Medical image fusion using m-PCNN

An Adaptive Sub Band Filtering Based Medical Image Fusion Through Non-subsampled Contourlet Transform

“…Image fusion methods are usually divided into transform domain [5][6][7] and spatial domain [8][9][10] techniques. Fusion methods in the spatial domain are directly on pixel graylevel or color space from the source images for fusion operation, so the spatial domain fusion methods are also known as singlescale fusion method.…”

An Improved Saliency for RGB-D Visual Tracking and Control Strategies for a Bio-monitoring Mobile Robot

“…Fusion can be seen in two ways based on the input variations; a) integration of different sensors in which the sensor data types may differ [1] b) fusion of the same type of data or features generated from the same type of source such as images or image features [2][3][4][5]. In addition, as stated by Luo and Chang [1], the grouping can be made based on the fusion algorithm as a different perspective: a) low level (estimation methods), b) medium level (classification methods), and c) high level (inference methods) fusion [1].…”

“…the MRI images [2], RGB color images [4], SAR images [6], active mm-wave images [7]), and the fusion algorithm is in medium level according to the method base grouping. Therefore, the main goal is to create a fused data with more informative data that could be used for further processing such as detection, classification, segmentation, noise reduction, data reduction, etc.…”

“…Considering the internal approaches, they are independent of the application, can utilize space projection, feature extraction, fusion concepts as a closed approach in an integrated manner, are easy to be used as a module and can be embedded in to a big application. Therefore, in this paper, we propose an internal approach that refers to the use of principal component analysis (PCA) [13] and multi-channel pulse coupled neural network (m-PCNN) fusion [2] with one of the most recent saliency computation method based on Wavelet transform (WT) [14] enhance the saliency map during fusion process. The local saliency model in [14] has saliency features from edge to textures as a good representation of saliency with local to global contrast for each feature channel.…”

Image Fusion Using a Parameterized Logarithmic Image Processing Framework

“…Moreover, the combination of MRI and CT images can provide images containing both dense bone structure and soft tissue information (Yang et al, 2010). Similarly, the combination of MRI-T1 images provides greater details of anamotical structures while MRI-T2 images provides greater contrast between normal and abmormal tissue matter, and thus, their fusion can also help to extract the features needed by physicians (Wang, 2008). In security applications, thermal/infrared images provide information regarding the presence of intruders or potential threat objects (Zhang & Blum, 1997).…”