The linear subspace algorithm and nonlinear subspace algorithm is explored to detect point targets. We call them as linear Eigentargets and nonlinear Eigentargets. Linear principal component analysis (LPCA) is based on the second-order correlations without taking higher-order statistics into account. So LPCA is only appropriate to represent the data with a Gaussian distribution. That results in the performance limitation of linear Eigentargets detection based on LPCA. For improving detection performance, we extend linear Eigentargets to its nonlinear version, nonlinear Eigentargets, in this paper. Because the nonlinear PCA is capable of capturing the part of higher-order statistics, the better detection performance can be achieved.