求助：从复杂图像中提取黄色集装箱方形端面的技术难题

从彩色图像中提取黄色集装箱的方形端面时遇到技术难题：

• 目标方形区域与背景色调相近，无明显视觉分界
• 集装箱上方及右侧的树木加剧了识别干扰

尝试通过HSV黄色阈值生成掩码，但无法让边界框准确贴合集装箱端面，原代码及识别情况如下：

原代码 #

import cv2 as cv

im = cv.imread("check1.jpg")
# Convert BGR ( not RGB ) to HSV as OpenCV works in BGR, not RGB.
hsv = cv.cvtColor(im, cv.COLOR_BGR2HSV)
cv.imshow('BGR2HSV__hsv-1',hsv)

# set range of yellowish colour, in HSV 
lower = (8, 0, 0)
upper = (78, 255, 255)

# Threshold the hSV range to get only yellow colours
mask1 = cv.inRange(hsv, lower, upper)

#erode the image to get rid of any background dross.
mask2 = cv.erode(mask1, kernel=None, iterations=2)
    
(x, y, w, h) = cv.boundingRect(mask2)

canvas = cv.cvtColor(mask2, cv.COLOR_GRAY2BGR)

final = cv.rectangle(canvas, (x,y), (x+w, y+h), color=(0,0,255), thickness=3)
cv.imshow('bounding box-final',final)

cv.waitKey(0)

图像情况 #

• 原始图像：画面左侧是黄色集装箱方形端面，上方及右侧有树木遮挡，背景为浅色调地面/墙面，与集装箱黄色色调接近
• 识别结果图像：掩码误包含部分背景和树木区域，边界框覆盖范围远大于集装箱端面，无法精准定位

1. 精细化HSV阈值，动态匹配目标色彩 #

原HSV范围过于宽泛，导致误识别背景。黄色在HSV中的典型有效范围可通过动态调参工具精准匹配：

import cv2 as cv
import numpy as np

def nothing(x):
    pass

im = cv.imread("check1.jpg")
hsv = cv.cvtColor(im, cv.COLOR_BGR2HSV)

cv.namedWindow('Trackbars')
cv.createTrackbar('H Lower','Trackbars',20,179,nothing)
cv.createTrackbar('H Upper','Trackbars',30,179,nothing)
cv.createTrackbar('S Lower','Trackbars',100,255,nothing)
cv.createTrackbar('S Upper','Trackbars',255,255,nothing)
cv.createTrackbar('V Lower','Trackbars',100,255,nothing)
cv.createTrackbar('V Upper','Trackbars',255,255,nothing)

while True:
    h_low = cv.getTrackbarPos('H Lower','Trackbars')
    h_high = cv.getTrackbarPos('H Upper','Trackbars')
    s_low = cv.getTrackbarPos('S Lower','Trackbars')
    s_high = cv.getTrackbarPos('S Upper','Trackbars')
    v_low = cv.getTrackbarPos('V Lower','Trackbars')
    v_high = cv.getTrackbarPos('V Upper','Trackbars')
    
    lower = np.array([h_low, s_low, v_low])
    upper = np.array([h_high, s_high, v_high])
    mask = cv.inRange(hsv, lower, upper)
    
    cv.imshow('Mask', mask)
    if cv.waitKey(1) & 0xFF == ord('q'):
        break
cv.destroyAllWindows()

调参完成后，将最终阈值替换到原代码中即可。

2. 形态学操作组合，清除干扰并补全目标区域 #

原代码仅用腐蚀会缩小目标区域，建议采用开运算+闭运算组合：

# 替换原形态学操作部分
kernel = cv.getStructuringElement(cv.MORPH_RECT, (5,5))
# 开运算：先腐蚀后膨胀，清除背景噪点
mask_open = cv.morphologyEx(mask1, cv.MORPH_OPEN, kernel, iterations=1)
# 闭运算：先膨胀后腐蚀，填补集装箱内部缝隙
mask_closed = cv.morphologyEx(mask_open, cv.MORPH_CLOSE, kernel, iterations=2)

3. 轮廓筛选，基于方形特征过滤无效区域 #

利用集装箱端面的方形特性，通过轮廓的面积、长宽比、多边形近似值筛选目标：

# 替换原boundingRect部分
contours, _ = cv.findContours(mask_closed, cv.RETR_EXTERNAL, cv.CHAIN_APPROX_SIMPLE)
canvas = cv.cvtColor(mask_closed, cv.COLOR_GRAY2BGR)

for cnt in contours:
    area = cv.contourArea(cnt)
    # 过滤过小的噪点轮廓
    if area < 5000:
        continue
    x, y, w, h = cv.boundingRect(cnt)
    # 筛选接近方形的轮廓（长宽比0.8-1.2）
    aspect_ratio = float(w)/h
    if 0.8 < aspect_ratio < 1.2:
        # 验证是否为四边形
        epsilon = 0.04 * cv.arcLength(cnt, True)
        approx = cv.approxPolyDP(cnt, epsilon, True)
        if len(approx) == 4:
            cv.rectangle(canvas, (x,y), (x+w, y+h), (0,0,255), 3)
            break

cv.imshow('bounding box-final', canvas)
cv.waitKey(0)

4. 可选：边缘检测+霍夫直线拟合 #

若色彩过滤效果仍不理想，可结合边缘检测提取方形边界：

gray = cv.cvtColor(im, cv.COLOR_BGR2GRAY)
blur = cv.GaussianBlur(gray, (5,5), 0)
edges = cv.Canny(blur, 50, 150)
# 霍夫直线检测提取水平/垂直直线
lines = cv.HoughLinesP(edges, 1, np.pi/180, threshold=50, minLineLength=100, maxLineGap=20)
# 筛选集装箱的四条边并拟合边界框（需根据实际直线数据调整筛选逻辑）

内容的提问来源于stack exchange，提问作者steve