5 行代码,用 Python 批量做海量小姐姐的素描图

你好,我是悦创。这次呢,我给大家带来的是 50行代码,生成一张素描图。让自己也是一个素描“大师”。那废话不多说,我们直接先来看看效果吧。

上图的右边就是我们的效果,那具体有哪些步骤呢?

1. 流程分析

对于上面的流程来说是非常简单的,接下来我们来看看具体的实现。

2. 具体实现

安装所需要的库:

pip install opencv-python

导入所需要的库:

import cv2

编写主体代码也是非常的简单的,代码如下:

import cv2
SRC = 'images/image_1.jpg'

image_rgb = cv2.imread(SRC)
image_gray = cv2.cvtColor(image_rgb, cv2.COLOR_BGR2GRAY)
image_blur = cv2.GaussianBlur(image_gray, ksize=(21, 21), sigmaX=0, sigmaY=0)
image_blend = cv2.divide(image_gray, image_blur, scale=255)
cv2.imwrite('result.jpg', image_blend)

那上面的代码其实并不难,那接下来为了让小伙伴们能更好的理解,我编写了如下代码:

"""
project = 'Code', file_name = 'study.py', author = 'AI悦创'
time = '2020/5/19 8:35', product_name = PyCharm, 公众号:AI悦创
code is far away from bugs with the god animal protecting
    I love animals. They taste delicious.
"""
import cv2

# 原图路径
SRC = 'images/image_1.jpg'

# 读取图片
image_rgb = cv2.imread(SRC)
# cv2.imshow('rgb', image_rgb) # 原图
# cv2.waitKey(0)
# exit()
image_gray = cv2.cvtColor(image_rgb, cv2.COLOR_BGR2GRAY)
# cv2.imshow('gray', image_gray) # 灰度图
# cv2.waitKey(0)
# exit()
image_bulr = cv2.GaussianBlur(image_gray, ksize=(21, 21), sigmaX=0, sigmaY=0)
cv2.imshow('image_blur', image_bulr) # 高斯虚化
cv2.waitKey(0)
exit()

# divide: 提取两张差别较大的线条和内容
image_blend = cv2.divide(image_gray, image_bulr, scale=255)
# cv2.imshow('image_blend', image_blend) # 素描
cv2.waitKey(0)
# cv2.imwrite('result1.jpg', image_blend)

那上面的代码,我们是在原有的基础上添加了,一些实时展示的代码,来方便同学们理解。

其实有同学会问,我用软件不就可以直接生成素描图吗?

那程序的好处是什么?

程序的好处就是如果你的图片量多的话,这个时候使用程序批量生成也是非常方便高效的。

这样我们的就完成,把小姐姐的图片变成了素描,skr~。

3. 百度图片爬虫+生成素描图

不过,这还不是我们的海量图片,为了达到海量这个词呢,我写了一个百度图片爬虫,不过本文不是教如何写爬虫代码的,这里我就直接放出爬虫代码,符和软件工程规范:

# Crawler.Spider.py
import re
import os
import time
import collections
from collections import namedtuple

import requests
from concurrent import futures
from tqdm import tqdm
from enum import Enum

BASE_URL = 'https://image.baidu.com/search/acjson?tn=resultjson_com&ipn=rj&ct=201326592&is=&fp=result&queryWord={keyword}&cl=2&lm=-1&ie=utf-8&oe=utf-8&adpicid=&st=-1&z=&ic=&hd=&latest=©right=&word={keyword}&s=&se=&tab=&width=&height=&face=0&istype=2&qc=&nc=1&fr=&expermode=&force=&pn={page}&rn=30&gsm=&1568638554041='

HEADERS = {
    'Referer': 'http://image.baidu.com/search/index?tn=baiduimage&ipn=r&ct=201326592&cl=2&lm=-1&st=-1&fr=&sf=1&fmq=1567133149621_R&pv=&ic=0&nc=1&z=0&hd=0&latest=0©right=0&se=1&showtab=0&fb=0&width=&height=&face=0&istype=2&ie=utf-8&sid=&word=%E5%A3%81%E7%BA%B8',
    'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/75.0.3770.100 Safari/537.36',
    'X-Requested-With': 'XMLHttpRequest', }


class BaiDuSpider:
    def __init__(self, max_works, images_type):
        self.max_works = max_works
        self.HTTPStatus = Enum('Status', ['ok', 'not_found', 'error'])
        self.result = namedtuple('Result', 'status data')
        self.session = requests.session()
        self.img_type = images_type
        self.img_num = None
        self.headers = HEADERS
        self.index = 1

    def get_img(self, img_url):
        res = self.session.get(img_url)
        if res.status_code != 200:
            res.raise_for_status()

        return res.content

    def download_one(self, img_url, verbose):
        try:
            image = self.get_img(img_url)
        except requests.exceptions.HTTPError as e:
            res = e.response
            if res.status_code == 404:
                status = self.HTTPStatus.not_found
                msg = 'not_found'
            else:
                raise
        else:
            self.save_img(self.img_type, image)
            status = self.HTTPStatus.ok
            msg = 'ok'

        if verbose:
            print(img_url, msg)

        return self.result(status, msg)

    def get_img_url(self):
        urls = [BASE_URL.format(keyword=self.img_type, page=page) for page in self.img_num]
        for url in urls:
            res = self.session.get(url, headers=self.headers)
            if res.status_code == 200:
                img_list = re.findall(r'"thumbURL":"(.*?)"', res.text)
                # 返回出图片地址,配合其他函数运行
                yield {img_url for img_url in img_list}
            elif res.status_code == 404:
                print('-----访问失败,找不到资源-----')
                yield None
            elif res.status_code == 403:
                print('*****访问失败,服务器拒绝访问*****')
                yield None
            else:
                print('>>> 网络连接失败 <<<')
                yield None

    def download_many(self, img_url_set, verbose=False):
        if img_url_set:
            counter = collections.Counter()
            with futures.ThreadPoolExecutor(self.max_works) as executor:
                to_do_map = {}
                for img in img_url_set:
                    future = executor.submit(self.download_one, img, verbose)
                    to_do_map[future] = img
                done_iter = futures.as_completed(to_do_map)

            if not verbose:
                done_iter = tqdm(done_iter, total=len(img_url_set))
            for future in done_iter:
                try:
                    res = future.result()
                except requests.exceptions.HTTPError as e:
                    error_msg = 'HTTP error {res.status_code} - {res.reason}'
                    error_msg = error_msg.format(res=e.response)
                except requests.exceptions.ConnectionError:
                    error_msg = 'ConnectionError error'
                else:
                    error_msg = ''
                    status = res.status

                if error_msg:
                    status = self.HTTPStatus.error

                counter[status] += 1

                if verbose and error_msg:
                    img = to_do_map[future]
                    print('***Error for {} : {}'.format(img, error_msg))
            return counter
        else:
            pass

    def save_img(self, img_type, image):
        with open('{}/{}.jpg'.format(img_type, self.index), 'wb') as f:
            f.write(image)
        self.index += 1

    def what_want2download(self):
        # self.img_type = input('请输入你想下载的图片类型,什么都可以哦~ >>> ')
        try:
            os.mkdir(self.img_type)
        except FileExistsError:
            pass
        img_num = input('请输入要下载的数量(1位数代表30张,列如输入1就是下载30张,2就是60张):>>> ')
        while True:
            if img_num.isdigit():
                img_num = int(img_num) * 30
                self.img_num = range(30, img_num + 1, 30)
                break
            else:
                img_num = input('输入错误,请重新输入要下载的数量>>> ')

    def main(self):
        # 获取图片类型和下载的数量
        total_counter = {}
        self.what_want2download()
        for img_url_set in self.get_img_url():
            if img_url_set:
                counter = self.download_many(img_url_set, False)
                for key in counter:
                    if key in total_counter:
                        total_counter[key] += counter[key]
                    else:
                        total_counter[key] = counter[key]

            else:
                # 可以为其添加报错功能
                pass

        time.sleep(.5)
        return total_counter


if __name__ == '__main__':
    max_works = 20
    bd_spider = BaiDuSpider(max_works)
    print(bd_spider.main())
# Sketch_the_generated_code.py
import cv2
def drawing(src, id=None):
    image_rgb = cv2.imread(src)
    image_gray = cv2.cvtColor(image_rgb, cv2.COLOR_BGR2GRAY)
    image_blur = cv2.GaussianBlur(image_gray, ksize=(21, 21), sigmaX=0, sigmaY=0)
    image_blend = cv2.divide(image_gray, image_blur, scale=255)
    cv2.imwrite(f'Drawing_images/result-{id}.jpg', image_blend)
# image_list.image_list_path.py
import os
from natsort import natsorted

IMAGES_LIST = []

def image_list(path):
    global IMAGES_LIST
    for root, dirs, files in os.walk(path):
        # 按文件名排序
        # files.sort()
        files = natsorted(files)
        # 遍历所有文件
        for file in files:
            # 如果后缀名为 .jpg
            if os.path.splitext(file)[1] == '.jpg':
                # 拼接成完整路径
                # print(file)
                filePath = os.path.join(root, file)
                print(filePath)
                # 添加到数组
                IMAGES_LIST.append(filePath)
    return IMAGES_LIST
# main.py
import time

from Sketch_the_generated_code import drawing
from Crawler.Spider import BaiDuSpider
from image_list.image_list_path import image_list
import os

MAX_WORDS = 20

if __name__ == '__main__':
    # now_path = os.getcwd()
    # img_type = 'ai'
    img_type = input('请输入你想下载的图片类型,什么都可以哦~ >>> ')
    bd_spider = BaiDuSpider(MAX_WORDS, img_type)
    print(bd_spider.main())
    time.sleep(10) # 这里设置睡眠时间,让有足够的时间去添加,这样读取就,去掉或者太短会报错,所以
    for index, path in enumerate(image_list(img_type)):
        drawing(src = path, id = index)

所以最终的目录结构如下所示:

C:.
│  main.py
│  Sketch_the_generated_code.py
│
├─Crawler
│  │  Spider.py
│  │
│  └─__pycache__
│          Spider.cpython-37.pyc
│
├─drawing
│  │  result.jpg
│  │  result1.jpg
│  │  Sketch_the_generated_code.py
│  │  study.py
│  │
│  ├─images
│  │      image_1.jpg
│  │
│  └─__pycache__
│          Sketch_the_generated_code.cpython-37.pyc
│
├─Drawing_images
├─image_list
│  │  image_list_path.py
│  │
│  └─__pycache__
│          image_list_path.cpython-37.pyc
│
└─__pycache__
        Sketch_the_generated_code.cpython-37.pyc

至此,全部代码已经完成。

AI悦创·创造不同!
AI悦创 » 5 行代码,用 Python 批量做海量小姐姐的素描图

Leave a Reply