Leveraging polynomial approximation for non-linear image transformations in real time

Display Omitted A hardware architecture for non-linear real-time image remapping is proposed.Finite differences are used to calculate the remapping function on-the-fly.It is shown that sub-pixel accuracy can be guaranteed at low hardware costs.The architecture features minimum latency and little utilization of FPGA resources.Results for lens distortion removal and a gesture recognition system are presented. Applying non-linear image transformations in real time remains a challenge for cost-sensitive embedded systems today. This paper offers a method to perform such transformations efficiently using polynomial approximations. It is based on a previously published method, but is not limited to transformations with little vertical variation, such as lens distortion removal. Instead, a wider range of image transformations is supported.Thus, a new, improved hardware architecture is presented and discussed. It can be configured to achieve virtually any required accuracy for practical transformation tasks at low hardware costs. As a proof of concept, the architecture has been integrated into a gesture recognition and gaze analysis system as an intelligent stereo camera delivering undistorted and rectified images.