Moments Descriptors are invariant under RST ( Rotation, Scaling, Translation) in PR.
Is Moment Descriptor almost invariant under Optimization?
Is Moment Descriptor almost invariant under Guassian Noise?
Can the reduction of pixels to draw an optimal shape of a planar curve shape using mouse on computer affect recognition rate? If yes then upto which extent ?
I'm not sure about your question, but an important factor if you are considering using moment descriptors is to make sure that moments exist in your domain. (For example Lorentzian curves, which are common in my work, have no defined moments).
Moment descriptors are invariant under RST. The second question is not clear. What you mean by moment descriptors invarinat under optimization. Moments of images are robust to different kinds of noise.
If by 'optimization' you mean image compression, then yes. However, it depends on the image compression scheme.
There are various kinds of invariants. Invariance to noise or radiometric degradations is also explored in the literature. It seems to me, achieving invariance to Gaussian noise, should be easy. Some clues may occur in the areas of "invariance to blurring" and convolution. best wishes
Hlo.. Guys...Im working on Handwriting recognition as my B.Tech 4th yr project... Plz somebody tell me about the AI algos used for that
The variety of algorithms used is considerable, check out the International Workshop on Frontiers in Handwriting Recognition (IWFHR) conference series
I'll search them...but if u could plz provide me the link....it wud be gr8....thnx in advance
It also depends what you are going to do, if you want to recognize the signature this is one thing, but to recognize the handwritten text-this is smth. different
Actually we are Converting the hand written Text to Standard Text that every one can understand
- Badges
- Science topic
- Similar topics
- Mathematics