119 lines
4.3 KiB
Python
119 lines
4.3 KiB
Python
|
import cv2
|
||
|
|
import numpy as np
|
||
|
|
import os
|
||
|
|
|
||
|
|
def detect_yellow_distance_from_bottom(image_path, visualize=False):
|
||
|
|
"""
|
||
|
|
检测鱼眼图像中垂线上最靠近下方的黄点到图像底部的距离
|
||
|
|
|
||
|
|
参数:
|
||
|
|
image_path: 图像路径
|
||
|
|
visualize: 是否显示检测过程可视化结果
|
||
|
|
|
||
|
|
返回:
|
||
|
|
distance: 黄点到图像底部的距离(像素)
|
||
|
|
center_x: 黄点的x坐标(用于垂线参考)
|
||
|
|
mask: 黄色区域掩模(可视化时使用)
|
||
|
|
"""
|
||
|
|
# 1. 读取图像并转换色彩空间
|
||
|
|
img = cv2.imread(image_path)
|
||
|
|
if img is None:
|
||
|
|
raise ValueError(f"无法读取图像，请检查路径: {image_path}")
|
||
|
|
|
||
|
|
hsv = cv2.cvtColor(img, cv2.COLOR_BGR2HSV)
|
||
|
|
height, width = img.shape[:2]
|
||
|
|
|
||
|
|
# 2. 定义黄色颜色范围 (考虑不同光照条件)
|
||
|
|
lower_yellow = np.array([20, 100, 100])
|
||
|
|
upper_yellow = np.array([30, 255, 255])
|
||
|
|
|
||
|
|
# 3. 创建黄色区域掩模
|
||
|
|
mask = cv2.inRange(hsv, lower_yellow, upper_yellow)
|
||
|
|
|
||
|
|
# 4. 形态学处理去除噪声
|
||
|
|
kernel = np.ones((5,5), np.uint8)
|
||
|
|
mask = cv2.morphologyEx(mask, cv2.MORPH_OPEN, kernel)
|
||
|
|
mask = cv2.morphologyEx(mask, cv2.MORPH_CLOSE, kernel)
|
||
|
|
|
||
|
|
# 5. 寻找轮廓
|
||
|
|
contours, _ = cv2.findContours(mask, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
|
||
|
|
|
||
|
|
if not contours:
|
||
|
|
print("未检测到黄色区域")
|
||
|
|
return None, None, mask
|
||
|
|
|
||
|
|
# 6. 找到所有黄色区域的中心点
|
||
|
|
yellow_points = []
|
||
|
|
for cnt in contours:
|
||
|
|
M = cv2.moments(cnt)
|
||
|
|
if M["m00"] > 100: # 忽略太小的区域
|
||
|
|
cx = int(M["m10"] / M["m00"])
|
||
|
|
cy = int(M["m01"] / M["m00"])
|
||
|
|
yellow_points.append((cx, cy))
|
||
|
|
|
||
|
|
if not yellow_points:
|
||
|
|
print("未找到有效的黄色中心点")
|
||
|
|
return None, None, mask
|
||
|
|
|
||
|
|
# 7. 计算图像中心垂线 (考虑鱼眼畸变，使用图像中心作为参考)
|
||
|
|
center_x = width // 2
|
||
|
|
vertical_line_threshold = width * 0.1 # 垂线左右10%的容差范围
|
||
|
|
|
||
|
|
# 8. 筛选在垂线附近的黄点
|
||
|
|
vertical_points = [p for p in yellow_points if abs(p[0] - center_x) < vertical_line_threshold]
|
||
|
|
|
||
|
|
if not vertical_points:
|
||
|
|
# 如果没有完全垂直的点，选择最接近垂线的点
|
||
|
|
vertical_points = sorted(yellow_points, key=lambda p: abs(p[0] - center_x))[:1]
|
||
|
|
print(f"警告: 没有严格垂直的点，使用最接近垂线的点: {vertical_points[0]}")
|
||
|
|
|
||
|
|
# 9. 找出最下方的黄点
|
||
|
|
lowest_point = max(vertical_points, key=lambda p: p[1])
|
||
|
|
|
||
|
|
# 10. 计算到图像底部的距离
|
||
|
|
distance = height - lowest_point[1]
|
||
|
|
|
||
|
|
# 可视化结果
|
||
|
|
if visualize:
|
||
|
|
vis = img.copy()
|
||
|
|
# 标记所有黄点
|
||
|
|
for (cx, cy) in yellow_points:
|
||
|
|
cv2.circle(vis, (cx, cy), 5, (0, 255, 255), -1)
|
||
|
|
# 标记垂线区域
|
||
|
|
cv2.line(vis, (center_x, 0), (center_x, height), (0, 255, 0), 1)
|
||
|
|
cv2.line(vis, (int(center_x - vertical_line_threshold), 0),
|
||
|
|
(int(center_x - vertical_line_threshold), height), (0, 255, 0), 1)
|
||
|
|
cv2.line(vis, (int(center_x + vertical_line_threshold), 0),
|
||
|
|
(int(center_x + vertical_line_threshold), height), (0, 255, 0), 1)
|
||
|
|
# 标记最低黄点
|
||
|
|
cv2.circle(vis, lowest_point, 10, (0, 0, 255), -1)
|
||
|
|
cv2.line(vis, (lowest_point[0], lowest_point[1]),
|
||
|
|
(lowest_point[0], height), (0, 0, 255), 2)
|
||
|
|
cv2.putText(vis, f"Distance: {distance}px", (10, 30),
|
||
|
|
cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 0, 255), 2)
|
||
|
|
|
||
|
|
# 确保保存目录存在
|
||
|
|
os.makedirs("saved_images", exist_ok=True)
|
||
|
|
|
||
|
|
# 保存原始图像和结果图像
|
||
|
|
cv2.imwrite("saved_images/Original_Image.jpg", img)
|
||
|
|
cv2.imwrite("saved_images/Detection_Result.jpg", vis)
|
||
|
|
|
||
|
|
|
||
|
|
return distance, center_x, mask
|
||
|
|
|
||
|
|
# 使用示例
|
||
|
|
try:
|
||
|
|
distance, center_x, _ = detect_yellow_distance_from_bottom(
|
||
|
|
"/home/mi-task/saved_images/right_20250820_120541_263023.jpg",
|
||
|
|
visualize=True
|
||
|
|
)
|
||
|
|
|
||
|
|
if distance is not None:
|
||
|
|
print(f"黄点到图像底部的距离: {distance} 像素")
|
||
|
|
print(f"参考垂线x坐标: {center_x}")
|
||
|
|
else:
|
||
|
|
print("未能检测到有效的黄色点")
|
||
|
|
|
||
|
|
except Exception as e:
|
||
|
|
print(f"发生错误: {str(e)}")
|