I've seen some contributions about monitoring traffic pattern using TerraSAR-X sensor, I don't know whether it is possible to achieve this aim with lower spatial resolution; just try googling the keyword :"traffic monitoring TerraSAR-X" and you will find a lot papers.

Good luck with your project.

Hi Amirhossein. I've checked many of these papers. Usually the data is simulated and experiments were done with airborne SAR.

I have actual Radarsat-2 data and it seems quite impossible to detect passenger vehicles and sometimes even trucks. I am using standard SLC data, though.

TerraSAR-X is a space-borne platform (with about 515km height). As mentioned before in enclosed paper (please see the link) : "In the case of two-channel systems, like TerraSAR-X or the Canadian RADARSAT-2, interferometric approaches can be used for vehicle detection." Therefore, I think you will find some approaches for vehicles detection. However I think you should have at least a pair SLC.There are also some difficulties that you should take into account, such as if vehicles are moving in along-track direction, some approaches are incapable of the detection and should be developed. That was all I know.

http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=4563055&url=http%3A%2F%2Fieeexplore.ieee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D4563055

Dear Waqas, Before sending my comments, I also had this assumption that vehicles can easily be identified as the approaches of ship detection do. In the case of ship detection, we should take into account that the shape is very similar to a cone reflector and sea acts as a mirror. However in streets there are many objects that also have high back scatter coefficient and I don't think it would be an easy task to distinguish between them. In order to show how much it is difficult, I enclosed two images; the first one is TerraSAR-X (left) and the second is PAN band of worldview2 (right). You can see some intensive points denoting they are vehicles. However most of them are only street lights!

I found also another paper that its main goal is to detect moving targets. From my point of view, Napeloen should firstly specify that whether he is willing to detect moving targets or fixed ones?!

Paper : Sikaneta, I. and Gierull, C., 2005. Two-Channel SAR Ground

Moving Target Indication for Traffic Monitoring in Urban

Terrain. In: International Archives of Photogrammetry, Remote

Sensing and Spatial Information Sciences, Vol. 36number

3/W24, pp. 95–101.

I am mostly interested in finding fixed ones. Moving ones are not necessary.

I would assume that standard resolution (11.8 x 5.1m) would not be enough for normal passenger cars, am I correct. I mean, it appears that the radar cross section is too small. Am I correct?

I guess it would only be able in very ideal conditions. Like the car placed in an open grassland.. and even then the car would have to be properly aligned, right?

Dear Waqas, shouldn't a target like a vehicle "dominate" the pixel it is located in? Sorry for some much questions, just I'm trying to find a technical justification to provide our clients (engineers).

Just one question..why you call standard data something is not. Surely is a multi-look SAR image and in my humble opinion you should not use multi-look data ! Why do you want to waste your full-resolution? It sounds a bit crazy, don't you think?

The question did not specify, so I am presuming a stationary vehicle...  The resolution given suggests that most vehicles will be a dot or very small blob in a SAR image...  RadarSat is C-band, so typical vehicle might average +5 dBsm RCS or so, so would be detectable...  Then you have to decide if your dot is a vehicle or not, which is a different question...  and very difficult without additional knowledge...