Abstract:
To obtain useful focal content from different focal planes in the scene and fuse them into
an image which has complimentary information of the source scene thus creating an all in
focus focal planes is the main initiative of multi-focus image fusion. Lens have limited
depth of field (DOF) and due to acquisition conditions it is not always possible to get all
scene in focus thus there is a need of efficient and cost effective multi-focus image fusion
algorithm for obtaining all in focus image and thus rendering of more information from a
scene. In this thesis, new efficient multi-focus image fusion technique is proposed for different
applications. The proposed technique utilizes the concept of gradients, morphological
filtering and construction of refined focus maps to better the quality of fused image. Some
source image in multi-focus image fusion may be subjected to illumination variance which
is catered in this thesis such images are passed through a pre-processing stage. Images not
falling in this category are directly subjected to calculation of clarity, construction of focus
maps, refinement through Laplacian focus measure and final fused map construction.
To demonstrate the significance of proposed technique, visual and quantitative comparison with other existing techniques is presented. Simulation results reveal that the proposed technique is more accurate compared to existing technique.
