Tuesday, December 18, 2007

Image retrieval using Incomplete Patterns

This should be definitely interesting !!! :)

If not, you didn't understand it completely

How is an image retrieved that has been registered before?
Images that are seen in consciousness with the mind are registered in the brain. When the same image is seen again neurons – the processors in the brain help to arrive at a conclusion that this kind of an image has been seen before or this image is the very exact ( same ) image that has been seen before or this is the very first time that this image is being seen and now its registered. So, there are at least 3 possibilities. {Exactly the same, new one, kind of that has been seen before}.

How do we recognize or come to a conclusion that this is a known image? I would say its not that the image has been completely stored in the brain cells ( am not sure where it would be stored exactly ) How is the brain able to recognize the image that is observed just with the pattern ( which is really not complete ) or the traces ? Had it been a complete comparison to the image viewed from the environment to the image stored in the brain, retrieval of the complete image including the shades, color, depression, skin color variants and many thousand different factors. When the attributes of the image are registered for the first time the environmental effects gets registered along with the image attributes.

Infact attributes of any image that is viewed is not stand alone as it definitely depends upon the environmental attributes and the viewer’s perspective. But that is not happening or absolutely not possible. So that is possibly ruled out. I don’t think so even the mother who is taking care of her beloved child will not be .. I take it authoritatively … cannot be able to reproduce the image of her child. Recognition and reproducing the entire image is entirely different. Both should not be confused as the first one is what happens in the world and current discussion is about the reproducing the entire image.

If it is possible that we recognize the image and not able to reproduce what are the traces that get registered / stored in the brain that enables to recognize the image when viewed from the environment ? So even if there is an attempt to reproduce the image that is stored without viewing the same image again, how would the environmental effects be imbibed to the retrieved image during the process of retrieving or is the brain and its army of processors that intelligent to map the environmental effects to suite the current environment where the image is retrieved. So there is necessarily a threshold value which is just enough to retrieve an image and we need not store the entire image. Its only the traces of the image that is needed to store and complete image need not be necessarily be stored.

No comments: