| Summary: | 碩士 === 國立高雄應用科技大學 === 電子工程系 === 105 === his thesis presents the real-time video restoration system for linear motion-blurred frames. The proposed method can be applied to handheld imaging device, vehicle-based camera、wearable imaging device, and digital video recording system. It can be also implanted in the intelligent vision system, robot vision system, and unmanned vehicle vision system to restore the motion-blurred images for the purpose of the later object detection and recognition processing. This system consists of five parts：(1) Image preprocessing：Firstly, a motion-blurred image is transformed to the cepstrum domain using Fourier transform (2) Point spread function estimation： blur angle and blur length of PSF are directly estimated in the cepstrum. (3) Single-image deblurring: Both estimated blur angle and blur length are employed to normalize PSF and then the image restoration is achieved by performing the iterative deconvolution with the normalized PSF. (4) Image register: The selected clear images and the motion-deblurred images are stored in a register. (5) Multi-image deblurring: It is first to search the corresponding feature points between the current image and the latest image stored in the image register. These feature points are used for deducing a homography matrix required by the perspective transformation which can align the corresponding pixels of those two images. Then, the weights of pixels among such two images are calculated and compared for generating the restored pixels in such a way that the high-weight pixel replaces the low-weight pixel, where each weight of pixel is generated according to the temporal information. Thus, the final restored image is produced. From the experimental results, the propose video restoration system can restore most linear motion-blurred in video.
|
|---|
