Computer Vision News - May 2016

Tool That’s all! Now you can invoke the region growing loaded image with the preselected seed points. This process is performed on the entire image (3D) propagated through the axial slice. It performs quite fast (less than 10 sec) for our 300x500x500 scan. You can interactively examine, change the parameters and observe the result immediately after the changes are made. This makes the development process quite easy and efficient: We will finish this demonstration with another variant of the region growing algorithm which allows the user to specify the threshold based on the statistics estimated from the seed points. This algorithm parameters are: (a) The "multiplier" parameter is the constant c from the formula; (b) a region around each seed point “initialNeighborhoodRadius“, from which the statistics are estimated, can be specified; (c) the "numberOfIterations" allows you to rerun the algorithm. In the first run, the bounds are defined by the seed voxels specified by you, while in the following iterations they are estimated from the segmented points and the region growing is updated accordingly. The results for the same CT scan with the same initial points is: COMPUTER VISION NEWS 14