A 2D Histogram-Based Image Thresholding Using Hybrid Algorithms for Brain Image Fusion

In this article, an effort is made to identify brain tumor disease such as neoplastic, cerebrovascular, Alzheimer's, lethal, sarcoma diseases by successful fusion of images from magnetic resonance imaging (MRI) and computed tomography (CT). Two images are fused in three steps: the two images are independently segmented by hybrid combination of particle swam optimization (PSO), genetic algorithm, and symbiotic organisms search (SOS) named as hGAPSO-SOS by maximizing 2-dimensional Renyi entropy. Image thresholding with 2-D histogram is stronger in the segmentation than 1-D histogram. The segmented regions with scale invariant feature transform (SIFT) algorithm are removed. Also, after image rotation and scaling, the SIFT algorithm is excellent at removing the features. The fusion laws are eventually rendered on the basis of type-2 blurry interval (IT2FL), where ambiguity effects are reduced unlike type-1. The uniqueness of the proposed study is evaluated on specific data collection of benchmark image fusion and has proven stronger in all criteria of scale.