Exposure fusion is the process of generating a pseudo-HDR (high dynamic range) image by directly fusing the images in a multi-exposure sequence. The main weakness of exposure fusion is the appearance of ghost artifacts caused by moving objects while shooting the sequence. Therefore, this article presents an improved temporal consistency assessment based on intensity and hue consistency maps for ghost-free exposure fusion. An intensity consistency map is generated by multiplying a multi-level threshold binary map and an intensity binary map, while a hue consistency map is generated by the difference of hue angle. The two maps are then multiplied to obtain a temporal weight map. A spatial weight map is generated by multiplying three image quality measures. Finally, the temporal and spatial weight maps are multiplied together, resulting in the final weights needed to produce the pseudo-HDR. Experiments show a reduction of ghost artifacts in the pseudo-HDR images produced by the proposed method when compared with the images produced by other conventional methods.