如何使用Python结合PIL分离3位验证码中的单个数字

1. 背景转为纯白色 #

先将图像转为灰度模式，通过阈值过滤把浅灰色背景替换为纯白色，保留深色数字：

# 转为灰度图
gray_im = im.convert('L')
# 阈值调整：根据验证码背景亮度设置，这里200适配你的验证码
threshold = 200
white_bg_im = gray_im.point(lambda x: 255 if x > threshold else x)

2. 裁剪多余空白区域 #

定位图像中所有非白色像素的边界，裁剪出仅包含验证码的有效区域：

pixels = white_bg_im.load()
width, height = white_bg_im.size

# 初始化边界值
left, right = width, 0
top, bottom = height, 0

# 遍历像素找有效区域边界
for x in range(width):
    for y in range(height):
        if pixels[x, y] != 255:
            left = min(left, x)
            right = max(right, x)
            top = min(top, y)
            bottom = max(bottom, y)

# 裁剪（注意right和bottom需+1，保证包含最后一个像素）
cropped_im = white_bg_im.crop((left, top, right + 1, bottom + 1))

3. 分割为单个数字图像 #

将裁剪后的图像按宽度三等分，得到三个独立的数字图像：

crop_width = cropped_im.width // 3

# 分别裁剪三个数字
digit1 = cropped_im.crop((0, 0, crop_width, cropped_im.height))
digit2 = cropped_im.crop((crop_width, 0, crop_width * 2, cropped_im.height))
digit3 = cropped_im.crop((crop_width * 2, 0, cropped_im.width, cropped_im.height))

# 保存结果（可选）
digit1.save('digit1.jpg')
digit2.save('digit2.jpg')
digit3.save('digit3.jpg')

完整代码 #

from PIL import Image

im = Image.open("captcha.jpg")

# 背景转纯白
gray_im = im.convert('L')
threshold = 200
white_bg_im = gray_im.point(lambda x: 255 if x > threshold else x)

# 裁剪空白
pixels = white_bg_im.load()
width, height = white_bg_im.size
left, right, top, bottom = width, 0, height, 0

for x in range(width):
    for y in range(height):
        if pixels[x, y] != 255:
            left = min(left, x)
            right = max(right, x)
            top = min(top, y)
            bottom = max(bottom, y)

cropped_im = white_bg_im.crop((left, top, right + 1, bottom + 1))

# 分割数字
crop_width = cropped_im.width // 3
digit1 = cropped_im.crop((0, 0, crop_width, cropped_im.height))
digit2 = cropped_im.crop((crop_width, 0, crop_width * 2, cropped_im.height))
digit3 = cropped_im.crop((crop_width * 2, 0, cropped_im.width, cropped_im.height))

# 保存分割后的图像
digit1.save('digit1.jpg')
digit2.save('digit2.jpg')
digit3.save('digit3.jpg')

内容的提问来源于stack exchange，提问作者ali emili