Description

ABSTRACT:

        Here we proposed method to segmenting and tracing the single neuron. We develop Tree2Tree, a robust automatic neuron segmentation and morphology generation algorithm.  Tree2Tree uses a local medial tree generation strategy in combination with a global tree linking to build a maximum likelihood global tree. The single image neuron will be the input to the process. In preprocessing we remove the noise from the image. And then we enhance the neuron image. We find the gradient of the image. The binary image is generated for the neuron image. The binary image will be segmented by the hessian based method. We use a Hessian-based scheme to detect pixels belonging to lengthened tubular structures in an intensity image. Here it determines whether the pixel is belonging to the neuron structure.  If it is belong to the neuron. The pixel will marked as one and the other pixels are marked as zero. After segmentation, we are tracing the neuron with the original image. Here we trace the neuron by the use of α-ß graph pruning algorithm. Finally we map the segmented neuron to the original image.