Artifacts Removing in Ultrasonic Images by Fusion Processes

Abstract

Allmost airborne ultrasonic images used in robotics have artefacts and distorstions, when comparing real objects with those presented in ultrasonic images. The differences between reral and reference images have various causes, as assymetries in directivity of ultrasonic transducers, reflections on the explored environment, non-linearities of the processing blocks. Previous works on the problem, showed that such distortions could be removed or attenutated by classical or adaptive filtering of ultrrasonic images. This paper presents the results artefacts removing made by image fusion approach and wavelet transform. Results are accurate and at the same level of quality, comparing with other explored methods, i.e. based on filtering.