taurus

I was more thinking about a low-cost software solution. In the solution proposed by you, I would need a lot of additional hardware: a multiplexer, another computer and another a card. This is way over my budget. What I need is to get the image from the GeoVision software.

Although for my application it is sufficient to check the actual image 1 or 2 times per second, it requires real time access, as it is sending an alarm to COM1 based on what's happening.

Not very much, it's not an issue.

That's a good idea, but I think it's quite had to take out certain frames from .avi files. I would like to have an easier way of getting snapshots...

The algorithm that I would like to use is already implemented in a stand-alone application written in C++. This application would run in the background and do its processing automatically (and in certain situations send data to COM1). I would not like to rewrite the application in another language. So the question is how do I get images from the GV System and input them to my application? It can be through COM, through a file, through a TCP/IP socket on the localhost etc. I don't know much about GV System. Can it be controlled by a script somehow in order to save the actual image of certain cameras every second in .bmp or .jpg format? This would be the most convenient for me, as my application currently takes a .bmp as input. If this is not feasible, can I use one of the ActiveX components given by GeoVision to "query" the actual image?

I am using a special image recognition algorithm that has to recognize certain patterns on the image. The concept is similar to that of the license plate number recognition. So what I need, is the raw pixel data at a rate of 1-2 images/sec. When I say image, I am really talking about snapshots or frames or still images. The problem is that at the same time the camera functions as surveillance, so all the standard functionality present in DV System has to be available.

I am trying to solve a similar problem, so I will "revive" this topic... I have a computer with a GV-600 card and GV System installed. I would like to develop a utility running in the background on the same computer. This utility would do some processing based on the image got from certain cameras at certain intervals. My first idea was to use an open source driver for the GV-600 card and get the image directly from the card. This works, but unfortunately I cannot have two different drivers installed at the same time for the same device, so either my driver is installed and the GV System is not working, or my utility doesn't work with the GV driver. Right now I am thinking about a different solution: install the GV System in the normal way and query the image of the cameras using the SDK or one of the ActiveX components. Does anybody know which way would it be easier? Is there any documentation about this? Thank you in advance! taurus