**7. Conclusions**

The clutter measurement density is a parameter required to calculate the data association probability of the measurement and target existence probability of a track and has a large impact on target tracking performance even with small changes. This paper presented the SCMDE with clutter measurement probability to estimate the clutter measurement density adaptively for non-parametric multi-target tracking in environments where there is no prior information about clutter distribution. The algorithm was developed by analyzing the causes of estimation performance deterioration of the existing SCMDE. The proposed clutter measurement density estimation method calculated the sparsity of the measurements by probabilistically classifying adjacent measurements as a target measurement or as a clutter measurement. We demonstrated the effectiveness of the proposed clutter measurement density estimation method, which was designed to achieve more accurate and robust clutter measurement density estimation by showing the performance improvement for multi-target tracking through simulation studies in various environments and a test with real radar data.

**Author Contributions:** Conceptualization, S.H.P. and T.L.S.; methodology, S.H.P.; software, S.H.P. and S.Y.C.; validation, S.H.P. and T.L.S.; formal analysis, H.J.K. and T.L.S.; investigation, T.L.S.; resources, S.H.P.; data curation, S.H.P. and S.Y.C.; writing, original draft preparation, S.H.P. and H.J.K.; writing, review and editing, S.H.P.; visualization, S.H.P. and H.J.K.; supervision, T.L.S.; project administration, T.L.S.; funding acquisition, T.L.S.

**Funding:** This work was supported by Hanwha Research and Development Center.

**Conflicts of Interest:** The authors declare no conflict of interest.
