Worst spot in the box2.2% vs 4.1%
Detail at the centre0.118 vs 0.122
Moves and turns0 vs 36
Coverage 3 → 16+ cameras
Detail 0.09 (sharp) → 0.22 mm/px (coarse)
seen by fewer than 3 cameras: a gap in the scan
Views in each strip, left to right: front, right side, back, left side.
| Person at the centre | V2 | V3.1c |
|---|---|---|
| Body seen by fewer than 3 cameras | 0.7% | 0.7% |
| Cameras seeing the whole body | 12 | 13 |
| Median cameras per spot | 16 | 15 |
| Body detail, finest view (mm/px) | 0.122 | 0.118 |
| Face detail (mm/px) | 0.087 | 0.087 |
V2as drawn: 13 cameras moved, 23 turned
V3.1cno camera moved or turned
| Person at the front-left corner | V2 | V3.1c |
|---|---|---|
| Body seen by fewer than 3 cameras | 2.6% | 2.0% |
| Cameras seeing the whole body | 9 | 11 |
| Median cameras per spot | 9 | 9 |
| Body detail, finest view (mm/px) | 0.152 | 0.130 |
| Face detail (mm/px) | 0.123 | 0.116 |
V2as drawn: 13 cameras moved, 23 turned
V3.1cno camera moved or turned
| Person at the front-right corner | V2 | V3.1c |
|---|---|---|
| Body seen by fewer than 3 cameras | 2.7% | 1.9% |
| Cameras seeing the whole body | 11 | 8 |
| Median cameras per spot | 9 | 10 |
| Body detail, finest view (mm/px) | 0.139 | 0.136 |
| Face detail (mm/px) | 0.119 | 0.118 |
V2as drawn: 13 cameras moved, 23 turned
V3.1cno camera moved or turned
| Person at the back-left corner | V2 | V3.1c |
|---|---|---|
| Body seen by fewer than 3 cameras | 4.1% | 2.2% |
| Cameras seeing the whole body | 9 | 11 |
| Median cameras per spot | 7 | 9 |
| Body detail, finest view (mm/px) | 0.130 | 0.123 |
| Face detail (mm/px) | 0.125 | 0.115 |
V2as drawn: 13 cameras moved, 23 turned
V3.1cno camera moved or turned
| Person at the back-right corner | V2 | V3.1c |
|---|---|---|
| Body seen by fewer than 3 cameras | 3.9% | 2.1% |
| Cameras seeing the whole body | 10 | 10 |
| Median cameras per spot | 7 | 9 |
| Body detail, finest view (mm/px) | 0.135 | 0.125 |
| Face detail (mm/px) | 0.121 | 0.113 |
V2as drawn: 13 cameras moved, 23 turned
V3.1cno camera moved or turned
How this was tested
- Body: the scan HSR0349 (walking pose, 1.77 m), placed at the centre and at each corner of the 104 × 104 cm floor box.
- Each camera's view is simulated from its calibrated position. A spot counts as seen when it is in frame, not hidden by the body and faced at less than 70°.
- V2 is the checklist as drawn. V3.1c keeps every camera's position and landscape/portrait orientation: 12 zooms at 18 mm for the whole body, 12 zooms set for detail (A05 stays at 105 mm on the face), 105 mm macros back on the other face cameras, and pan/tilt changes.
- Averaged over the centre and four corners: body seen by fewer than 3 cameras 2.8% (V2) and 1.8% (V3.1c); finest detail 0.136 and 0.126 mm/px.