Enhanced image fusion using directive contrast with higher‐order approximation

One of the most obvious characteristics of a good quality fused image is high spectral resolution. It is used to discriminate inner information in an image of low and high resolution by computing respective frequency components. The high‐resolution image results in high correlation between the adjacent bands. The authors have utilised resolution, contrast, and correlation using the wavelet fusion approach to achieve the enhanced image. A wavelet scheme is used to extract low‐ and high‐frequency data, forming wavelet domain components. These components are combined to create a fused wavelet coefficient map. To further improve the fused coefficients, we have used a higher‐order approximation to compute directive details of an image. The high‐frequency components are extracted to obtain edge and texture information to avoid too much spectral information loss because block‐wise wavelet transform decomposition and reconstruction introduces visible blocking artefacts in the image. In this work, phase variation in wavelets is also discussed and error analysis is performed. Comparison of various performance measures using different methods is shown in both qualitative and quantitative results.