With the increasing problem of water pollution, oil-water separation technology has attracted widespread attention worldwide. In this study, we proposed laser electrochemical deposition hybrid preparation of an oil-water separation mesh and introduced a back-propagation (BP) neural network model to realize the regulation of metal filter mesh. Among them, the coating coverage and electrochemical deposition quality were improved by laser electrochemical deposition composite processing. Based on the BP neural network model, the pore size after electrochemical deposition could be obtained only by inputting the processing parameters into the model, enabling the prediction and control of the pore size of the processed stainless-steel mesh (SSM), and the maximum residual difference between the predicted value and the experimental value was 1.5%. According to the oil-water separation theory and practical requirements, the corresponding electrochemical deposition potential and electrochemical deposition time were determined by the BP neural network model, which reduced the cost and time loss. In addition, the prepared SSM was found to achieve efficient separation of oil and water mixtures, reaching 99.9% separation efficiency in a combination with oil-water separation, along with other performance tests without chemical modification. The prepared SSM showed good mechanical durability and the separation efficiency exceeded 95% after sandpaper abrasion, thus, still maintaining the separation ability of oil-water mixture. Compared to other similar preparation methods, the method proposed in this study has the advantages of controllable pore size, simplicity, convenience, environmental friendliness, and durable wear resistance, offering important application potential in the treatment of oily wastewater.