The plasma current density profile plays a key role in the development of a high poloidal beta scenario, which is essential for long-pulse and high-performance plasma operation on a tokamak. Based on the polarimetry technique, a Motional Stark Effect (MSE) diagnostic has been built on the Experimental Advanced Superconducting Tokamak. To be prepared for real-time (RT) feedback control of the plasma current density profile in the future, a RT signal processing system has been developed. The RT signal processing system is composed of three functional modules: analog-to-digital conversion (ADC) module, polarization information extraction module, and digital-to-analog conversion (DAC) module. The final objective of this system is to acquire the polarization information of the MSE. Based on the field-programmable gate array unit, fast Fourier transformation is adopted to process the Photoelastic Modulator (PEM) digital signal, which was converted from a PEM signal via the ADC module. By means of frequency spectrum separation, the components around double modulating frequencies are restored through inverse fast Fourier transformation. Furthermore, the two amplitudes of their corresponding components can be obtained through a digital harmonic analyzer technique. Afterward, the ratio of the two amplitudes is calculated by arc tangent so that the polarization angle is obtained. Finally, the information of this polarization angle is converted into a voltage signal by the DAC module and then output in RT. The test results based on the RT signal processing system are in good agreement with those based on the phase lock-in amplifiers. The working cycle of this system is shorter than 10 ms, which meets the requirements of the MSE diagnostic as a RT controller. The algorithm of RT signal processing and the relevant technology applied for building this system are presented in the main body of this paper in detail.