求助：如何编写条件语句判断图像ROI像素值范围并输出状态

Hey there! Let's sort out that condition for your ROI pixel check. The main thing to note is that your ROI variable is a 2D array of pixels (not a single color value), so we can't just compare it directly to a range. We need to check how many pixels in the ROI fall within your target color bounds, then use that to decide which status to show.

Here's the corrected code with explanations:

import cv2
import numpy as np

frame = cv2.imread('frame 8 sec.jpg')
ROI = frame[450:500, 380:391]

# Define your color range (note: OpenCV uses BGR order, not RGB!)
lower_bound = np.array([105, 105, 105], dtype=np.uint8)
upper_bound = np.array([255, 255, 255], dtype=np.uint8)

# Create a mask where pixels in the range are white (255), others black (0)
mask = cv2.inRange(ROI, lower_bound, upper_bound)

# Choose your validation logic:
# Option 1: Check if ALL pixels in ROI are within the range
# all_valid = np.all(mask == 255)

# Option 2: Check if MOST pixels (e.g., >50%) are within the range (more practical for real images)
total_pixels = ROI.shape[0] * ROI.shape[1]
valid_pixels = np.count_nonzero(mask)
valid_ratio = valid_pixels / total_pixels
is_valid = valid_ratio > 0.5  # Adjust this threshold based on your needs

# Set status and text color
if is_valid:
    status = 'left'
    text_color = (0, 255, 0)  # Green for valid status
else:
    status = 'write'
    text_color = (0, 0, 255)  # Red for invalid status

# Add text to the frame (removed unused alpha channel from color tuple)
cv2.putText(frame, f"Status {status}", (10, 20), cv2.FONT_HERSHEY_SIMPLEX, 1, text_color, 3)

cv2.imshow('img', frame)
cv2.waitKey(0)
cv2.destroyAllWindows()

Key Details to Know: #

• OpenCV Color Format: cv2.imread loads images in BGR order, so your gray-scale bounds (105,105,105) work perfectly here since all channels are equal.
• Mask Usage: cv2.inRange is the standard tool for detecting pixels within a color range—it generates a binary mask to highlight valid pixels.
• Flexible Validation:
  • Use the "all pixels valid" check if you need strict adherence to the color range.
  • Use the ratio check for real-world images, where minor noise or small color variations are common. Adjust the 0.5 threshold to make the check more or less strict.
• Text Rendering: The alpha channel in your original color tuple wasn't necessary for standard OpenCV windows, so I removed it to avoid confusion.

内容的提问来源于stack exchange，提问作者Khaled