#!/usr/bin/env python # -*- coding: utf-8 -*- """ ------------------------------------------------- @File Name: utils.py @Author: Luyao.zhang @Date: 2023/5/16 @Update Date: 2024/11/22 修改和中文化 By 陳會安 @Description: ------------------------------------------------- """ from ultralytics import YOLO import streamlit as st import cv2 from PIL import Image import tempfile def infer_uploaded_IMX219(conf, model, task_type): """ Raspberry Pi 5 + IMX219 舊版 Picamera2：不使用 transform，改用 OpenCV 翻轉 """ try: from picamera2 import Picamera2 import cv2 except ImportError: st.error("缺少 picamera2，請執行：sudo apt install python3-picamera2") return picam2 = Picamera2() config = picam2.create_video_configuration( main={ "format": "RGB888", # ⭐ 避免 RGB → BGR 色彩錯誤 "size": (1280, 720) } ) picam2.configure(config) picam2.start() st_frame = st.empty() stop_flag = st.button("停止 IMX219 推論") # IMX219 在 RPi5 上通常需要水平與垂直翻轉 hflip = True vflip = True while not stop_flag: frame = picam2.capture_array() bgr = frame # 已經是 BGR888 # ⭐ 用 OpenCV 做翻轉 → 相容所有舊版 picamera2 if hflip and vflip: bgr = cv2.flip(bgr, -1) # 上下 + 左右 elif hflip: bgr = cv2.flip(bgr, 1) # 左右 elif vflip: bgr = cv2.flip(bgr, 0) # 上下 _display_detected_frames(conf, model, st_frame, bgr, task_type) picam2.stop() picam2.close() def _display_detected_frames(conf, model, st_frame, image, task_type): """ 使用 YOLO11 模型在影片幀上標記出偵測到的物體。 :param conf (float): 物體偵測的信心指數閾值。 :param model (YOLO11): YOLO11 模型的實例。 :param st_frame (Streamlit 物件): 用於顯示偵測影片結果的 Streamlit 物件。 :param image (numpy 陣列): 影片幀的 numpy 陣列。 :param task_type (string): YOLO 任務類型。 :return: 無返回值。 """ # Resize the image to a standard size image = cv2.resize(image, (720, int(720 * (9 / 16)))) # Predict the objects in the image using YOLO11 model res = model.predict(image, conf=conf) # Plot the detected objects on the video frame res_plotted = res[0].plot() st_frame.image(res_plotted, caption='偵測結果影片', channels="BGR", width = 'stretch' ) @st.cache_resource def load_model(model_path): """ 在指定 model_path 目錄下載 YOLO 模型。 參數： model_path (str): YOLO 模型檔案的路徑。 返回： YOLO 模型。 """ model = YOLO(model_path) return model def infer_uploaded_image(conf, model, task_type): """ 執行上傳圖片的推論 :param conf (float): 物體偵測的信心指數閾值。 :param model (YOLO11): YOLO11 模型的實例。 :param task_type (string): YOLO 任務類型。 :return: 無返回值。 """ source_img = st.sidebar.file_uploader( label="選擇圖檔...", type=("jpg", "jpeg", "png", 'bmp', 'webp') ) col1, col2 = st.columns(2) with col1: if source_img: uploaded_image = Image.open(source_img) # adding the uploaded image to the page with caption st.image( image=source_img, caption="上傳圖檔", use_container_width=True ) if source_img: if st.button("執行"): with st.spinner("執行中..."): if task_type == "物體偵測": res = model.predict(uploaded_image, conf=conf) res_plotted = res[0].plot()[:, :, ::-1] with col2: st.image(res_plotted, caption="偵測結果影像", use_container_width=True) try: with st.expander("偵測結果"): #boxes = res[0].boxes #for box in boxes: # st.write(box.xywh) for box in res[0].boxes: cords = box.xyxy[0].tolist() cords = [round(x) for x in cords] class_id = int(box.cls[0].item()) conf = box.conf[0].item() conf = round(conf*100, 2) st.write("分類編號:", class_id) st.write("分類名稱:", model.names[class_id]) st.write("邊界框座標:", cords) st.write("信心指數:", conf, "%") st.write("------------------------") except Exception as ex: st.write("尚未有圖檔上傳!") st.write(ex) elif task_type == "姿態評估": res = model.predict(uploaded_image, conf=conf) res_plotted = res[0].plot()[:, :, ::-1] with col2: st.image(res_plotted, caption="偵測結果影像", use_container_width=True) try: with st.expander("偵測結果"): for r in res: boxes = r.boxes kps = r.keypoints for idx, p in enumerate(kps): list_p = p.data.tolist() st.write("姿態評估偵測的17個關鍵點:", idx) for i, point in enumerate(list_p[0]): st.write(f"關鍵點: {i}: ({int(point[0])}, {int(point[1])})") st.write("------------------------") except Exception as ex: st.write("尚未上傳圖檔!") st.write(ex) def infer_uploaded_video(conf, model, task_type): """ 執行上傳影片檔的推論 :param conf (float): 物體偵測的信心指數閾值。 :param model (YOLO11): YOLO11 模型的實例。 :param task_type (string): YOLO 任務類型。 :return: 無返回值。 """ source_video = st.sidebar.file_uploader( label="選擇上傳影片檔..." ) if source_video: st.video(source_video) if source_video: if st.button("執行"): with st.spinner("執行中..."): try: tfile = tempfile.NamedTemporaryFile() tfile.write(source_video.read()) vid_cap = cv2.VideoCapture( tfile.name) st_frame = st.empty() while (vid_cap.isOpened()): success, image = vid_cap.read() if success: _display_detected_frames(conf, model, st_frame, image, task_type ) else: vid_cap.release() break except Exception as e: st.error(f"載入影片檔錯誤: {e}") def infer_uploaded_webcam(conf, model, task_type): """ 執行網路攝影機Webcam的推論 :param conf (float): 物體偵測的信心指數閾值。 :param model (YOLO11): YOLO11 模型的實例。 :param task_type (string): YOLO 任務類型。 :return: 無返回值。 """ try: flag = st.button( label="停止執行" ) vid_cap = cv2.VideoCapture(0) # local camera st_frame = st.empty() while not flag: success, image = vid_cap.read() if success: _display_detected_frames( conf, model, st_frame, image, task_type ) else: vid_cap.release() break except Exception as e: st.error(f"載入網路攝影機錯誤: {str(e)}") def infer_uploaded_IPcam(conf, model, task_type): """ 執行網路監控攝影機IP Camera的推論 :param conf (float): 物體偵測的信心指數閾值。 :param model (YOLO11): YOLO11 模型的實例。 :param task_type (string): YOLO 任務類型。 :return: 無返回值。 """ # tw live可搜尋台灣即時影像：https://trafficvideo2.tainan.gov.tw/b596d902 # 在側邊欄中輸入 URL url = st.sidebar.text_input("輸入URL") if url: st.write(f"您輸入的URL是: {url}") try: flag = st.button( label="停止執行" ) vid_cap = cv2.VideoCapture(url) # IP camera st_frame = st.empty() while not flag: success, image = vid_cap.read() if success: _display_detected_frames( conf, model, st_frame, image, task_type ) else: vid_cap.release() break except Exception as e: st.error(f"載入網路攝影機錯誤: {str(e)}") else: st.error(f"請在側邊欄[輸入URL]欄位輸入網路監控攝影機IP Camera的URL網址後, 按Enter鍵")