Add PS Move HID probe
This commit is contained in:
parent
b8d5d9f667
commit
a79bb1aa9f
7 changed files with 478 additions and 0 deletions
|
|
@ -426,6 +426,16 @@ a named invariant that the native runtime cannot protect yet.
|
|||
generic KS exposure/gain writes and only recovered two LED components; both
|
||||
PS3 Eye feeds currently score fixed-control frames because their native
|
||||
wrapper does not yet expose gain/exposure.
|
||||
PS Move is now a concrete active-marker input candidate rather than a lore
|
||||
object on a cable. `Mimir.PsMoveProbe` enumerates Sony `VID_054C` /
|
||||
`PID_03D5` over Windows HID with no third-party dependency. The connected
|
||||
controller exposes three HID collections: one generic desktop gamepad
|
||||
collection with 49-byte input/output/feature reports, plus two vendor-defined
|
||||
feature-report collections. The safe `--read` path uses overlapped `ReadFile`
|
||||
plus `CancelIoEx` so idle input reads time out instead of hanging the probe.
|
||||
Next useful cut is report decoding and controlled LED/sphere command
|
||||
provenance, then optical sphere observations from PS3 Eyes/Kiyos can enter
|
||||
the same calibration-evidence path as LED spline constraints.
|
||||
PS3 Eyes now have the first sparse tracking surface:
|
||||
`MimirSparseFeatureTracker` detects grid-suppressed Bayer/luma corner
|
||||
features, links them frame-to-frame by bounded nearest-neighbor motion, and
|
||||
|
|
|
|||
|
|
@ -33,6 +33,10 @@ flowchart TD
|
|||
- Stable LED index identity requires temporal/color/address coding or another
|
||||
correspondence source; uncoded identical lights are spline constraints, not
|
||||
metric depth authority.
|
||||
- PS Move controllers are active tracked-marker witnesses. USB HID owns
|
||||
identity, buttons, IMU/report topology, and later LED/rumble control; the
|
||||
glowing sphere is visual calibration evidence only when observed by cameras
|
||||
and fused by the global residual owner.
|
||||
- Exposure/gain sweeps optimize recovered spline quality: detected LED count,
|
||||
curve coverage, spacing coherence, smoothness, exposure fitness, and
|
||||
saturation penalty. Bright clipped blobs are not good evidence.
|
||||
|
|
|
|||
Loading…
Reference in a new issue