Real world scenes often contain large variations in brightness values making it difficult for accurate digital imaging. An approach to accurately image high dynamic ranges is to capture the same scene successively with different exposure times, and then combine these varying exposures into one final image. But if this capture time is too long, the image will be distorted by changes in the scene. Proper selection of the different exposure times thus increases image quality by capturing the widest amount of information with the fewest number of frames while limiting the amount of scene distortion. This technology prescribes exposure times based on the Fibonacci sequence while correcting for motion in the scene. The final image is a combination of these processes, resulting in improved efficiency and accuracy for high dynamic range imaging.
Current methods that combine multiple exposures do not efficiently determine which set of exposure values to use, potentially leading to excessive image capture times. This technology makes uses the Fibonacci sequence for selecting exposure times, where subsequent exposures are equal to the sum of the previous two exposures. Such patterning ensures that a sufficiently wide range of data is captured for high quality imaging. Additionally, this technology estimates and corrects for the amount of motion that occurs between each frame. It can also be implemented in video systems, where each frame consists of a high dynamic range image.
Experimental results have shown that this method can be effectively implemented on basic camera systems to obtain high dynamic range images.
Tech Ventures Reference: IR CU13156