K.K. Kim, S.Y. Chi, and Y.K. Chung (Korea)
Local thresholding, camera document recognition
This paper presents region-oriented local thresholding for recognizing camera document image, which is captured under inconsistent lighting condition. Camera image is very sensitive to illumination that result in difficulties for recognizing character. Image enhancement and region oriented local thresholding method are proposed to reduce illumination effect, which produces misclassification of character region into background region. Camera document image is binarized using the difference of maximum and minimum intensity value of pixels in sub window. It is an efficiency method to extract character from background of camera document image and to provide high speed processing time. To evaluate the proposed method, we have experimented with camera document image of the ETRI database. An encouraging binarization results have been obtained.
Important Links:
Go Back
