本次代码开发调试基于简化版的，相关代码说明见文档：
二次解读的博客-CSDN博客
一、直接使用高仿车牌生成车牌
高仿车牌生成的代码见：+生成较真实的车牌号码图片的博客-CSDN博客
从网上下载了一批真实的质量较高的车牌，每张车牌都有车牌号码的文字信息，如下图所示：
首先，根据上图中的车牌号码生成高仿的车牌，作为训练集的A目录 。
然后，将真实车牌作为训练集的B目录 。
我们将上述训练集通过进行训练，200次迭代后，效果依然不佳 。
效果大致如下：
然后将迭代次数提高到400次，以及增加的个数，效果依然不佳 。
所以提出了疑问：
1、的示例中的图片（比如苹果橘子）是不涉及文字的，那么这种涉及文字的是否可行？
2、自己训练集的图片是矩形的，而代码中会为正方形，这块是否有影响 。
带着这两个疑问，我们重新设计了训练集进行测试 。
二、最简单的带文字的GAN试验
源码中图片最后会被到256*256，所以这次我们直接生成256*256的图片 。
源图片和目标图片样式分别见下面两张图：
生成代码：
#!/usr/bin/env python#coding=utf-8import cv2import osimport numpy as npfrom PIL import ImageDraw, ImageFont, Imagechar_province_list = ["京","沪","津","渝","冀","晋","蒙","辽","吉","黑","苏","浙","皖","闽","赣","鲁","豫","鄂","湘","粤","桂","琼","川","贵","云","藏","陕","甘","青","宁","新"]char_alphbet_list = ["0","1","2","3","4","5","6","7","8","9","A","B","C","D","E","F","G","H","J","K","L","M","N","P","Q","R","S","T","U","V","W","X","Y","Z"]def generate_license_plate_number():"""随机生成车牌号码:return:"""lp_list = []for province in char_province_list:for _ in range(100):lp_str = provincenumber_list = np.random.choice(char_alphbet_list, 6)lp_str += "".join(number_list)lp_list.append(lp_str)return lp_listdef write_chinese(img, font_type, font_size,color, position, content):# 图像从OpenCV格式转换成PIL格式img_PIL = Image.fromarray(cv2.cvtColor(img, cv2.COLOR_BGR2RGB))# 字体字体*.ttc的存放路径一般是： /usr/share/fonts/opentype/noto/ 查找指令locate *.ttcfont = ImageFont.truetype(font_type, font_size)# 字体颜色# 文字输出位置# 输出内容draw = ImageDraw.Draw(img_PIL)draw.text(position, content, font=font, fill=color)# 转换回OpenCV格式img_OpenCV = cv2.cvtColor(np.asarray(img_PIL), cv2.COLOR_RGB2BGR)return img_OpenCVdef generate_plate_rectangle(plate_str):"""绘制矩形边框:param plate_str::return:"""width = 256height = 256img = np.ones((height, width, 3), dtype=np.uint8)img *= 255# white backgroundcv2.rectangle(img, (20, 20), (236, 236), (255, 0, 0), 4)img = write_chinese(img, 'font/SimHei.ttf', 42, (0, 0, 0), (50, 100), plate_str)return imgdef generate_plate_filled_rectangle(plate_str):width = 256height = 256img = np.ones((height, width, 3), dtype=np.uint8)img *= 255# white backgroundcv2.rectangle(img, (20, 20), (236, 236), (255, 0, 0), -1)img = write_chinese(img, 'font/SimHei.ttf', 42, (0, 0, 0), (50, 100), plate_str)return imgdef generate_plate_tuple(plate_str, index):dir_path = "D:\\workspace\\ncz-python-algo\\com\\ncz\\algo\\license-plate-generator\\simple_images"path_A = os.path.join(dir_path, "A", str(index) + "_" + plate_str + ".png")path_B = os.path.join(dir_path, "B", str(index) + "_" + plate_str + ".png")img1 = generate_plate_rectangle(plate_str)img2 = generate_plate_filled_rectangle(plate_str)cv2.imencode('.png', img1)[1].tofile(path_A)cv2.imencode('.png', img2)[1].tofile(path_B)# # 生成个别图片# license_plate_str = '浙A5B5T3'# generate_plate_tuple(license_plate_str, 1)# 随机生成一堆图片license_plate_list = generate_license_plate_number()for index, license_plate in enumerate(license_plate_list):print(index + 1, license_plate)generate_plate_tuple(license_plate, index + 1)