Hello, everyone i am beginner to this field, and i am working on script identification from camera based multi-lingual document images.for this i need to do the pre-processing to the input images which are suffering from blurred, uneven illumination, warping and etc. here i am attaching the input images for your consideration. i need suggestions and tips from you people.
first you should use geometric transformation to obtain rectangle/ square shape and the brightness transformation.
brightness transformation should depend on amount of non uni formality in illumination. this may be obtained using proper window size for averaging brightness and running it over complete image.
then we may go for further processing.
we may continue discussion after above steps implementations.
Thank you sir, Sure sir i will try to implement the geometric transformation on my images.
thanks for your valuable suggestion sir.
- Badges
- Science topic
- Similar topics
- Statistics
- Time Series