In this paper, the direction of arrival (DOA) and polarization parameter estimation algorithm with sparse collocated loop and dipole cross array is proposed. According to the rotational invariance relationship between the dipoles and the loops, the polarization parameter estimations are obtained. On the basis of the subspace theory, the array steering vectors are got, the phase differences of the virtual array elements are acquired according to the feature that the phase differences between the array elements of two ends of diameter and the coordinate origin are opposite. The virtual short baselines between the virtual and actual array elements are obtained, the rough but unambiguous estimates of DOA are herein achieved. The precise but ambiguous estimates of DOA are got by using the long baseline of actual array elements, these above-mentioned ambiguities can be disambiguated by using the virtual baseline method, the high-precision estimations of DOA are herein acquired. The parameters of this method can be automatically matching. Simulation examples have verified the correctness and validity of our method.