Python 在弱光图像中应用高斯模糊和Canny边缘检测后的白线检测

我有两张图像，一张是明亮的，边缘检测效果很好，白线清晰可见

然而，还有一张照片是在暗处拍摄的，光线很暗，白线很暗，几乎看不见

原始图像

img = mpimg.imread(filenm)

# first convert to grayscale
grayscale = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY)

#apply gaussian blur before canny edge detection
kernel_size=125
gaussian_blurred = cv2.GaussianBlur(grayscale,(kernel_size, kernel_size), 0)

# Canny edge detection threshold ratio 1:3
low_threshold = 50
high_threshold = 150
edges = cv2.Canny(gaussian_blurred, low_threshold, high_threshold)


plt.imshow(edges, cmap='Greys_r')

弱光：

好光：

Canny边缘检测

微光

好光

我可以将哪些图像过滤器应用于这两幅图像，以便为这两幅图像检测到白线。我尝试过灰度和高斯模糊，然后再进行canny边缘检测，但效果不太好

我的代码

img = mpimg.imread(filenm)

# first convert to grayscale
grayscale = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY)

#apply gaussian blur before canny edge detection
kernel_size=125
gaussian_blurred = cv2.GaussianBlur(grayscale,(kernel_size, kernel_size), 0)

# Canny edge detection threshold ratio 1:3
low_threshold = 50
high_threshold = 150
edges = cv2.Canny(gaussian_blurred, low_threshold, high_threshold)


plt.imshow(edges, cmap='Greys_r')

---

最好尝试自适应thresholding@min2bro您的代码仅在高亮度下生成黑屏，请再次检查您的代码，尤其是内核大小

kernel\u size=125

！我没有和你一样的优势！