A Novel Region-Based Method For Intensity Inhomogeneities Image Segmentation And Denoising
R. Jaya Subalakshmi · 2014
Every image is implausibly vital for the medical field to grasp the facts of the user/patient. throughout this report we've got a bent to propose a general methodology (PURE-LET) to vogue and optimize a decent class of thresholding algorithms for denoising footage corrupted by mixed Poisson-Gaussian noise. to boot, this report jointly proposes a singular region-based technique for image segmentation, that's in degree extraordinarily position to vary intensity inhomogeneities inside the segmentation. The methods related to the images with intensity inhomogeneities, we've got a bent to derive a district intensity agglomeration property of the image intensities, and description a district agglomeration criterion operate for the image intensities in associate passing neighborhood of each purpose. In the level set associate formulation, thes criterion defines associate energy in terms of the amount set method that represent a partition of the image domain and a bias field that accounts for the intensity irregularity of the image. For minimizing the energy, this technique is in degree extraordinarily position to at constant time section the image and estimate the bias field, and also the calculable bias field could even be used for intensity irregularity correction (or bias correction). As associate application, our technique presents denoising results obtained on real footage of low-count analysis and has been used for segmentation and bias correction of resonance (MR) and CT footage with promising results.