Figure 3

Finding contours of objects (neural structures). (A) The original image. The contour that needs to be extracted is encircled in red, which forms the initial snake. The image size is 101 × 101 pixels, with each pixel of size 27 nm. (B) An expanded image showing the normalized gradient vector field around the contour. Note that the vectors indicate the direction of deformation of the snake as it accommodates itself to the shape of the desired contour. (C) The final result, showing the snake acquires the contour shape of the neural structure.