69980d8073
- Refactor app API and schemas; adjust surgery pipeline, repository, and session manager. - Improve consumption TSV logging and consumable vision integration; trim voice resolution. - Add Baidu Face 1:N search script, .env.example entries, and client API integration doc. - Update demo client, staging checklist, surgery interface doc, and related tests; add sample face image. Made-with: Cursor
162 lines
9.0 KiB
Plaintext
162 lines
9.0 KiB
Plaintext
# Copy to `.env` and adjust. Loaded by pydantic-settings (field names → UPPER_SNAKE_CASE env vars).
|
||
# Used by local `start.sh` and Docker Compose. See also docs/video-backends.md.
|
||
|
||
# --- PostgreSQL ---
|
||
POSTGRES_USER=postgres
|
||
POSTGRES_PASSWORD=postgres
|
||
POSTGRES_DB=operation_room
|
||
POSTGRES_HOST=localhost
|
||
POSTGRES_PORT=35432
|
||
|
||
# Optional: full async SQLAlchemy URL (overrides POSTGRES_* when set and matches defaults logic — see Settings).
|
||
# DATABASE_URL=postgresql+asyncpg://postgres:postgres@localhost:35432/operation_room
|
||
|
||
# --- YOLO 视觉推理（内部调用，无独立 HTTP）---
|
||
# 耗材分类权重默认 app/resources/consumable_classifier.pt；手部检测为空时退化为全帧分类。
|
||
# CONSUMABLE_CLASSIFIER_WEIGHTS=/absolute/path/to/consumable_classifier.pt
|
||
CONSUMABLE_CLASSIFIER_IMGSZ=224
|
||
CONSUMABLE_CLASSIFIER_DEVICE=
|
||
# 视觉待确认 options 最多为模型 top3 与 candidate_consumables 的交集（非本变量）。
|
||
CONSUMABLE_CLASSIFIER_TOPK=5
|
||
# CONSUMABLE_MIN_CLS_CONFIDENCE=0.5
|
||
# 时间窗（秒）：窗内多次推理取众数后再走自动记账 / 待确认。
|
||
# CONSUMABLE_VISION_WINDOW_SEC=15
|
||
# 可选：Excel「商品名称」「产品编码」表；空则物品 id 用名称。
|
||
# CONSUMABLE_CATALOG_XLSX_PATH=/path/to/视频中的商品信息表.xlsx
|
||
# HAND_DETECTION_WEIGHTS=/absolute/path/to/hand_detect.pt
|
||
# HAND_DETECTION_IMGSZ=640
|
||
# HAND_DETECTION_CONF=0.25
|
||
# HAND_DETECTION_PAD_RATIO=0.30
|
||
# HAND_DETECTION_MIN_CROP_PX=64
|
||
# HAND_DETECTION_DEVICE=
|
||
# Device: empty → auto (macOS MPS if available; Linux CUDA if available). Docker image uses CPU torch unless you change it.
|
||
|
||
# --- Surgery recording API retries ---
|
||
# SURGERY_RECORDING_MAX_ATTEMPTS=3
|
||
# SURGERY_RECORDING_RETRY_DELAY_SECONDS=1.0
|
||
|
||
# --- Video: RTSP + optional Hikvision HCNetSDK (Linux x86_64 + glibc recommended) ---
|
||
# Client `camera_ids` must match keys in your RTSP map (sample IDs: or-cam-01, or-cam-02).
|
||
#
|
||
# VIDEO_DEFAULT_BACKEND=rtsp
|
||
# Values: rtsp | hikvision_sdk | auto (auto: SDK .so loaded and HIKVISION_SDK_ENABLED=true → prefer SDK)
|
||
#
|
||
# Per-camera backend override (JSON object):
|
||
# VIDEO_CAMERA_BACKEND_OVERRIDES_JSON={"or-cam-01":"rtsp","or-cam-02":"hikvision_sdk"}
|
||
#
|
||
# RTSP URL resolution (first match wins per camera_id):
|
||
# 1) VIDEO_RTSP_URLS_JSON_FILE — JSON file: {"or-cam-01":"rtsp://...","or-cam-02":"rtsp://..."}
|
||
# 2) merged with VIDEO_RTSP_URLS_JSON (inline JSON string; overrides same keys from file)
|
||
# 3) else VIDEO_RTSP_URL_TEMPLATE with {camera_id}
|
||
# Example file (committed): app/resources/camera_rtsp_urls.sample.json
|
||
# VIDEO_RTSP_URLS_JSON_FILE=app/resources/camera_rtsp_urls.sample.json
|
||
# In Docker (WORKDIR /app): VIDEO_RTSP_URLS_JSON_FILE=/app/app/resources/camera_rtsp_urls.sample.json
|
||
# WARNING: if VIDEO_RTSP_URLS_JSON_FILE is set but the path is missing, the app will fail to start.
|
||
#
|
||
# VIDEO_RTSP_URL_TEMPLATE=rtsp://user:pass@192.168.1.64:554/Streaming/Channels/101
|
||
# VIDEO_RTSP_URLS_JSON={"or-cam-01":"rtsp://user:pass@192.168.1.101:554/Streaming/Channels/101","or-cam-02":"rtsp://user:pass@192.168.1.102:554/Streaming/Channels/101"}
|
||
#
|
||
# VIDEO_OPEN_TIMEOUT_SEC=15
|
||
# 连续读帧失败次数达到阈值后释放 RTSP 并重连。
|
||
# VIDEO_READ_FAILURE_RECONNECT_THRESHOLD=15
|
||
# VIDEO_RECONNECT_BACKOFF_SECONDS=1.0
|
||
# VIDEO_INFERENCE_INTERVAL_SEC=2
|
||
# VIDEO_INFERENCE_CONFIDENCE_THRESHOLD=0.35
|
||
# 置信度 >= 此值且命中候选清单时自动 vision 记账。提高到 0.9 可减少自动记账、更多走待确认。
|
||
# 默认 0.9：Top1 置信度不足该值时入队待确认；达到且标签在 candidate_consumables 内则直接记 vision。
|
||
# VIDEO_AUTO_CONFIRM_CONFIDENCE=0.9
|
||
# 与 VIDEO_VOICE_CONFIRM_MIN_CONFIDENCE 共同决定何时自动 / 待确认（见 app/config 注释）。
|
||
# VIDEO_VOICE_CONFIRM_MIN_CONFIDENCE=0.35
|
||
# 待确认话术由服务端生成（prompt_text），TTS 一般在客户端播放；医生 WAV 上传后服务端 ASR 解析。
|
||
# 解析顺序：① pending 里展示的 topk（序号/名称）；② 仍不匹配时，对「开始手术」请求体中的 candidate_consumables 全文做名称子串匹配——医生报清单内其它耗材也以医生为准入账。
|
||
# 是否启用低置信度人工确认（客户端播报 + resolve 回传；服务端无麦克风/扬声器要求）。
|
||
# VOICE_CONFIRMATION_ENABLED=true
|
||
# 同一条待确认在语音/文本「解析失败」时累计的允许失败轮次（默认 2：首败后再给 1 次重试提示；见 422 的 detail.retry_remaining）。
|
||
# VOICE_CONFIRM_MAX_FAILED_PARSE_ROUNDS=2
|
||
# VIDEO_VOICE_CONFIRM_DOCTOR_ID=voice
|
||
# （已弃用）服务端本机录音 / ffmpeg 音频输入；当前闭环不依赖。
|
||
# VOICE_RECORD_SECONDS=5
|
||
# VOICE_FFMPEG_INPUT=
|
||
# 停录后写库失败时，后台重试落库间隔（秒）。
|
||
# ARCHIVE_PERSIST_RETRY_INTERVAL_SECONDS=30
|
||
# VIDEO_DETAIL_COOLDOWN_SEC=15
|
||
# VIDEO_JPEG_QUALITY=85
|
||
# VIDEO_RESULT_DOCTOR_ID=vision
|
||
# 每次单帧分类得到 top1～3 时打一条 INFO（联调开；生产建议 false）
|
||
# VIDEO_LOG_INFERENCE_RESULTS=true
|
||
# 时间窗级消耗文本日志（制表符列；每例手术 start 时截断+表头，窗内结果追加；终端 Rich 为可读时间戳，文件内为 ISO 时间戳列）
|
||
# CONSUMPTION_TSV_LOG_ENABLED=true
|
||
# 须含 {surgery_id}，如 logs/consumption_{surgery_id}.txt
|
||
# CONSUMPTION_TSV_LOG_PATH=logs/consumption_{surgery_id}.txt
|
||
# 同一时间窗结果在终端以 Markdown 表格打印（Top1～3 分列 id / 名称 / 置信度）
|
||
# CONSUMPTION_LOG_MARKDOWN_TERMINAL=true
|
||
# 消耗日志时间戳列的时区（IANA，如 Asia/Shanghai）；不设置则用运行环境的系统时区
|
||
# CONSUMPTION_LOG_TIMEZONE=Asia/Shanghai
|
||
#
|
||
# 语音确认：stderr 中可 grep 的 `VoiceConfirm ...` 行 + 每例手术 TSV（与 `start_surgery` 同次截断写表头；成功/ASR/解析失败均追加一行）
|
||
# VOICE_FILE_LOG_ENABLED=true
|
||
# 须含 {surgery_id}，如 logs/voice_{surgery_id}.txt
|
||
# VOICE_FILE_LOG_PATH=logs/voice_{surgery_id}.txt
|
||
|
||
# --- Hikvision: mount vendor Linux x86_64 .so at runtime (do not commit proprietary binaries) ---
|
||
# HIKVISION_LIB_DIR=/opt/hikvision/lib
|
||
# Optional: single library path (overrides directory search in code)
|
||
# HIKVISION_LIB_PATH=
|
||
# HIKVISION_SDK_ENABLED=false
|
||
# HIKVISION_DEVICE_IP=
|
||
# HIKVISION_DEVICE_PORT=8000
|
||
# HIKVISION_USER=
|
||
# HIKVISION_PASSWORD=
|
||
# HIKVISION_CHANNEL=1
|
||
# After SDK login, OpenCV still pulls frames via RTSP; template placeholders: {ip} {user} {password} {channel} {camera_id}
|
||
# HIKVISION_PREVIEW_RTSP_TEMPLATE=rtsp://{user}:{password}@{ip}:554/Streaming/Channels/101
|
||
# Per-camera RTSP when using SDK path (same shape as VIDEO_RTSP_URLS_JSON):
|
||
# HIKVISION_CAMERA_RTSP_URLS_JSON={"or-cam-01":"rtsp://...","or-cam-02":"rtsp://..."}
|
||
# HIKVISION_SDK_FALLBACK_TO_RTSP=true
|
||
|
||
# --- Baidu Speech（可选，遗留；当前手术闭环由客户端完成 TTS/ASR，服务端可不配置）---
|
||
# BAIDU_SPEECH_APP_ID=
|
||
# BAIDU_SPEECH_API_KEY=
|
||
# BAIDU_SPEECH_SECRET_KEY=
|
||
# BAIDU_SPEECH_CONNECTION_TIMEOUT_MS=
|
||
# BAIDU_SPEECH_SOCKET_TIMEOUT_MS=
|
||
# 短语音识别模型：固定普通话（默认 1537；勿用 1737 英语等）。代码会始终带上此 dev_pid。
|
||
# BAIDU_SPEECH_ASR_DEV_PID=1537
|
||
|
||
# --- Baidu Face（可选；仅 `scripts/baidu_face_1n_search.py` 批量人脸 1：N 搜索；需在控制台创建应用并开通人脸识别）---
|
||
# BAIDU_FACE_APP_ID=
|
||
# BAIDU_FACE_API_KEY=
|
||
# BAIDU_FACE_SECRET_KEY=
|
||
# 搜索的人脸组 id，逗号分隔，最多 10 个；未传命令行 --groups 时使用此项
|
||
# 仅允许英文/数字/下划线（与控制台「用户组 id」一致），不能中文；否则 API 会报 222005
|
||
# BAIDU_FACE_GROUP_ID_LIST=my_group_1
|
||
# BAIDU_FACE_MAX_USER_NUM=1
|
||
# BAIDU_FACE_MATCH_THRESHOLD=80
|
||
# BAIDU_FACE_QUALITY_CONTROL=NONE
|
||
# BAIDU_FACE_LIVENESS_CONTROL=NONE
|
||
# BAIDU_FACE_CONNECTION_TIMEOUT_MS=
|
||
# BAIDU_FACE_SOCKET_TIMEOUT_MS=
|
||
|
||
# --- MinIO（语音 WAV 存桶；`docker-compose.dev.yml` 内已含 `minio` 服务；本机只跑 API 时填 127.0.0.1:9000）---
|
||
# docker compose -f docker-compose.dev.yml up -d minio
|
||
# MINIO_ENDPOINT=127.0.0.1:9000
|
||
# MINIO_ACCESS_KEY=minioadmin
|
||
# MINIO_SECRET_KEY=minioadmin
|
||
# MINIO_BUCKET=operation-room-voice
|
||
# MINIO_SECURE=false
|
||
# optional: MINIO_REGION=
|
||
|
||
# --- Demo 浏览器客户端 / 一键联调假 RTSP（仅开发；生产关）---
|
||
# demo_client/index.html 跨源访问本服务 / 一键开录
|
||
# DEMO_CORS_ENABLED=true
|
||
# DEMO_CORS_ORIGINS=*
|
||
# 为 true 时提供 POST /internal/demo/orchestrate-and-start；需能执行 docker+ffmpeg 的**同一进程**内起 MediaMTX（通常=在宿主机直接跑 main.py，或容器挂载 /var/run/docker.sock）
|
||
# DEMO_ORCHESTRATOR_ENABLED=false
|
||
# VIDEO_RTSP_URLS_JSON_FILE 必须设成**可写**的 JSON 文件；Docker 中请 bind-mount 宿主机文件，与一键覆盖写入的映射一致
|
||
# DEMO_ORCHESTRATOR_RTSP_PORT=18554
|
||
# 手配假流、只改 JSON 给「另一进程」用时：可把 127.0.0.1 换成 host.docker.internal 等。
|
||
# 一键联调 orchestrate-and-start 在本进程起流+拉流，固定写 127.0.0.1，不读此项。
|
||
# DEMO_ORCHESTRATOR_RTSP_JSON_HOST=host.docker.internal
|
||
# 一键起 MediaMTX 后，等待本机 RTSP 端口可连接的最长时间（秒）
|
||
# MEDIAMTX_TCP_READY_SEC=30
|