Based on the information processing functionalities of spiking neurons, a spiking neural network model is proposed to extract features from a visual image. The network is constructed with a conductance-based integrate-and-fire neuron model and a set of specific receptive fields. The properties of the network are detailed in this paper. Simulation results show that the network is able to perform image feature extraction within a time interval of 100 ms. This processing time is consistent with the human visual system. The demonstrations show how the network can extract right-angle contours in a visual image. Based on this principle, many other image features can be extracted by analogy. The parallel processing mechanism of this network model is very promising for a hardware implementation based on VLSI or FPGA technology.