yolov5detect的检测结果进行目标提取的代码

把检测结果进行目标提取的代码可以参考以下代码：

import cv2
import numpy as np
from PIL import Image

# 加载模型
model = cv2.dnn.readNetFromDarknet('yolov5.cfg', 'yolov5.weights')
# 获取输出层信息
ln = model.getLayerNames()
ln = [ln[i[0] - 1] for i in model.getUnconnectedOutLayers()]

# 检测函数
def detect(image):
    # 获取图像尺寸
    (H, W) = image.shape[:2]

    # 构建一个blob
    blob = cv2.dnn.blobFromImage(image, 1 / 255.0, (416, 416), swapRB=True, crop=False)

    # 设置模型输入
    model.setInput(blob)

    # 运行前向传递
    layerOutputs = model.forward(ln)

    # 初始化检测结果列表
    boxes = []
    confidences = []
    classIDs = []

    # 处理每个输出层
    for output in layerOutputs:
        # 处理每个检测
        for detection in output:
            # 提取类别ID和置信度
            scores = detection[5:]
            classID = np.argmax(scores)
            confidence = scores[classID]

            # 过滤掉置信度低的检测
            if confidence > 0.5:
                # 获取检测框坐标
                box = detection[0:4] * np.array([W, H, W, H])
                (centerX, centerY, width, height) = box.astype("int")

                # 计算检测框的左上角坐标
                x = int(centerX - (width / 2))
                y = int(centerY - (height / 2))

                # 更新检测结果列表
                boxes.append([x, y, int(width), int(height)])
                confidences.append(float(confidence))
                classIDs.append(classID)

    # 进行非极大值抑制
    idxs = cv2.dnn.NMSBoxes(boxes, confidences, 0.5, 0.3)

    # 提取目标
    targets = []
    if len(idxs) > 0:
        for i in idxs.flatten():
            (x, y) = (boxes[i][0], boxes[i][1])
            (w, h) = (boxes[i][2], boxes[i][3])
            targets.append(image[y:y+h, x:x+w])

    return targets

# 测试
image = cv2.imread('test.jpg')
targets = detect(image)
for i, target in enumerate(targets):
    cv2.imwrite(f'target{i}.jpg', target)