| 摘要 |
Ultrasound images are stitched together by extracting target image regions within each ultrasound image, identifying interest points for the extracted target images and then identifying interest point correspondences between the interest points of all the extracted target image regions. The interest point correspondence is used to define a common, word coordinate for all extracted target images, and the extracted target images are warped onto the world coordinates. A distance weight, w1, gradient weight w2, and histogram weight w3 is then determined for each pixel of the warped, target images. The three weights are then combined and normalized across all the warped, target images. The combined, normalized weights are used in the blending and stitching of the warped, target images. |
| 主权项 |
1. A method of stitching together a respective image region from each of a plurality of input images, comprising:
detecting feature points in each image region, and identify candidate corresponding feature points among the image regions, reject incorrect correspondences among the candidate corresponding feature points to define corresponding features points; warping the corresponding feature points onto a defined world coordinates spanning the image regions; For each of said image regions,
(a) defining a first weight, i.e. w1, as a “distance weight” wherein each pixel is assigned a distance weight dependent upon its proximity to an edge of its respective image region;(b) defining a second weight, i.e. w2, as a “gradient weight” wherein pixels having a higher gradient, i.e. difference in intensity value as compared to neighboring pixels, receiving a higher weight value;(c) defining a third weight, i.e. w3, as a “histogram weight” of the intensity of each pixel across all image regions, stitching the respective image regions by a blending function applied to overlapping areas of the image regions, wherein in said blending function, across each image region to be stitched, its three weights are normalized to weight values from 0 to 1, to define a first normalized weight, i.e. Norm(1), a second normalized weight, i.e. Norm(w2) and a third normalized weight, i.e. Norm(w3), and overlapping regions are combined by the following relationship: w=Norm(w1)+Norm(w2)+2*Norm(w3). |
