Hand-held Camera Based Document Image Mosaicing Algorithm

This paper presents a global alignment model based image mosaicing method for camera-captured document images,and it can be used to combine multiple overlapping document images into one large image. It corrects the perspective distortion with the estimated vanishing points,and there exists only an affine transform between two adjacent images. Then,it adjusts the distance of featurepoints to distribute them as evenly as possible in the overlapping regions. Thirdly,it uses local alignment constraints of all the overlapping image pairs to construct global alignment model,thus,to eliminate the error accumulation. In order to reduce alignment error of overlapping area,a binary weighted function is used to blend the overlapping region of image pairs. This method is unique because it does not require the calibration of the internal/external camera parameters in advance and does not restricting the camera position,thus allowing greater flexibility than scanner-based or fixed-camera-based approaches. It can produce a high resolution and accurate full page mosaic from small image patches of a document.