Capturing the nonlinear aspects of a dynamic texture is both challenging and tough. The proposed innovative algorithm focuses on capturing both the aspects of dynamic texture that is linear as well as nonlinear motion. The innovative part of the proposed model is computing a seasonal matrix using simple mathematical calculations to predict a new frame as well as insert a predicted frame for reconstructing a dynamic texture from the original ones by using innovative, dynamic texture synthesis models. The proposed approach is used for dynamic texture synthesis with a very fewer amount of model coefficients and very high visual quality. It is useful to create a huge number of frames with a compression ratio of 33% and 90% and PSNR values between 45dB to 65dB. The main advantage over the other widely available methods is that there is no need for transform coding since the algorithm directly applies to the raw video thereby, reducing the model size, model parameters, time, and computational complexity.