Image Fusion Technique Based on PCA and Fuzzy Logic Part 2

The output is a fusion image. Each input image consists of pixels. Since the range of each pixel value is [0, 255], there are 256 gray levels. The 256 gray level is divided into five fuzzy sets (VL, L, M, H, VH). Since input and output are 256 grayscale gray scale images, use the same fuzzy set for input and output. 2) Membership functions: Membership functions are used to demonstrate pixel value distribution and clustering, and allow optimal fusion operators and image fusion decision rules.

Fuzzy Logic: Fuzzy logic is a common method in fusion of control and sensor to represent uncertainty that reasoning is based on realism rather than absolute value. However, as sensor input increases, this approach becomes more complicated. In addition, verifying this method requires extensive testing where security is an important factor. Evidence theory (ET): The advantage of ET is that it can represent incomplete evidence, complete ignorance, and lack of necessity of prior probability. In the field of intelligent vehicle recognition, there are various incomplete information. Uncertain or inaccurate. For example, if an object is lost (hidden), the sensor can not measure all the relevant characteristics (hardware constraints) of the object and the observation is ambiguous (partial object detection). However, as the number of hypotheses increases, ET becomes more difficult to compute.

Integration of vehicle target detection based on deep learning network and future and multisensor fusion algorithm

Medical image fusion plays an important role in clinical diagnosis. This paper proposed a sparse representation theory and a multimode multidose medical image combination fusion model based on PCA. Visual and quantitative experimental results showed that this method effectively demonstrates the geometric shape and edges and prove that its performance is superior to PCA and OMP fusion. The modality can also be extended to merge multiple source images from multiple resolutions, multiple spectral frequencies, and multiple modalities.

In this research, we propose a fusion method using PCA transformation and sparse transformation. Please make effective use of the advantages of PCA and sparse fusion scheme. The proposed fusion framework is shown in FIG. First, we extract common innovative components from multiple images of geometric arrangement of the same scene. Secondly, different fusion rules are used to fuse these coefficients. To test the performance of the proposed joint fusion algorithm, we compared the quantitative and qualitative results with the two prior art methods. Qualitative measurements are made by visual inspection taking into account transparency and noise suppression. Since the proposed joint fusion uses both the PCA domain and the sparse domain, we use PCA which is the sparse OMP fusion method for comparison. For evaluation, we used a joint strategy for multiple resolution, multiple focus and multimode images, and compared the results with existing algorithms.