YOLOV3实现车牌检测

推荐阅读31个Python实战项目教你掌握图像处理，PDF开放下载opencv_contrib扩展模块中文教程pdf，限时领取

cd labelImg-masterbrew install qtbrew install libxml2make qt5py3python labelImg.py

import glob # Contains all .txt files except our listof classestxt_files = [file for file inglob.glob('images/*.txt') if file != 'images/classes.txt']# Read every .txt file and store it'scontent into variable currfor file in txt_files:   with open(file, 'r') as f:       curr = f.read()     # Replace class index 15 with 1 and storeit in a variable newnew = curr.replace('15 ', '1 ') # Once again open every .txt file and makethe replacementfor file in txt_files:   with open(file, 'w') as f:       f.write(new) 

from google.colab import drivedrive.mount('/content/gdrive')!ln -s /content/gdrive/My\ Drive/ /mydrive

！git clonehttps://github.com/AlexeyAB/darknet

%cd darknet!make

！cp cfg / yolov3.cfg cfg / yolov3-train.cfg

!echo -e 'license-plate' >data/obj.names!echo -e 'classes = 1\ntrain =data/train.txt\nvalid = data/test.txt\nnames = data/obj.names\nbackup =/mydrive/yolo-license-plates' > data/obj.data 

！mkdir数据/对象！unzip ../images.zip -d data / obj

import globimages_list = glob.glob('data/obj/*.jpg')with open('data/train.txt', 'w') as f:   f.write('\n'.join(images_list))

！wgethttps://pjreddie.com/media/files/darknet53.conv.74

import cv2import numpy as npimport matplotlib.pyplot as pltclass LicensePlateDetector:   def __init__(self, pth_weights: str, pth_cfg: str, pth_classes: str):       self.net = cv2.dnn.readNet(pth_weights, pth_cfg)       self.classes = []       with open(pth_classes, 'r') as f:           self.classes = f.read().splitlines()       self.font = cv2.FONT_HERSHEY_PLAIN       self.color = (255, 0, 0)       self.coordinates = None       self.img = None       self.fig_image = None       self.roi_image = None                 def detect(self, img_path: str):       orig = cv2.imread(img_path)       self.img = orig       img = orig.copy()       height, width, _ = img.shape       blob = cv2.dnn.blobFromImage(img, 1 / 255, (416, 416), (0, 0, 0),swapRB=True, crop=False)       self.net.setInput(blob)       output_layer_names = self.net.getUnconnectedOutLayersNames()       layer_outputs = self.net.forward(output_layer_names)       boxes = []       confidences = []       class_ids = []        for output in layer_outputs:           for detection in output:                scores = detection[5:]                class_id = np.argmax(scores)                confidence = scores[class_id]                if confidence > 0.2:                    center_x = int(detection[0]* width)                    center_y = int(detection[1]* height)                    w = int(detection[2] *width)                    h = int(detection[3] *height)                    x = int(center_x - w / 2)                    y = int(center_y - h / 2)                     boxes.append([x, y, w, h])                   confidences.append((float(confidence)))                    class_ids.append(class_id)        indexes = cv2.dnn.NMSBoxes(boxes, confidences, 0.2, 0.4)        if len(indexes) > 0:           for i in indexes.flatten():                x, y, w, h = boxes[i]                label = str(self.classes[class_ids[i]])                confidence =str(round(confidences[i],2))                cv2.rectangle(img, (x,y), (x +w, y + h), self.color, 15)                cv2.putText(img, label + ' ' +confidence, (x, y + 20), self.font, 3, (255, 255, 255), 3)       self.fig_image = img       self.coordinates = (x, y, w, h)       return         def crop_plate(self):       x, y, w, h = self.coordinates       roi = self.img[y:y + h, x:x + w]       self.roi_image = roi       return 

lpd = LicensePlateDetector(   pth_weights='yolov3_training_final.weights',   pth_cfg='yolov3_testing.cfg',   pth_classes='classes.txt') # Detect license platelpd.detect('001.jpg') # Plot original image with rectangle aroundthe plateplt.figure(figsize=(24, 24))plt.imshow(cv2.cvtColor(lpd.fig_image, cv2.COLOR_BGR2RGB))plt.savefig('detected.jpg')plt.show() # Crop plate and show cropped platelpd.crop_plate()plt.figure(figsize=(10, 4))plt.imshow(cv2.cvtColor(lpd.roi_image,cv2.COLOR_BGR2RGB)) 

个人微信（如果没有备注不拉群！）请注明：地区+学校/企业+研究方向+昵称


下载1：何恺明顶会分享
在「AI算法与图像处理」公众号后台回复：何恺明，即可下载。总共有6份PDF，涉及 ResNet、Mask RCNN等经典工作的总结分析
下载2：终身受益的编程指南：Google编程风格指南
在「AI算法与图像处理」公众号后台回复：c++，即可下载。历经十年考验，最权威的编程规范！

下载3 CVPR2020
在「AI算法与图像处理」公众号后台回复：CVPR2020，即可下载1467篇CVPR 2020论文
觉得不错就点亮在看吧