LR (116) OCC (237) 0.4 0.5 0.6 0.7 0.8 0.2 0.3 0.1 01020304050 Location-Error Threshold AutoTrack [0.700] ARCF-HC [0.696] CSR-DCF [0.686] ECO-HC [0.684] Staple [0.677] MCCT-H [0.676] Staple_CA [0.673] ARCF-H [0.672] STRCF [0.671] BACF [0.651] KCC [0.638] 0.4 0.5 0.6 0.7 0.8 0.2 0.3 0.1 0 0.2 0.4 AutoTrack [0.457] ARCF-HC [0.455] ECO-HC [0.442] STRCF [0.440] Staple [0.431] MCCT-H [0.430] Staple_CA [0.421] CSR-DCF [0.418] BACF [0.416] SRDCFdecon [0.415] ARCF-H [0.413] (c) Overlap Threshold SRDCF [0.409] KCC [0.406] DSST [0.397] fDSST [0.382] SAMF_CA [0.364] SAMF [0.346] CN [0.297] LCT2.0 [0.291] DCF [0.274] KCF [0.272] DSST [0.635] SRDCF [0.627] SRDCFdecon [0.624] SAMF_CA [0.615] fDSST [0.612] SAMF [0.582] CN [0.532] LCT2.0 [0.508] DCF [0.475] KCF [0.475] LR (116) 0.4 0.5 0.6 0.7 0.8 0.2 0.3 0.1 0.6 0.8 10 0.2 0.4 0.6 0.8 1 Overlap Threshold ECO-HC [0.459] STRCF [0.451] ARCF-HC [0.447] AutoTrack [0.436] CSR-DCF [0.431] MCCT-H [0.429] ARCF-H [0.421] Staple_CA [0.418] BACF [0.416] SRDCFdecon [0.413] Staple [0.412] (d) FIGURE 7. (Continued) The tracking results of each tracker under different attributes. (c) LR and (d) OCC. The number after the attribute indicates the total number of sequences with the specific attribute, e.g., the total number of sequences with the FM attribute in the six benchmarks is 69. The ranking standard in the precision plot is the precision (DP) under CLE = 20 pixels, and the standard in the successrate plot is the AUC. (Continued) 152 IEEE GEOSCIENCE AND REMOTE SENSING MAGAZINE MARCH 2022 SRDCF [0.409] KCC [0.385] SAMF_CA [0.375] DSST [0.369] fDSST [0.367] SAMF [0.354] LCT2.0 [0.317] DCF [0.296] CN [0.294] KCF [0.294] 0.4 0.5 0.6 0.7 0.8 0.2 0.3 0.1 01020304050 Location-Error Threshold ECO-HC [0.651] STRCF [0.638] CSR-DCF [0.635] ARCF-HC [0.634] AutoTrack [0.622] MCCT-H [0.615] ARCF-H [0.608] Staple_CA [0.597] Staple [0.590] BACF [0.587] SRDCF [0.579] SRDCFdecon [0.574] SAMF_CA [0.558] KCC [0.558] DSST [0.529] SAMF [0.521] fDSST [0.519] LCT2.0 [0.500] CN [0.466] DCF [0.460] KCF [0.456] OCC (237) Success Rate Precision Success Rate Precision