A recent meta-heuristic algorithm called Marine Predators Algorithm (MPA) is enhanced using Opposition-Based Learning (OBL) termed MPA-OBL to improve their search efficiency and convergence. A comprehensive set of experiments are performed to evaluate the MPA-OBL and prove the impact influence of merging OBL strategy with the original MPA in enhancing the quality of the solutions and the acceleration of the convergence speed, using IEEE CEC’2020 benchmark problems as recently complex optimization benchmark. In order to evaluate the performance of the proposed MPA-OBL, the effectiveness of conjunction of OBL with the original MPA and the other counterparts are calculated and compared with LSHADE with semi-parameter adaptation hybrid with CMA-ES (LSHADE_SPACMA-OBL), Restart covariance matrix adaptation ES (CMA_ES-OBL), Differential evolution (DE-OBL), Harris hawk optimization (HHO-OBL), Sine cosine algorithm (SCA-OBL), Salp swarm algorithm (SSA-OBL), and the original MPA. The extensive results and comparisons in terms of optimization metrics have revealed the superiority of the proposed MPA-OBL in solving the IEEE CEC’2020 benchmark problems and improving the convergence speed. Moreover, as a sequel to the proposed MPA-OBL, also, we have conducted experiments using two objective functions of Otsu and Kapur’s methods over a variety of benchmark images at different level of thresholds based on three commonly evaluation matrices namely Peak signal-to-noise ratio (PSNR), Structural similarity (SSIM), and Feature similarity (FSIM) indices are analyzed qualitatively and quantitatively. Eventually, the statistical post-hoc analysis reveal that the MPA-OBL obtains highly efficient and reliable results in comparison with the other competitor algorithms.