In order to improve the detection effect of the infrared dim target, an improved particle swarm optimization algorithm is proposed. Firstly, a quantum-behaved particle swarm algorithm is optimized based on the Gaussian distribution attraction factor, and the particle swarm mapping is optimized by the logistic chaos which can avoid the later evolution into the local optimum. Secondly, the reliability of the chaotic quantum-behaved particle swarm optimization algorithm is ensured according to the diversity determined by the average Euclidean distance of the particle swarm in the later stage. Finally, the infrared dim target is detected under the minimum mean variance criterion, and the prediction value is corrected which can ensure the validity of the detection. The experimental results show that the proposed algorithm is effective in detecting infrared dim targets with the largest signal noise ratio value, and the detection probability and false alarm probability are better than other algorithms.