Initial public release

backend/app/schemas/segmentation.py

@@ -0,0 +1,110 @@
+from __future__ import annotations
+
+from datetime import datetime
+from uuid import UUID
+
+from pydantic import BaseModel, ConfigDict, Field
+
+from app.schemas.detection import DetectionModelCapability
+
+
+SegmentationModelCapability = DetectionModelCapability
+
+
+class SegmentationModelsResponse(BaseModel):
+    models: list[SegmentationModelCapability]
+
+
+class SegmentationRunRequest(BaseModel):
+    model_config = ConfigDict(protected_namespaces=())
+
+    project_id: UUID
+    dataset_id: UUID
+    model_id: str
+    confidence_threshold: float = Field(default=0.5, ge=0.0, le=1.0)
+    class_filter: list[str] | None = None
+    tile_manifest_path: str | None = None
+    parameters_json: dict = Field(default_factory=dict)
+
+
+class SegmentationQaRequest(BaseModel):
+    reference_dataset_id: UUID
+    iou_threshold: float = Field(default=0.5, ge=0.0, le=1.0)
+    class_name: str | None = None
+    min_confidence: float | None = Field(default=None, ge=0.0, le=1.0)
+    # Read off the one matching pass, exactly as for detection.
+    calibration_thresholds: list[float] = Field(default_factory=list, max_length=32)
+
+
+class SegmentationRunResponse(BaseModel):
+    model_config = ConfigDict(protected_namespaces=())
+
+    analysis_run_id: UUID
+    job_id: UUID
+    project_id: UUID
+    dataset_id: UUID
+    model_id: str
+    status: str
+    segmentation_count: int
+    error_code: str | None = None
+    message: str
+
+
+class SegmentationRunRead(BaseModel):
+    model_config = ConfigDict(from_attributes=True, protected_namespaces=())
+
+    id: UUID
+    project_id: UUID
+    dataset_id: UUID | None = None
+    job_id: UUID | None = None
+    analysis_type: str
+    status: str
+    model_name: str | None = None
+    model_version: str | None = None
+    parameters_json: dict
+    result_json: dict | None = None
+    error_message: str | None = None
+    created_at: datetime | None = None
+    started_at: datetime | None = None
+    finished_at: datetime | None = None
+
+
+class SegmentationRunListResponse(BaseModel):
+    items: list[SegmentationRunRead]
+    # ``total`` counts every run; ``items`` is the most recent page of them.
+    total: int
+    limit: int | None = None
+    offset: int = 0
+    truncated: bool = False
+
+
+class SegmentationRead(BaseModel):
+    model_config = ConfigDict(from_attributes=True, protected_namespaces=())
+
+    id: UUID
+    project_id: UUID
+    dataset_id: UUID | None = None
+    analysis_run_id: UUID | None = None
+    job_id: UUID | None = None
+    model_name: str
+    model_version: str | None = None
+    class_name: str
+    confidence: float | None = None
+    bbox_json: dict | None = None
+    area_m2: float | None = None
+    mask_path: str | None = None
+    source_tile_path: str | None = None
+    tile_index: int | None = None
+    properties_json: dict | None = None
+    provenance_json: dict | None = None
+    created_at: datetime | None = None
+
+
+class SegmentationListResponse(BaseModel):
+    items: list[SegmentationRead]
+    # ``total`` describes the complete filtered population; ``items`` is one
+    # stable confidence-ranked page of it.
+    total: int
+    limit: int | None = None
+    offset: int = 0
+    truncated: bool = False