Content-aware image retargeting is the problem of adapting images to different display sizes and aspect ratios while minimizing distortions to the most important regions of those images. Seam carving is an operator for content-aware image retargeting that iteratively removes 8-connected pixel paths (seams) from an image until a target resolution is reached. Finding optimal seams for seam carving is computationally expensive. We have proposed the concept of minimum energy windows as an approach to reduce the computational load of seam finding. Our results demonstrate that it is possible to find nearly-optimal seams and obtain high quality results with a significant performance improvement.