JSEG

AcronymDefinition
JSEGJapan Society of Engineering Geology
References in periodicals archive ?
Algorithms SFF-IS 0.56 SFF 0.52 FPVVI 0.52 OASRG 0.51 SLIC 0.38 JSEG 0.37 GPU-SLIC 0.35 WS 0.28 GBIS 0.23 Algorithms SFF-IS 0.20 SFF 0.19 FPVVI 0.19 OASRG 0.18 SLIC 0.17 JSEG 0.17 GPU-SLIC 0.17 WS 0.12 GBIS 0.08 Note: Table made from bar graph.
Caption: Figure 5: Comparison of four segmentation methods (proposed, SAS, CTM, and JSEG).
Index Proposed SAS CTM JSEG 12084 PRI 0.7848 0.76155 0.7051 0.7026 VoI 2.4184 2.1261 3.6425 4.3114 GCE 0.2256 0.1753 0.1658 0.1209 BDE 7.7028 6.8194 7.4629 9.2061 130066 PRI 0.8209 0.7286 0.7710 0.7457 VoI 1.4075 2.2986 2.6194 3.6813 GCE 0.1416 0.3001 0.1117 0.1226 BDE 9.4618 18.6371 10.2823 13.3104 134052 PRI 0.7630 0.6454 0.6303 0.5833 VoI 1.2913 2.0665 2.6110 3.9948 GCE 0.1020 0.2263 0.0833 0.1302 BDE 6.1061 11.2896 9.4271 15.1976 15062 PRI 0.8882 0.8973 0.8673 0.8705 VoI 1.5428 1.4820 2.1843 2.6231 GCE 0.2303 0.1959 0.1386 0.1075 BDE 8.7102 21.2930 12.2526 13.9926 385028 PRI 0.9152 9.1108 0.9031 0.9036 VoI 1.5868 3.6813 1.8808 2.2501 GCE 0.1436 0.2236 0.0888 0.1072 BDE 9.1108 9.1016 8.6668 8.6987 Table 2: The average values of PRI, VoI, GCE, and BDE for BSD300.
The images in Fig.10-d-f are contours contrast of JSEG, UGC, and our method.
It is very clear that our algorithm performs better than JSEG and the UGC method in most of the test images.
In this part, we add two comparing methods Mean-Shift [18], and Bipartite-Graphcut [47] except JSEG and UCG.
In JSEG, it is not necessary to construct textons database for pixels classification, but it needs to estimate texture patterns called J-map for images.