找到
19
篇与
网站源码
相关的结果
-
python批量去除图片文字水印 #!/usr/bin/env python # -*- coding:utf-8 -*- # 需要安装的库 # pip install paddlepaddle -i https://mirrors.aliyun.com/pypi/simple/ # pip install paddleocr -i https://mirrors.aliyun.com/pypi/simple/ # pip install cv2 -i https://mirrors.aliyun.com/pypi/simple/ # pip install numpy -i https://mirrors.aliyun.com/pypi/simple/ # pip install Pillow -i https://mirrors.aliyun.com/pypi/simple/ import os import cv2 import numpy as np from PIL import Image from paddleocr import PaddleOCR, draw_ocr class DeleteImageWatermark: def __init__(self): pass def distinguish_string(self, img_path, lang='ch'): """ 得到文字识别结果列表 img_path: 图片路径 lang: 默认为识别中文 return: 返回所有被识别到的文字文本框坐标、文字内容和置信度 如：[ [[[1415.0, 977.0], [1482.0, 977.0], [1482.0, 1001.0], [1415.0, 1001.0]], ('小红书', 0.868567168712616)], [[[1441.0, 1001.0], [1493.0, 1001.0], [1493.0, 1024.0], [1441.0, 1024.0]], ('小红书', 0.9620211124420166)] ] """ orc = PaddleOCR(use_angle_cls=True, lang=lang) result = orc.ocr(img_path, cls=True) return result def save_distinguish_result(self, result, img_path, save_path): """ 将识别文字的结果输出图片 """ 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(save_path) def delete_watermark(self, result_list, kw_list, img_path, delete_path): """ 将符合目标的水印，模糊化处理 """ # 获取所有符合目标的文本框位置 text_axes_list = [] for line in result_list: for kw in kw_list: if kw in line[1][0]: min_width = int(min(line[0][0][0], line[0][3][0])) max_width = int(max(line[0][1][0], line[0][2][0])) min_hight = int(min(line[0][0][1], line[0][1][1])) max_hight = int(max(line[0][2][1], line[0][3][1])) text_axes_list.append([min_width, min_hight, max_width, max_hight]) break # 去除水印 delt = 10 # 文本框范围扩大 img = cv2.imread(img_path, 1) tmp_delete_path = delete_path.split('.')[0] + '_test.' + delete_path.split('.')[1] # 临时图片地址 cv2.imwrite(tmp_delete_path, img) for text_axes in text_axes_list: img = cv2.imread(tmp_delete_path, 1) hight, width = img.shape[0:2] # 截取图片 min_width = text_axes[0] - delt if text_axes[0] - delt >= 0 else 0 min_hight = text_axes[1] - delt if text_axes[1] - delt >= 0 else 0 max_width = text_axes[2] + delt if text_axes[2] + delt <= width else width max_hight = text_axes[3] + delt if text_axes[3] + delt <= hight else hight cropped = img[min_hight:max_hight, min_width:max_width] # 裁剪坐标为[y0:y1, x0:x1] cv2.imwrite(delete_path, cropped) # 保存截取的图片 imgSY = cv2.imread(delete_path, 1) # 图片二值化处理，把[200,200,200]-[250,250,250]以外的颜色变成0 start_rgb = 200 thresh = cv2.inRange(imgSY, np.array([start_rgb, start_rgb, start_rgb]), np.array([250, 250, 250])) # 创建形状和尺寸的结构元素 kernel = np.ones((3, 3), np.uint8) # 设置卷积核3*3全是1；将当前的数组作为图像类型来进⾏各种操作，就要转换到uint8类型 # 扩展待修复区域 hi_mask = cv2.dilate(thresh, kernel, iterations=10) # 膨胀操作，白色区域增大，iterations迭代次数 specular = cv2.inpaint(imgSY, hi_mask, 5, flags=cv2.INPAINT_TELEA) # imgSY：输入8位1通道或3通道图像。 # hi_mask：修复掩码，8位1通道图像。非零像素表示需要修复的区域。 # specular：输出与imgSY具有相同大小和类型的图像。 # 5：算法考虑的每个点的圆形邻域的半径。 # flags：NPAINT_NS基于Navier-Stokes的方法、Alexandru Telea的INPAINT_TELEA方法 cv2.imwrite(delete_path, specular) # 覆盖图片 imgSY = Image.open(delete_path) img = Image.open(tmp_delete_path) img.paste(imgSY, (min_width, min_hight, max_width, max_hight)) img.save(tmp_delete_path) os.remove(delete_path) os.rename(tmp_delete_path, delete_path) def has_kw(self, result_list, kw_list): """ 图片是否包含目标水印，返回匹配到的文字列表 """ result_str_list = [] for line in result_list: for kw in kw_list: if kw in line[1][0]: result_str_list.append(line[1][0]) break return result_str_list def main(kw_list, img_path, result_path): """ kw_list: 需要识别的文字列表 img_path: 输入的图片地址 result_path: 输出去水印的结果图片地址 """ d = DeleteImageWatermark() # 识别文字 result = d.distinguish_string(img_path) for line in result: print(line) # 打印识别结果：识别到的文字文本框坐标、文字内容和置信度 # 显示文字识别结果 d.save_distinguish_result(result, img_path, os.path.dirname(__file__) + '/test_01.jpg') # 是否含有指定水印 result_str_list = d.has_kw(result, kw_list) if len(result_str_list) > 0: # 删除水印 d.delete_watermark(result, kw_list, img_path, result_path) print('共有 %d 处水印，都已删除成功!' % len(result_str_list)) return True else: print('无指定水印！') return False if __name__ == '__main__': # 图片地址 #path = os.path.dirname(__file__) path=os.getcwd() img_path = path + '/去除水印.jpg' result_path = path + "/result.jpg" # 删除指定水印 kw_list = [ '快手', '抖音', '网易云'] main(kw_list, img_path, result_path) -
-
-
-
腾讯视频VIP一次解析api源码 第一步：代码第43行 $VIPCookies = '';// 必填VIP Cookies 使用会员帐号cookies第二步：上传到服务器，使用方法：https://域名/文件名/?url=视频地址返回格式JSON，需要用到M3U8播放器播放！代码如下！[hide]<?php /** * 腾讯视频VIP一次解析api源码 */ //使用教程 //参数 url=视频地址 type=输出方式(json/dplayer) error_reporting(0); // allow cross header('Access-Control-Allow-Origin: *'); header('Access-Control-Allow-Methods: GET'); $url = isset($_GET['url']) ? $_GET['url'] : ''; $type = isset($_GET['type']) ? $_GET['type'] : ''; $api = new Tvideo(); $data = $api->Tvideo_api($url); if ($type == 'dplayer') { $file_path = __DIR__ . '/src/DPlayer.html'; if (file_exists($file_path)) { echo file_get_contents($file_path); } exit; }else{ header('content-type:application/json;charset=utf8'); exit(json_encode($data,480)); } //腾讯视频解析 class Tvideo{ public function Tvideo_api($url){ $content = $this->curl($url); preg_match('#data-vid="(\w+)" data-cid="(\w+)"#iU',$content,$id); $cid = !isset($id[2])? id[2]; if (basename($url,'.html')==$cid) { $vid = !isset($id[1])? id[1]; }else { $vid=basename($url,'.html'); } $curl ='https://vv.video.qq.com/getinfo?encver=2&defn=shd&platform=10801&otype=ojson&sdtfrom=v4138&appVer=7&dtype=3&vid='.$vid.'&newnettype=1'; $VIPCookies = '';// 必填VIP Cookies $JsonInfo = $this->vip_curl($curl,$VIPCookies); $JsonData = json_decode($JsonInfo,true); $vurl = $JsonData["vl"]["vi"][0]['ul']['ui'][3]['url'].$JsonData["vl"]["vi"][0]['ul']['ui'][3]['hls']['pt']; if($JsonData['em']==61){ $result = array( 'code' => 1, 'msg' => '解析失败！' ); }else{ $result = array( 'code' => 0, 'msg' => '解析成功！', 'url' => $vurl, 'm3u8_to'=>$JsonData["vl"]["vi"][0]['ul']['ui'][3]['url'] ); } return $result; } public function curl($url){ $curl = curl_init(); $header = array( "X-FORWARDED-FOR:".long2ip(mt_rand(1884815360, 1884890111)), "CLIENT-IP:".long2ip(mt_rand(1884815360, 1884890111)), "X-Real-IP:".long2ip(mt_rand(1884815360, 1884890111)), "Accept:application/json, text/javascript, */*; q=0.01", "Accept-Language:zh-CN,zh;q=0.9", "Connection: Keep-Alive", "Referer:https://v.qq.com", "User-Agent:Mozilla/5.0 (Windows NT 6.3; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/63.0.3239.132 Safari/537.36 QIHU 360EE", "X-Requested-With:XMLHttpRequest", ); curl_setopt($curl, CURLOPT_URL, $url); curl_setopt($curl, CURLOPT_SSL_VERIFYPEER, FALSE); curl_setopt($curl, CURLOPT_SSL_VERIFYHOST, FALSE); curl_setopt($curl, CURLOPT_RETURNTRANSFER,true); curl_setopt($curl, CURLOPT_HTTPHEADER, $header); curl_setopt($curl, CURLOPT_FOLLOWLOCATION,1); curl_setopt($curl, CURLOPT_HEADER,0); curl_setopt($curl, CURLOPT_CONNECTTIMEOUT, 10); curl_setopt($curl, CURLOPT_TIMEOUT, 10); return curl_exec($curl); } public function getC($str, $leftStr, $rightStr) { $left = strpos($str, $leftStr); $right = strpos($str, $rightStr,$left); if($left < 0 or $right < $left) return ''; return substr($str, $left + strlen($leftStr), $right-$left-strlen($leftStr)); } public function vip_curl($url,$cookie='') { $header = array ( 0 => 'Accept: text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.9', 1 => 'Accept-Language: zh-CN,zh;q=0.9,en;q=0.8,zh-HK;q=0.7', 2 => 'Cache-Control: max-age=0', 3 => 'Connection: keep-alive', 4 => 'Sec-Fetch-Dest: document', 5 => 'Sec-Fetch-Mode: navigate', 6 => 'Sec-Fetch-Site: none', 7 => 'Sec-Fetch-User: ?1', 8 => 'Upgrade-Insecure-Requests: 1', 9 => 'User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/104.0.0.0 Safari/537.36', 10 => 'sec-ch-ua: ^\\^', 11 => 'sec-ch-ua-mobile: ?0', 12 => 'sec-ch-ua-platform: ^\\^', 13 =>'Cookie:'.$cookie ); $timeout = 10; $ch = curl_init($url); curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1); curl_setopt($ch, CURLOPT_ENCODING, "gzip"); curl_setopt($ch, CURLOPT_FOLLOWLOCATION, true); if(substr($url, 0, 8) === 'https://') { curl_setopt($ch, CURLOPT_SSL_VERIFYPEER, false); curl_setopt($ch, CURLOPT_SSL_VERIFYHOST, false); } if(!empty($postData)) { curl_setopt($ch, CURLOPT_POST, 1); curl_setopt($ch, CURLOPT_POSTFIELDS, $postData); } if(!empty($cookie)) { $header[] = $cookie; } if(!empty($header)) { curl_setopt($ch, CURLOPT_HTTPHEADER, $header); } curl_setopt($ch, CURLOPT_TIMEOUT, (int)$timeout); $content = curl_exec($ch); if($error = curl_error($ch)) { error_log($error); } curl_close($ch); return $content; } public function rand_ip(){ $ip_long = array( array('607649792', '608174079'), //36.56.0.0-36.63.255.255 array('975044608', '977272831'), //58.30.0.0-58.63.255.255 array('999751680', '999784447'), //59.151.0.0-59.151.127.255 array('1019346944', '1019478015'), //60.194.0.0-60.195.255.255 array('1038614528', '1039007743'), //61.232.0.0-61.237.255.255 array('1783627776', '1784676351'), //106.80.0.0-106.95.255.255 array('1947009024', '1947074559'), //116.13.0.0-116.13.255.255 array('1987051520', '1988034559'), //118.112.0.0-118.126.255.255 array('2035023872', '2035154943'), //121.76.0.0-121.77.255.255 array('2078801920', '2079064063'), //123.232.0.0-123.235.255.255 array('-1950089216', '-1948778497'), //139.196.0.0-139.215.255.255 array('-1425539072', '-1425014785'), //171.8.0.0-171.15.255.255 array('-1236271104', '-1235419137'), //182.80.0.0-182.92.255.255 array('-770113536', '-768606209'), //210.25.0.0-210.47.255.255 array('-569376768', '-564133889'), //222.16.0.0-222.95.255.255 ); $rand_key = mt_rand(0, 14); $huoduan_ip= long2ip(mt_rand($ip_long[$rand_key][0], $ip_long[$rand_key][1])); return $huoduan_ip; } } [/hide]
