### Table 2: Discontinuity angles of torus surfaces.

"... In PAGE 6: ...04 0.11 Studying the one point method, Table2 shows the maximum dis- continuity angle along all the boundary curves using our scheme as we increase the sampling rate. We conjecture that the angle be- tween the normals of adjacent patches is O(h4), as long as certain derivatives are bounded.... ..."

### Table 2: Magnitude of Normal Angle Discontinuity (radian)

1996

"... In PAGE 16: ... Methods 3{5 have smaller approximation errors than Methods 1 and 2; however, they need a convolution curve segment, Cv(t) + r v, to ll the gap between each pair of nearby convolution curve segments, Ci(t) + r Qi(s(t)), i = 1; 2, so that the o set curve segments are connected with G1-continuity. Table2... ..."

Cited by 9

### Table 1. Restoration of the height map from various smoothed surface normal vector fields. Surface Normal Number of

2002

"... In PAGE 3: ... The result- ing heights have been smoothed by averaging in a a0 a40 a0 neighbourhood in two passes over the entire image. The MCE between the vector fields obtained from the esti- mated height map and those provided initially are shown in Table1 . We can observe from this table that we obtain a quite accurate representation of the surface normals de- rived from the estimated height map.... ..."

Cited by 1

### Table 2. Accuracy Increase When Surface Normals Are Considered

"... In PAGE 8: ... Here, the model is represented by 4,468 points, and the multiresolution mesh contains 205 elements. Surface Normals in the ICP Algorithm Table2 shows the advantage of adding surface normal measurements. An arbitrary transformation is applied to the bone surface and is then sampled with N points.... ..."

### Table 4: Performance of merging rules at crease discontinuities with changing . Table shows the minimum required by merging rules to correctly detect a crease discontinuity with 100% success.

1998

"... In PAGE 20: ... The minimum for 100% success increases with . Table4 shows these values at different . For artificial (non-existent) discontinuities, data are generated for surfaces A and B from the line z = 100 + x.... ..."

Cited by 5

### Table 5: Percentage success in merging artificial discontinuities to fit from correct model.

1998

"... In PAGE 20: ... for 100% success increases with . Table 4 shows these values at different . For artificial (non-existent) discontinuities, data are generated for surfaces A and B from the line z = 100 + x. Table5 compares performance of different merging rules at a region size of 25 pixels per surface. The results show that merging rule based on RISS, BAYES, BMSC- RISS, BMSC-BAYES perform the best, followed by RUNS, CAIC, CHI.... ..."

Cited by 5

### Table 2. Control Surface Areas

"... In PAGE 6: ... The axial force characteristics (Fig. 8(a)) remain essentially the same in trend, but with a positive increment that scales proportionally with projected control surface area (see Table2 ). The normal force characteristic (Fig.... ..."

### Table 4: Performance of merging rules at crease discontinuities with changing #1B. Table shows

1998

"... In PAGE 20: ... The minimum #0B for 100% success increases with #1B. Table4 shows these #0B values at different #1B. For artificial (non-existent) discontinuities, data are generated for surfaces A and B from the line z = 100 + x.... ..."

Cited by 5

### Table 5: Percentage success in merging artificial discontinuities to fit from correct model.

1998

"... In PAGE 20: ...0B for 100% success increases with #1B. Table 4 shows these #0B values at different #1B. For artificial (non-existent) discontinuities, data are generated for surfaces A and B from the line z = 100 + x. Table5 compares performance of different merging rules at a region size of 25 pixels per surface. The results show that merging rule based on RISS, BAYES, BMSC- RISS, BMSC-BAYES perform the best, followed by RUNS, CAIC, CHI.... ..."

Cited by 5