Identification cascade · selected frames → event

From selected frames to a carrier event.

Tollscopic DOT# uses focused extraction and catalog scoring to turn a vehicle track into a carrier identity, a ranked candidate set, or an explicit refusal. Stronger reasoning is spent only when the evidence justifies it.

01 · The cascade

Six stages. Each one earns its cost.

Lightweight, deterministic work happens first. Heavier semantic work runs on prepared evidence, not raw video. Scoring happens against a finite catalog. Refusal is a first-class output.

01 Preprocess

Crops, contrast, denoise, normalize. The panel is prepared before any model reads it.

02 Per-frame extraction

Each frame returns field candidates with confidence — not a final answer.

03 Cross-frame consensus

Reads are compared across frames. Agreement, disagreement, and frame diversity are surfaced.

04 Catalog retrieval

Plausible FMCSA candidates are retrieved from every observed field.

05 Joint scoring

The scorer evaluates joint evidence against candidates plus the out-of-catalog null.

06 Decision + event

Confidence class, usage policy, evidence references, version tags. The event leaves the system.

02 · Preprocessing

Prepare the evidence before the model reads.

Cropping, contrast, denoising, and frame normalization make the door panel more legible and reduce how much irrelevant image area the semantic extractor sees. Plain-English labels — no algorithm names on the public page.

01
Original
02
Geometry correct
03
Placard zoom
04
Contrast boost
05
Denoise · sharpen
06
Upscaled crop
03 · Per-frame extraction

A frame can say 'I see these possible digits' — or 'the panel is not visible.'

That uncertainty is kept. The system does not turn the first plausible string into a carrier.

What each frame returns
USDOT
observed digit hypotheses with per-digit confidence
Authority
MC / MX / FF prefix + digits, where visible
Name tokens
character spans and consensus across multiple reads
City · state
string + standardized state code
Frame usefulness
whether this frame should weight the consensus
Placard visible?
explicit signal — not assumed yes
frame_extract · 03
{
  "frame_id":  "03",
  "usdot": {
    "digits":     ["3","8","4","7","1","2","8|9"],
    "confidence": [0.95,0.94,0.92,0.91,0.92,0.88,0.55]
  },
  "name": {
    "consensus": "NORTHRIDGE FREIGHT CO",
    "edit_dist": 1
  },
  "authority": "MC-71458",
  "city_state": ["TOLEDO", "OH"],
  "placard_visible": true,
  "frame_quality":   0.92
}
04 · Cross-frame consensus

Reads are compared. Diversity matters. Duplicates don't.

Per-frame extractions are aggregated into one observation for the vehicle. Frames that look almost identical do not produce fake confidence. Frames that look different and still agree produce the most.

Selected frames
FRAME 03 384712 8 NORTHRIDGE FREIGHT 0.92
FRAME 04 384712 9 NORTHRIDGE FREIGHT CO 0.88
FRAME 07 384712 9 NORTHRIDGE FREIGHT CO 0.84
Consensus observation
USDOT
3847128 (0.45) · 3847129 (0.55)
last digit ambiguous
Name
NORTHRIDGE FREIGHT CO
3 frames agree
City/state
TOLEDO · OH
high confidence
Authority
MC-71458
consistent
Diversity
3 frames · spread 1.2s
independent reads
05 · Catalog retrieval and scoring

Which carrier explains the joint evidence best?

The catalog stage retrieves plausible carriers from all observed fields. The scorer asks whether the observed evidence is more likely if a candidate carrier is the true carrier — or if the out-of-catalog option explains it better than any carrier.

Candidate
Evidence supporting
Score
Verdict
1
Northridge Freight Co
name + state + MC + DOT variant
0.94
verified
2
Atlas Long-Haul LLC
literal DOT match · name mismatch
0.21
3
Northland Carriers Inc
partial name · weak agreement
0.08
⟨out of catalog⟩
null hypothesis
0.03
06 · Walkthrough

See each stage on a single track.

A schematic walkthrough of how the cascade handles three different tracks, with the verdict each pass produces at the end.

01
verified
Track 01
multi-field agreement
02
ambiguous
Track 02
two candidates close
03
catalog_miss
Track 03
foreign carrier · not in snapshot
Three schematic tracks, three outcomes. Each card pairs a representative side-fire view with the verdict the cascade would emit at the end of the pass.

The cascade is not a demo. It is the system.

If you want to see your traffic mix go through it, we can stand up an audit-only run on selected lanes and walk through the events together.