PROJECT TITLE :

Combining inertial measurements with blind Image deblurring using distance transform - 2016

ABSTRACT:

Camera motion throughout exposure results in a blurry image. We tend to propose an image deblurring method that infers the blur kernel by combining the inertial measurement unit (IMU) information that track camera motion with techniques that seek blur cues from the image sensor information. Specifically, we tend to introduce the notion of IMU fidelity value designed to penalize blur kernels that are unlikely to possess yielded the observed IMU measurements. When combined with the image knowledge-primarily based fidelity and regularization terms used by the standard blind image deblurring techniques, the energy function is nonconvex. We have a tendency to solved this nonconvex energy minimization downside by a completely unique use of distance rework, recovering a blur kernel and sharp image that are per the IMU and image sensor measurements.