High frequency assisted fusion for infrared and visible images through sparse representation

In this paper, an improved fusion method based on sparse representation is proposed to restrain the defects of infrared and visible image fusion, such as blocking artifact, edge blur and low contrast. Comparing with other fusion methods based on sparse representation, we propose two contributions in this paper. First, we present a novel strategy for the low frequency fusion assisted by high frequency, which suppresses blocking artifact and low contrast. Second, we design dictionaries suit for the properties of high frequency and low frequency respectively under the fusion framework. Experiments show that the performance of the proposed method is better than current state-of-the-art fusion algorithms, both on the visual quality and objective assessment.