Nonlinear computer image scene and target information extraction based on big data technology

Abstract To explore the extraction of computer image scene and target information, a nonlinear method based on big data technology is proposed. The can decompose into plurality components when SAR processed such as compression, which represent different captured features, respectively. Selecting most suitable processing according to characteristics greatly improve performance. Using diffusion method, decomposed structural representing large-scale information texture small-scale detailed automatic threshold estimation in process studied. LAIDA criterion introduced solution diffusion-based decomposition test evaluate various parameter forms. results show that experimental outcome very close each index, shows using estimation, no matter what index used, be obtained. Specifically, for algorithm, l outliers plays an obvious role. third degree initiative process. larger L , outlier, will lead greater extent process, resulting continuous decrease similarity compositional correlation. It proved algorithm has strong global search ability, effectively avoid premature convergence, fast convergence speed, good long stability. widely used optimization multimodal functions.