An Optimized Segmentation of Images uisng Modified Level Set Method
P. Shivachandana, B Rushikesh, Ganta Raghotham Reddy, Kama Ramudu · 2024
Image segmentation is a crucial step in processing digital images, aimed at partitioning an image into distinct regions or segments. The Additive Bias Correction (ABC) model addresses challenges in image segmentation arising from intensity inhomogeneity. While existing models, such as multiplicative bias field correction, have made progress, they encounter limitations such as slow segmentation speeds and limited applicability. The ABC model starts by defining local regions and a local clustering criterion to address intensity inhomogeneity. This criterion is then transformed into an energy function using the level set method. During segmentation, the bias field component and reflection edge structure are estimated to minimize this energy function. Compared to traditional multiplicative models, the additive approach achieves significantly faster computation speeds. Experimental results demonstrate that the proposed ABC model surpasses traditional methods in terms of robustness and speed, representing a significant advancement in image segmentation techniques.