EfficientDet数据集可视化

EfficientDet数据集可视化

import cv2
import random
import json, os
from pycocotools.coco import COCO
#from skimage import io
from matplotlib import pyplot as plt

train_json = './datasets/jam/annotations/instances_train.json'
train_path = './datasets/jam/train/'

def visualization_bbox2(num_image, json_path, img_path):
    coco = COCO(json_path)
    list_imgIds = coco.getImgIds(catIds=2) # 获取含有该给定类别的所有图片的id
    print(list_imgIds)
    img = coco.loadImgs(list_imgIds[num_image-1])[0]  # 获取满足上述要求，并给定显示第num幅image对应的dict
    image = cv2.imread(img_path + img['file_name'])  # 读取图像
    image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB) 
    image_name =  img['file_name'] # 读取图像名字
    image_id = img['id'] # 读取图像id
    img_annIds = coco.getAnnIds(imgIds=image_id)
    img_anns = coco.loadAnns(img_annIds)
    print(len(img_annIds))
    for i in range(len(img_annIds)):
        x, y, w, h = img_anns[i]['bbox']  # 读取边框
        label = img_anns[i]['category_id']
        label = coco.loadCats(label)[0]['name']
        image = cv2.rectangle(image, (int(x), int(y)), (int(x + w), int(y + h)), (255, 255, 255), 2)
        cv2.putText(image,str(label),(int(x), int(y)),cv2.FONT_HERSHEY_SIMPLEX,1.2, (255, 255, 255), 2)
    plt.rcParams['figure.figsize'] = (20.0, 20.0)
    plt.imshow(image)
    plt.show()
 
 visualization_bbox2(10, train_json, train_path)