Mitigating Sensor Saturation through Image Processing Techniques

ABSTRACT:The objective of this SBIR research effort is to investigate algorithms for mitigation of sensor saturation, including the effects of laser dazzlers,in mid-wave infrared (MWIR) cameras using advanced image-processing techniques.The optical irradiance present in dazzled imagery spans several orders of magnitude more than conventional MWIR focal planes can reproduce. The large variation in irradiance is manifested in the imagery as severely under-exposed and over-exposed regions of the captured image as well as artifacts due to the scattering of rays within the imaging device.Image processing based solutions offer the potential for dazzler mitigation without prior knowledge of any dazzler characteristics and offer an alternative solution in applications where the addition of optical filters to the collection device is impractical or undesirable.Our approach employs High-Dynamic-Range image processing to combine multiple frames of varying exposure in a statistically rigorous manner in order to capture information in both low and high-light regions, and maximize the information content in a single image.In addition, a two step pre-processing scheme is utilized to separate and remove dynamic (lens flare) as well as static (main beam) contributions from the corrupting high energy source.BENEFIT:There will be immediate benefits in military applications of this technology to surveillance, reconnaissance, and target acquisition and tracking. For example, this technology may be used to mitigate saturation effects of directed energy laser dazzlers on guided missile electro-optics. Since methods developed in this project will not be specific to the MWIR band, they may be more generally applied over the visible to LWIR range of imaging devices. Programmable consumer digital cameras in the visible band could eventually be controlled with these algorithms, presenting a large market for this technology. Multiple frames with exposure times covering the full dynamic range of intensities would be automatically taken, with a composite HDR image then constructed from these frames. Lens flare due to reflections internal to the camera would also be reduced using the methods devised here.