The Landweber image reconstruction algorithm for electrical capacitance tomography (ECT) images combining with the particle swarm optimization (PSO) can improve image quality. However, the PSO is easy to fall into local optimum when it is used for image reconstruction. Aiming at the problem, we propose an image reconstruction algorithm based on the improved PSO and Landweber algorithm. The new algorithm adds the PSO with exponentially decayed inertial weights，and adds the exponentially decayed restraint factors in particle speed updating formula,  so that the algorithm has strong global search ability in prophase and strong local search ability in later stage, which further optimizes the initial reconstruction result of the Landweber algorithm to improve the quality of reconstructed images. We compare the ECT image reconstruction results of our algorithm with those of the typical LBP, the improved Tikhonov iterative algorithm and Landweber image reconstruction algorithm, and the comparison results verify the proposal’s effectiveness.  Experimental results also show that the new algorithm outperforms the other algorithms in terms of subjective and objective qualities of reconstructed images.