earticle

영어

Video generation is the strategy of making videos by discovering moving pictures (videography) and making blends in live creation. During this aspect of videos creation, sequence of images gets contorted in the image acquisition and transmission phase. This paper presents a non-linear patch model for enhancing the degraded quality of video sequences. The proposed method navigates the frames of videos by selecting a patch of considerable width and with the assistance of this patch, a regression model is applied in order to have a robust effect while performing de-noising, de-blurring or super-resolution. The proposed work implements on three models i.e. search model, regression model and non-linear patch model. Various filters have been propounded e.g. kernel filters, total variation filters, adaptive median filter etc. Regression model is applied to every frame with predefined number of iterations with estimated number of frames. The proposed method inculpates two kinds of noise i.e. Gaussian noise and impulse noise. Various performance comparison metrics have been evaluated to check the coherence and productivity of imaging system like Peak signal to noise ratio (PSNR), Mean squared error (MSE), Root mean squared error (RMSE), Standard deviation (SD), Linear correlation and structural entropy.