Inspired by the direction selectivity mechanism of vision system, a new image clutter modeling method was proposed and applied to the target acquisition performance evaluation of imaging system. Firstly, the gradient direction of pixels was used to simulate the response direction of neurons in the local receptive field of the visual system. By comparing the gradient direction similarity between the central pixel and the adjacent pixel, the relationship between neuron "excitation" and "inhibition" response was simulated. The visual pattern was designed based on directional selectivity. Secondly, considering the sensitivity of human visual system to brightness contrast, the brightness contrast was used as a factor to weight the visual pattern histogram. The background clutter was modeled in the weighted histogram space. Finally, the relationship between background clutter and target acquisition performance was deduced, and the target acquisition performance evaluation model of photoelectric imaging system was established. The experimental results show that the target acquisition performance evaluated by the proposed method is consistent with the actual target acquisition performance in the field, and is superior to the existing performance evaluation methods in terms of root mean square error and correlation.