The invention belongs to the field of signal processing, in particular to an impulse interference internal modulation characteristic real-time extraction and classification system based on an FPGA. The impulse interference internal modulation characteristic real-time extraction and classification system based on the FPGA is higher in real-time performance and higher in precision. The impulse interference internal modulation characteristic real-time extraction and classification system based on the FPGA comprises a signal generating module, a pulse forming module, a recognition module, a frequency measurement module and a comprehensive distinguishing module, wherein the recognition module comprises a phase modulation signal recognition module and a frequency modulation signal module, and the frequency measurement module relates to FFT frequency measurement and instantaneous phase difference frequency measurement. According to the impulse interference internal modulation characteristic real-time extraction and classification system based on the FPGA, rough recognition and fine recognition are conducted on signals in sequence, rough recognition is conducted on the signals through a 3dB bandwidth, so that the signals are divided into phase modulation signals and frequency modulation signals, the phase modulation signals comprise BPSK signals, QPSK signals and conventional radar signals, and the frequency modulation signals comprise LFM signals, NLFM signals, VFM signals and SFM signals; fine recognition is then conducted on the signals, the frequency modulation signals are further classified through time frequency analysis, and the phase modulation signals are further classified according to the time domain accumulation instantaneous self-correlation method.