A novel jamming identification method based on bispectrum and Zernike moment feature is proposed to effectively deal with two new kinds of jamming of smeared spectrum (SMSP) and chopping and interleaving (C＆I). In this method, the bispectrum analysis of received radar signal under three cases is given firstly through a series of dimensionality reduction and normalization, and a threedimensional feature information is turned into twodimensional feature information. Then the twodimensional feature spectrum is transformed into a gray image through a series of image processing, and Zernike moment feature is utilized for extracting image shape feature to constitute feature vector for signal recognition. The result shows that the proposed approach can achieve satisfying recognition. Particularly the method is less affected by SNR and the recognition rate can still reach 90％ under low signal to noise ratio. In comparison with paper［4~6］, this algorithm is the best in the recognition results, and this algorithm is also good in recognition rate under low signaltonoise ratio. The method is feasible in the field of radar jamming signal recognition.