Baylab

The network DVR (actually any kind of DVR) IS a computer (for some kind of DVR, it IS a PC). No extra PC is needed for your application.

imatest is a good choice the evaluate the quality of image. there are some free software can do similar job. ImageJ, developed by people in NIH, should be the most suitable software to do this. http://rsbweb.nih.gov/ij/download.html of course, you need another Plug-in called SE MTF, which is free too. http://rsbweb.nih.gov/ij/plugins/se-mtf/index.html

I think so

Actually, even two sensors have same pixel number and same pixel pitch, and used with same lens with same focusing, you still can get different result. for the high resolution sensor, especially the one with small pixel, the crosstalk between the neighboring pixel is quite significant (light will leak from one pixel to another pixel). such kind of crosstalk will degrade the image quality definitely. I don't think so... Crosstalk between pixels "possible" in over lighted CCD, but not in CMOS. Different real resolution, but not image size in pixels caused more optical and little bit compression effects... That phenomena happen in CCD is called "blooming", it is another story ( overflow of electric charge VS light leakage) . there is no totally opaque wall between neighboring pixels (so researchers proposed to add aluminum (or other reflective )wall between pixel) . the colour sensor is worse due to the bayer filter.

You see, the CCD in the low res camera has larger pixel, that means it can collection more light than that on high res CCD.

Actually, even two sensors have same pixel number and same pixel pitch, and used with same lens with same focusing, you still can get different result. for the high resolution sensor, especially the one with small pixel, the crosstalk between the neighboring pixel is quite significant (light will leak from one pixel to another pixel). such kind of crosstalk will degrade the image quality definitely.

It may take millions of dollar and years of time to just develop some usable prototypes.

This picture is quite misleading: no camera can output the picture (even the camera with 3CCD) as sharp as this one (because it is resized from a larger one). In order to get similar result, you may need multiply the number of ppf by 1.3 or even 1.5

What you need is just a mobile speed camera system without speed detection device (laser or radar). Police (or other people) has to pay more than 10K for even a "entry-level" product. If you plan to operate this system in night, you also need a really powerful narrow angle IR illuminator ( which is actually hard to source). you also need two cameras point to the target vehicle and they should use different exposure setting, 1st use shorter exposure to capture the image of number plate ( which is reflective). 2nd on use longer exposure to capture the image of driver and front seat passenger. forget the day/night camera with a "sealed" housing. all of them just use cheap M12 mount lens. it is impossible to find any M12 lens which is narrow enough for you application.

For most DVR, the image quality ( the sharpness) will be degraded by the compression algorithm, even you use a D1 format machine, you still need to make sure the number plate will occupy 1/6 of the picture. for a CIF format, 1/4 is essential ( as Soundy said). another issue is, in order to get sharp image without any blur introduced by vehicle's motion, you need to fix the shutter speed of the camera to 1/500 or shorter. (this figure depends on the speed)

Catching the number plate on the moving vehicle is a quite challenging job for any video surveillance system. In order to identify the number plate without any guess work, a minimum of 100pixel width is essential. that means the width your FOV should be less than 8'. so the only affordable choice should be block camera ( the only with integrated varifocal lens) based on SONY 480 or SONY 980 or something else similar.

Actually, the 650TVL has to achieved with the help of a CCD with higher resolution ( 976x582, panasonics's model, SONY has a similar one). In order to achieve 800TVL, the camera will need a CCD with a horizontal resolution of 1200 pix. or with 3 mono CCD ( even more expensive).

two choice: 1st: use the camera with larger sensor(1/2'' or 1/1.8''), you can try to get some cheap 135 format photograph lens ( Canon's 135mm F2.0 or Sigma's 150 mm F2., and then you can find some EF mount to C mount converter. 2nd, use the camera with smaller sensor (1/2.5''), there are ton's 75mm machine vision lens. they have very good resolution( of course most them are manual one, you can't change the aperture and focusing after installation).

Extra illumination (visible one or IR one) seems the only working solution.

Lot of times I find it easier to get it "close" with the focus ring, then fine-tune focus using the zoom ring - generally seems to be less "sensitive" on most lenses. I use same trick to adjust the focusing ring. " title="Applause" />

a real WDR is achieved by double exposure: one exposure with longer exposure time to get the detail in dark area, another is taken with shorter exposure time to get the detail in very bright area, then the two pictures is combinen in the DSP of camera to a single frame ( actually one field). the Digital WDR, to my understanding, should be just some kind manipulation of digitized picture in the camera, it may make the output of camera looks better ( if it does work). but won't provide any extra detail.

I think that the author means the camera with two different sensor ( of course, there will be two lenses for 2 sensor board). one is colour one, another is B/W one . At visible band, the B/W camera's the sensitivity triples that of colour camera, however, at IR band, it may only 10% or 20% better than colour camera (due to the bayer filter over the CCD sensor, of course, I mean the colour camera without IR cutting filter). Maybe that is why the dual head camera is not so attractive for IR based application.

It seems that what you need is suitable lens for your application. you need to figure out the the width ( or height) of the FOV (field of view) first, they you can calculate the the view angle. I don't think that there is any CCTV video camera which can output the video as good as that from current camcorder. if the resolution ( clarity , sharpness, whatever) is critical for, you may have to use some camera with digital output (or expensive 3CCD block camera).