PIL imagedraw， imagefont工具使用实战

        利用paddleocr字符识别模块，做字符检测，以下是paddleocr提供的示例代码：

from paddleocr import PaddleOCR, draw_ocr

# Paddleocr目前支持的多语言语种可以通过修改lang参数进行切换
# 例如`ch`, `en`, `fr`, `german`, `korean`, `japan`
ocr = PaddleOCR(use_angle_cls=True, lang="ch")  # need to run only once to download and load model into memory
img_path = './imgs/11.jpg'
result = ocr.ocr(img_path, cls=True)
for idx in range(len(result)):
    res = result[idx]
    for line in res:
        print(line)

# 显示结果
from PIL import Image
result = result[0]
image = Image.open(img_path).convert('RGB')
boxes = [line[0] for line in result]
txts = [line[1][0] for line in result]
scores = [line[1][1] for line in result]
im_show = draw_ocr(image, boxes, txts, scores, font_path='./fonts/simfang.ttf')
im_show = Image.fromarray(im_show)
im_show.save('result.jpg')

这段代码运用paddle内置的权重，对示例图片进行检测，显示结果为原图和生成图片的拼接，图例如下： 

 在原图上，对检测字符加上选中框，在生成图片上排练打印出检测得到字符以及置信度。

但由于项目应用时，GUI界面显示有限，会对拼接图片压缩，以至于看不出识别信息，所以想把输出改一下，不要原图，只要检测输出字符。

 

代码改动如下：

from pathlib import Path
from paddleocr import PaddleOCR, draw_ocr
from PIL import Image
import os
os.environ["KMP_DUPLICATE_LIB_OK"] = "TRUE"
# os.environ["OMP_NUM_THREADS"] = 1
# Paddleocr目前支持的多语言语种可以通过修改lang参数进行切换
# 例如`ch`, `en`, `fr`, `german`, `korean`, `japan`


"""改进图片显示，不显示原图，只显示检测字符"""

file = Path(__file__).resolve()
root = file.parents[1]
def ximing_ocr():
    image_folder_path = os.path.join(root, 'sm', 'data_image')
    # save_path=os.path.join(root, 'sm', 'result')
    for file in os.listdir(image_folder_path):
        if file.endswith((".jpg", ".jpeg", ".png", "bmp")):
            image_path = os.path.join(image_folder_path, file)
            print(image_path)
            # 读取图像大小
            image = Image.open(image_path)
            Width, Height = image.size
            blank_image = Image.new('RGB', (20, Height), color='white')

            # img = image_path[]
            ocr = PaddleOCR(use_angle_cls=True, lang="ch")  # need to run only once to download and load model into memory
            result = ocr.ocr(image_path, cls=True)
            for idx in range(len(result)):
                res = result[idx]
                for line in res:
                    print(line)
            global txts
            # 显示结果
            # 如果本地没有simfang.ttf，可以在doc/fonts目录下下载
            result = result[0]
            image = blank_image
            boxes = [line[0] for line in result]
            txts = [line[1][0] for line in result]
            scores = [line[1][1] for line in result]
            im_show = draw_ocr(image, boxes, txts, scores, font_path='doc/fonts/simfang.ttf')
            im_show = Image.fromarray(im_show)
            im_show.save('result.jpg')

            return txts

ximing_ocr()

用image.new生成一块新的画布，颜色设置为白色，宽度为20（可根据个人喜好），代替原图位置与后续生成的字符画布拼接，结果如下：

 

 此时，字符信息可以清晰看出。

但是忽觉太丑，为什么不利用一些工具，去做一做美化显示呢！

优化一：改进图片显示，在原图上显示框，和字符

 

代码实例：

def ximing_ocr():
    global txt
    image_folder_path = os.path.join(root, 'sm', 'data_image')
    save_path = os.path.join(root, 'sm', 'result')
    for file in os.listdir(image_folder_path):
        if file.endswith((".jpg", ".jpeg", ".png", "bmp")):
            image_path = os.path.join(image_folder_path, file)

            ocr = PaddleOCR(use_angle_cls=True, lang="ch")
            result = ocr.ocr(image_path, cls=True)

            # Load the image
            image = Image.open(image_path).convert('RGB')
            draw = ImageDraw.Draw(image)

            for idx in range(len(result)):
                res = result[idx]
                for line in res:
                    # Extract information
                    box, (txt, score) = line[0], line[1]

                    # Draw bounding box
                    box = [tuple(point) for point in box]
                    draw.polygon(box, outline=(0, 255, 0), width=10)

                    # Draw text and score
                    font = ImageFont.truetype("./Gidole-Regular.ttf", size=200)  # You can use other fonts if needed
                    draw.text((50, 50), f"{txt} ({score:.2f})", fill=(0, 255, 0), font=font)
                    # print(txt)
                    # print(score)
            # Save the image with annotations
            save_path = os.path.join(save_path, f"annotated_result.jpg")
            image.save(save_path)

加入一个for循环嵌套用于取用字符信息， box框固定值为10，字符显示大小固定值200

ImageFont有一个要提的点，若是用默认字符，不可更改大小，这里用了truetype字符显示

：字体包地址

找到.ttf结尾的即为字体包，放到项目文件夹下，引用路径即可使用。

如果这里使用，图片像素值还不错，那么应该不会有啥大问题，若是比较低（低于400像素），就会：

 哈哈哈，固定字符值显然不能兼容不同大小图片。

优化二：改进字符显示兼容性问题，字符大小根据图片长宽显示

代码：

"""改进字符显示兼容性问题，字符大小根据图片长宽显示"""
def ximing_ocr():
    global txt
    image_folder_path = os.path.join(root, 'sm', 'data_image')
    save_path = os.path.join(root, 'sm', 'result')
    for file in os.listdir(image_folder_path):
        if file.endswith((".jpg", ".jpeg", ".png", "bmp")):
            image_path = os.path.join(image_folder_path, file)

            ocr = PaddleOCR(use_angle_cls=True, lang="ch")
            result = ocr.ocr(image_path, cls=True)

            # Load the image
            image = Image.open(image_path).convert('RGB')
            draw = ImageDraw.Draw(image)

            # Calculate font size based on image size
            image_width, image_height = image.size
            font_size = int(min(image_width, image_height) * 0.10)
            font = ImageFont.truetype("./Gidole-Regular.ttf", size=font_size)

            for idx in range(len(result)):
                res = result[idx]
                for line in res:
                    # Extract information
                    box, (txt, score) = line[0], line[1]

                    # Draw bounding box
                    box = [tuple(point) for point in box]
                    draw.polygon(box, outline=(0, 255, 0), width=10)

                    # # Draw text and score
                    # font = ImageFont.truetype("./Gidole-Regular.ttf", size=20)  # You can use other fonts if needed
                    draw.text((50, 50), f"{txt} ({score:.2f})", fill=(0, 255, 0), font=font)
                    # print(txt)
                    # print(score)
            # Save the image with annotations
            save_path = os.path.join(save_path, f"annotated_result.jpg")
            image.save(save_path)

效果：

 文字检测，可能是受字体包的影响。

加入优化："""改进框宽度自适应显示"""

"""改进字符显示兼容性问题，多字符识别时显示字符往下兼容"""

"""改进单个图片检测问题，可以兼容检测整个文件夹"""