In this study, a magnetic actuator is controlled to obtain a constant flow rate micropumping by a magnetic nanofluid flow inside a semicircular-shaped microchannel with a constant square cross-section. A closed-loop control strategy is developed to achieve adequate control by utilizing a model based on experimental data. A novel digital image processing–based real-time velocity estimation algorithm is proposed as feedback information in the closed-loop control system. This paper aims to evaluate the success of the proposed real-time velocity estimation algorithm. The results indicate that the proposed algorithm can supply the fluid velocity inside the microchannel with a maximum deviation of less than 4% in the current micropumping system. Three different controllers—that is, proportional (P), proportional–integral (PI), and proportional–integral–derivative (PID)—are also implemented to test the influence of the control logic used in the closed-loop control strategy. Compared with the previous study, the implementation of a closed-loop control strategy in the present work provided a constant flow rate pumping. The findings reveal that the PID controller can maintain a constant flow rate with transient time values around 186, 94, and 60 seconds for reference values of 100, 200, and 300 µm/s, respectively. In addition, the PID controller could supply a constant flow rate within the microchannel 51% longer than the PI controller.