are your individual tasks in these images. I have experience with Matlab and Python. If you have a licence for Matlab you will have less stress with libriaries and dependencies. If you have no licence, Python is reasonable too and you can install and work wherever you want. If you need algorithms using artificial intelligence (object detection, automatic (semantic) segmentation, in Python there is probably more code and repositories available. I would only recommend ImageJ or other standalone (bio) applications (e.g. Icy - Bioimage analysis) if you never want a program a line of code.

“Faint” particles presumably means low signal-to-noise ratio (SNR) for tracking, which means that there will always be some uncertainty about track assignments for close-passing particles.

Two possible solutions are:

Perfect tracking requires perfect information (high SNR and/or faultless heuristics). Your track assignments will always be to uncertain, and therefore subject to errors, while SNR remains low and heuristics are incomplete or imperfect.

Do not fall into the common trap of endlessly revising/tweaking tracking algorithms to deal with yet another error you happened to notice, because then your work will churn on endlessly, with no stopping point, and without confidence in your results.

The way ahead with confidence is to squarely address the question: “How do I use an automatic tracking algorithm that I know is going to make mistakes?”

It has to be addressed statistically and practically:

Your goal is presumably not to develop perfect tracking in low SNR conditions. Your goal is rather to address some other more interesting aspects of particle motion or dynamics. The fact that you see errors occurring on rare occasions (assuming they are rare) does not necessarily mean that you have to fix your tracking algorithm. If particle crossings are rare, then your current algorithm may already be at acceptable operation with your current algorithm.

One can imagine situations, for instance, where simply discarding particle tracks that have crossed from further analysis is a serviceable strategy for handling crossings. All you need is a crossing detector. This may or may not apply in your case.

Or you might flag all crossings for visual inspection by you in a manual stage of processing, in which you fix any errors using your expert judgement about each crossing event (assuming that you are a perfectly reliable judge, or at least a better judge that the algorithm, of track continuation).

The points to be made in any case are that, given finite SNR and imperfect heuristics, you generally need to accept that automatic tracking errors will occur, and you need to show that their ultimate negative influence on the objectives of your project is acceptably low and suitably managed.

"... in the optically cleared sample there must be no light scattering"

This is not true.

Even in pure liquids, light is scattered due to density fluctuations (ie. compressibility).

Dear Neema,

I suggest you see these links and attached file:

There is a window of transparancy- from ~1.2 to 2 microns - i which the wavelengths are too long for electronic absoption and too short for bond vibrations. This region is very popular in Optical Coherence Tomography. Attenuation in this region is mostly due to scattering, and is usually describedas photon diffusion. Be careful, however, since in a lot of cases the so-called "diffusion model" employed in random-walk simulation is applicable, but the corrersponding "diffusion approximation", which employs paraboloc equations of diffusion, is not.

Dear Anup, please note that a SOLUTION being sensititive to the depth does not mean that the focal mechanism of the actual earthquake depends on the depth. If one uses waveform inversion to obtain a focal mechanism, one will get a solution that is independent of any prior determination of the depth and representative of in situ conditions.  

You can insert an optical fiber into the tissue and measure the attenuation as a function of the distance... Of course you will damage the tissue, but in ex vivo samples you can try to do it.

Refer to PIV. As the others have mentioned before, use the cylindrical lens. The details of its application can be easily found in the frame of PIV.

Not at all.

I'am waiting you'r remarks. 

If you are interested in ground deformation, use HH images in StripSAR (they could be FBD or FBS, does not matter). 

Thank you

This is absolutely true, what i ve tried to say is that lensless imaging for certain cases might allow such a high NA registration when pixel size would matter. When it usually doesnt if a common type glass objectives are used. So conciderations about sampling conditions might not be important here.

Mihail,    Hot spot is harmful, when you want perfect uniformity.  In my experience of miscrospcope image photos, they have been always accompanied with this.  So, I recommend you diffuser dome.  It should be as uniform as in integrating sphere in principle.  But your intension is relatively uniform illumination, Thorlabs ring is not bad.   Regards, Shigeo

If you want to measure OD with ImageJ first you have to produce first a calibration image (an 8-bit grayscale or color image with defined areas, the mean gray value of each corresponding to a known OD value under your experimental conditions). Once you have it, load the calibration image into ImageJ and measure the mean gray value of each of its areas using an appropriate area selection tool. Then, go to Analyze/Calibrate, enter "OD" in the measurement unit box, and in the two boxes below enter the mean gray values in the left and their corresponding, known ODs in the right. Select a best-fitting equation, tick the "Global calibration" box, and press OK (if you are not satisfied with the chosen equation, you can select a different one by going to Analyze/Calibrate again, no need to enter grayscale-OD value pairs all over).

Once you have fittted the OD values satisfactorily to a curve, load the image to be analyzed and measure whatever it is that you want to measure. Main, min and max gray values will now correspond to optical densities.

Hope it helps!

G655 fibre typically has very slightly slower group velocity than G652 (effective index of Corning LEAF is 1.4693, compared with 1.4679 for SMF-28e+).  However, if the G652 link requires one or more dispersion compensation modules, each fibre DCM will introduce additional latency, and the G655 solution should perform better in this respect.

Does your application require fibre-based compensation?  Bragg grating dispersion compensation will introduce significantly less latency than an equivalent fibre compensator.

How much control do you have over cable manufacture and installation?  An alternative to discrete DCMs is to incorporate compensating fibre with negative dispersion into the transmission cable so that the total length of fibre is unchanged and there is minimal impact on latency.

Other options to improve dispersion tolerance without appreciable impact on latency include use of lower symbol rates.  For example via multiple WDM channels, or multi-level modulation formats such as differential QPSK, dual polarisation QPSK or higher level QAM.

Using QAM with coherent detection, high levels of dispersion can be compensated electronically or by digital signal processing.  This can introduce some additional latency, but potentially much less than a fibre-based compensation module.

Note that the lower dispersion of G655 fibre increases susceptibility to both Kerr-effect nonlinearity and to stimulated Raman crosstalk.  This may require a reduction in the per-channel launch power, particularly in dense WDM systems.  Fibre effective area is typically lower in G655, and this degrades the non-linear crosstalk further.

Hello sir,

How r u...

I would like to tell you that I don't hv enough knowledge about this topic but I told my coalgees and senior about that ?. Whenever I will get the solution for question  I will inform you as soon as possible....

Regards 

Ajay yadav

Thank you so much Ronald. 

Laser range finding for other vessels near the surface of the sea might run into a bit of an issue: you need it most when visibility (e.g., light transmission in the air) is the worst.  My understanding is that most laser range finding systems rely on ballistic (singly scattered or reflected) photons.  On a clear night, laser range finding would do the job.  Any time when there is significant spray in the air (windy conditions, storms) the laser beam would be heavily scattered and therefore be less useful.

Importantly, for vessel detection, you don't need a lot of spatial precision- is a meter accuracy in your maps good enough?  10m?  My inclination would be to think about a modality which is not impacted by near-surface water conditions (above or below).

There is an introductory article on ocean spray in the context of weather and climate in Physics Today, Nov. 2016.  Figure 2 showed the common drop sizes, which are in a range to heavily scatter most commonly available lasers. 

regards,

David

The classic way to study this problem is to use interference reflection microscopy to measure the distance between the cell membrane and the underlying substrate (5 to 100nm distance range). There is an extensive body of literature on the subject, and there is broad consensus that the plasma membrane is NOT uniformly pancaked against the substratum. So you cannot conceive of a single distance measurement that describes membrane-to-substrate distance.

Instead, cells contact the substrate at focal adhesions (5-10nm from the glass surface), and the plasma membrane between these adhesion sites assumes an undulating pattern with variable distances between it and a 2D extracellular matrix. Electron microscopy of lamellipodial structures also show strong adhesion near the base of the membrane at the leading edge, but the protruding/extending membrane can tower a fair distance above the substratum. So 200nm might be plausible, depending on the cell type and the region of plasma membrane under consideration.

If your goal is simply to image the ventral (substrate-facing) plasma membrane, then TIRF imaging modality with a fluorescently tagged molecule is your best bet. Confocal would be the next choice if you can't do TIRF.

From wikipedia: "The cross section is an effective area that quantifies the essential likelihood of a scattering event when an incident beam strikes a target object, made of separate particles."

In other words,be careful when you define transmission as forward scattered light. I would rather say that light is either scattered, absorbed or transmitted. Extinction is defined as the sum of scattering and absorption,

C_ext = C_abs + C_sca,

in scattering cross sections. 

Also, the Stokes parameters describe the light, not scattering by the particle. You will have a set of Stokes parameters for your incident light, and a set of Stokes parameters for the scattered light. Note that the S₀ is really the light intensity! I assume your modulation system describes the relation between the two sets of Stokes parameters i.e. the scattering by your particle (see Mueller Calculus).

Now, an easy way to find the scattering cross section would be to look at the intensities of the incident light and the scattered light:

C_sca = I_sca/I_inc

Just be sure to integrate the scattered intensity over all angles.

I am only want to clear this concept. 

VV. VH, HH and HV interacts with a single tree one by one that is heighten from surface then what will be response ? 

Search for papers experimentally comparing the two ways of imaging and, if you don't find good publications, cary out the experiments with the compositions you've peaked in your lab - it will be a good paper.

One point maybe still has to be mentioned - oxides are not perfect hosts for upconversion. Although there is some logic behind a direct comparison of yttria-based down-shifting and up-conversion, it would be more relevant if you compare not the same host with different doping but "best in class" compositions for both types of emission. You will need something like sodium yttrium fluoride as a host for upconversion (to be put against an optimised oxide-based material exhibiting Stokes emission).

High end microscope objective lenses, like those found on most confocals, are designed to image through a number 1.5 glass coverslip.  You can image through plastic but be aware the image quality will be worse.  Ibidi makes plastic bottom dishes that are more similar to glass. They are designed for high end microscopy applications.

If you use standard plastic, as long as the objective lens has enough working distance to get through the thickness of the plastic, then it can be done.  Long WD lenses are typically dry, low magnification lenses.  Look for the WD marked on the lens (at least it is for Nikon optics). 

in this paper the researchers measure some optical properties of different types of hydroxyapatite. Maybe it helps.

 You should consider the distance from the center of a dark area to the center of the next dark area thats means 4.38um.

Hi Muhammed,

I assume you mean HR coating if you want to obtain a 99.93% reflection on the back mirror? I don't know which wavelength you're aiming at but these HR coatings normally consist of combinations of high and low index coatings (e.g. SiO2-Si3N4, SiO2-Ta2O5). There is good software to calculate these coatings for your specific wavelength (TF_calc, Filmstar, McLeod). Very often the commercial companies who deliver you the mirrors will use this software if you give them the specs (wavelength, reflectance, polarisation, etc). For a 99.93% reflectance my guess is you will need 5-6 stacks of the material combinations I mentioned. For AR coatings the same software is used and there you probably need a single thin film (e.g SiO2) or a single combination of SiO2 and Ta2O5, depending on central wavelength and wavelength window. 

Dear Shubhda Srivastava,

You might had use Ar during annealing steps before the growth by keeping Cu foils directly on a quartz boat (tube). The uneven surface indicates the presence of oxygen in the reaction chamber. You can try once by flowing H₂ for few min(eg. 10min) after you checked your proper vacuum to ensure there is no more Oxygen. After that, you can replace H₂ by Ar. You can check the purity of Ar gas used (what % of oxygen ppm).

We had the similar problem for both APCVD and LPCVD at 1050^oc. But that technique effectively worked for us. Even on that surface we succeed to grow graphene crystals but difficult to transfer.

Maybe this technique helpful to you too.

I think, it is important to read this file 

Hello Hong,

in order to perform any automatic correction, you need two things: a image metric, and an optimization algorithm. In the case of focusing, a good metric can be the edge contrast, often referred to as acutance (https://en.wikipedia.org/wiki/Acutance), which is, in simple words, the mean square difference in intensity between adjacent pixels. Other metrics could be the variance of the image, or the high spatial frequency content of the image, computed through its fourier transform.

Once you have your metric, which is maximum (or minimum) when the image is on focus, you need an optimization algorithm. If you are not in a hurry, you can simply scan through the focus, compute the metric for many positions, and fit the curve with a polynomial (usually a parabola) to find the maximum. In alternative you can use more advanced algorithms such as stochastic gradient descent, or simplex optimization, in order to find your minimum in less iterations.

I hope this was useful, good luck with your experiment!

Dear Dr. Sharma,

You can use metallographic software such as Dewinter Material Plus to analyse the porosity through ASTM B276. Alternatively, you can also analyse the porosity in by ImageJ software tool. ImageJ is available freely (follow the link below).

Please see the attachment for help on ImageJ.

I hope it helps

With Best Regards,

Sunny Zafar

Apologies Paul, I miss read your question. 

If you draw a ray diagram for your device, given a known the size of your proximal mirror, which I imagine will be of order of the size of an eye piece, you should be able to calculate the appropriate size of your distal mirror. 

If you want to generlize your simulation to any geometry you can check up Nirfast an open source software

Hi Alba, answers to your questions. 1. we use a different dye, RH482, (NK3630), which is an absorption dye. The good thing about absorption dye is that it is not excited by the light so is not phtotoxic. Another good thing about NK3630 is that it works in near infrared, 705-720nm.  In contrast,  Di-4-ANEPPS is a fluorescent dye excitation 520nm emission 610 and longer. It has larger signals but the recording time is shorter due to bleaching and phototoxicity. 2. We incubate 2 hrs in diluted dye solution. In order to incubate slice in a long time you will need to circulate the ACSF to keep the slice viable. 3. all dyes work on exact wavelength. RH482 works on 705 and longer, RH1691 works on excitation 630 and emission >690; Di-4-ANEPPS works on excitation 520 and emission > 610. you can get LED light to work. Laser sometimes are noisier then LED. We had a method paper for RH482 :

Jin, W., Zhang, R.J., and Wu, J.Y. (2003). Voltage-sensitive dye imaging of population neuronal activity in cortical tissue. J Neurosci Methods 115, 13-27.

A method paper for in vivo RH1691:

Lippert, M.T., Takagaki, K., Xu, W., Huang, X., and Wu, J.Y. (2007). Methods for voltage-sensitive dye imaging of rat cortical activity with high signal-to-noise ratio. J Neurophysiol 98, 502-512.    

Yes, it should be possible - if you have 3 independent components of observation you can derive the full 3D displacement. You can first write out the InSAR LOS unit vector with its EW, NS, and vertical components, and then solve for the vertical component in terms of the others. Insert the LOS, EW and NS observations into your equation, and you can get the vertical displacement. Note that all the LOS unit vector coefficients are spatially dependent, so you will have to do it for each pixel separately. If you have even more observations, then you can do it in a least squares sense to help minimize the error.

@Farouk,This not just by perform only in FEM also in FDTD, but the real problem is that how i can define refractive index of the material over a range of wavelength, i don't know material??? 

Have a look at TrakEM2, which is a plugin in ImageJ (Fiji):

In my opinion the best tool for image alignment at the moment (elastic alignment) and manual segmentation.

SAR cares about imaging geometry, and not the vehicle/platform on which the radar is located...  Once you have an image, you can't tell what the vehicle was from which the data was collected...  So, no difference...

Use mCT analyses. If you want, you can send sample to our research centre (I belive you have possibilities of mCT in your university). You can obtain 3D information. The small round artefacts are metalurgical pores typical for small laser speed (high energy). Bigger one, shape similar to keyhole are typical pro higher speed of laser and unsufficient hatch strategy.

You can use a diffractive optical element, but that still acts as a lens. So I'm assuming you are wanting to minimize loses.  You can design some sort of all reflective system that can provide collimated light and also combine the beams.  This system could be done in a way it could be molded from plastics as well.  Keep in mind that you're going to need aspheric off-axis elements to design this system.  You're not likely to easily find what you want in a catalog.  Systems like this are typically designed and built as specialty systems.  If you're wanting to do this on a lab bench, it is possible, but will require many more elements.  

Is it a miicroscopic image? How much resolutions you need? Nowadays digital cameras have a very wide range of selections. Finalising one depends on many factors like the exact use, fund available etc. It is also not clear what sort of images are you looking for...............

 which laser excitation wavelength are you using?

is the objective coverslip corrected (0.17)?

If you are using 532nm, glass is no problem (and objective corrected)

For the 785 nm, use quartz or CAF2 coverslips also.

     

Nishidate et. al has reported a multi-spectral diffuse reflectance imaging method based on a single snap shot of Red-Green-Blue images for estimating melanin concentration, blood concentration, and oxygen saturation in human skin tissue. Multiple regression analysis of the absorbance spectrum and the extinction coefficients of melanin, oxygenated hemoglobin and deoxygenated hemoglobin provides concentrations of melanin and total blood;

Sensors 2013, 13(6), 7902-7915; doi:10.3390/s130607902

this file can help you

Image data compression is concerned with minimizing the number of bits required to represent an image. Image compression algorithms aim to remove redundancy i.e. repetition of messages to reduce the probability of errors in transmission, in data in a way which makes image reconstruction possible. Image compression algorithms calculate which data needs to be kept in order to reconstruct the original image and therefore which data can be "thrown away". By removing the redundant data, the image can be represented in a smaller number of bits, and hence can be compressed.

Hi Bjarke.

Have you looked in 'Fundamentals of Light Microscopy and Electronic Imaging' by Douglas B. Murphy and Michael W. Davidson, 2nd Edition, Wiley-Blackwell 2013 [ISBN 978-0-471-69214-0]; ?    E.g., pp. 9, 40 & 220.

Best wishes,

Richard

I've built one before with a qimaging camera for $12,000. Now that the Basler Ace cameras are out, I'm going to try an acA1300-30um Monochrome. USB3 interface. $695 from Edmund Optics. Use free ephus software from Svoboda lab that runs in matlab for the frame triggering, data acquisition and data processing. Should have the whole rig built for under $3k. Will post a build to my blog once its done. Cheers.

I think -.Georeferencing, preprocessing, and other basic stuff can be done using ENVI

If you want to do more analysis like interferogram etc ..Yes, it can be done however, you have to load SARscape module in your ENVI.

I believe the answer to your question is: no.

The reason is that when you try to do cloaking (eg. using metamaterials) all the techniques are narrowband. To get true invisibility you need your technique to work over a wide band. I there are no good wideband techniques.

Strictly speaking, focal point of a lens is defined as the point where plane wave is collected into a point, in the geometrical optics approximation. So, focal point of a lens is a constant. If lens is used for transformation of a wave front different from plane wave, than that wave front can be collected, in general, not to a point, but to an area with nonzero size. The position of minimum size occupied by the light may not coincide with the focal plane.

I'm not sure I understand the sentence about "sub-wavelength technique in optical fiber for communication." Perhaps, you meant how one can actually beat the diffraction limit or achieve better than (roughly) 200nm resolution? Let's talk nano-optics then.

Nano-optics deals with optical phenomena on nanometer scale, that is beyond the diffraction limit of light. The spatial confinement that can be achieved for photons is inversely proportional to the spread of wavevector components in the corresponding spatial direction. Such spread occurs in light field that converges towards the focus (behind the lens). Thus, one can obtain some finite spread in the wavevector components for highest possible NA that can be obtained with oil-based microscope objectives. If one would be able to to achieve a wider spread of wavevectors, then light could be focused to the spot smaller than the one dictated by the diffraction limit and one can achieve sub-wavelength resolution...

In 1928 Synge published an article that outlined the concepts of what is currently known as scanning near-field optical microscopy. He stated if one takes an opaque film with a sub-wavelength aperture in it and places it within a sub-wavelength distance from a sample, then the produced illumination spot will not be limited by diffraction but the size of the aperture!

In modern days this concept is the basis of the near-field scanning optical microscopy (NSOM) or scanning near-field optical microscopy (SNOM). Both abbreviations have been used. The tip can be made in several ways: metal coated tapered fiber that has an aperture 10-100 nm (achieved by so called shadow coating evaporation technique) or micro-fabricated probes (for example, based on standard AFM cantilever technology). These are just examples because the technology is evolving and so do the methods of making probes with sub-wave apertures.

So, in the nutshell this is a combination of AFM with optical probing. One brings such a probe to a close contact with a sample and scans it across the sample. Collection is done in far-field with a diffraction limited optics. As in optical microscopy, various modes of operation are possible: transmission, reflection etc. One can add non-linear effects as contrast mechanism (SHG, for example), fluorescence etc.

To get more info, do a literature search for NSOM, SNOM and you end up with excellent reviews and original articles on the subject. Alternatively, you may look at some books on the topic like L. Novotny and B. Hecht "Principles of Nano-Optics", Cambridge University Press (2006)

Hope it helps!

The main advantages of using a ‘liquid crystal tunable filter’ (LCTF) for a multi-modal imaging system are: wide wavelength (tuning) range (from the visible to the near-infrared, NIR), large apertures, high switching speed, wide field of view and two-dimensional imaging.

Well, it is not my field of research, but I would say yes to the first question, and no to the second:

The state of a quantum system is described by its quantum state function. These functions involve all the degrees of freedom, so grows in complexity with the number of photons, becoming useless. Nevertheless, the light is almost always incoherent, and photons behave independently. It is true that lasers generate coherent radiation, but generally no entanglement, except in a few cases (e.g. two-photon sources). Two photons definitely not form images, so, in realistic cases, the intensity should be proportional to photon count on a detector.

The quantum description is useful for systems with a few photons. The double slit experiment can be understood building the quantum state function with even one photon in the quantum regime, but diffraction is described by the Huyggens principle. I have found this text about the link between both representations:

I won't talk about single photon detectors or extremely weak sources, let's wait for a true expert. It seems that the kind of representation you need may exist, but restricted to some situations. I would also like to know it.

Regards, Nasser