The methods based on statistical model are usually used to suppress se a clutter and detect target by photoelectric imaging system,but their performance of target detection is generally unstable in the s ituation of high ocean wave especially.A novel sea clutter cancellation approach for target image at sea based on chaotic neu ral network is proposed in this paper.The motion of sea clutter is chaotic,which induces that the photoelectric image of sea clutter is also chaotic.The dynamic model of sea clutter is constructed after the sea clutter phase space of chaotic dynamical system is rec onstructed.The model parameters used to predict sea clutter are obtained by radial basis function (RBF) neural network.Then the sea clutter suppression can be performed by clutter cancellation between the predicted and real sea clutters.The real sea clutter ph otoelectric image is used to test the model,and the results show that this method is more effective to suppress sea clutter compared with the least mean square and the largest Lyapunov exponent methods,the mean deviation is reduced by 30%,and the signal-to-noi se ratio is increased by 4－6dB.At the same time, a novel strategy based on chaos theory will be adopted to improve the performanc e of dim target detection in sea clutter by the photoelectric imaging system.