Datasets:
File size: 5,756 Bytes
cf1cafc |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 |
# Download panorama images (0, 90, 180, 270 stitching) with duplicate detection and 404 error handling
# Input: CSV file (e.g., selected_data.csv) containing panoid field
# Output: Panorama images saved to save_dir
import os
from PIL import Image
import requests
import io
import csv
from tqdm import tqdm
import dotenv
import pandas as pd
import re
from time import sleep
import random
dotenv.load_dotenv("../.env")
API_KEY = os.getenv('GOOGLE_MAP_API_KEY')
base_url = 'https://maps.googleapis.com/maps/api/streetview'
save_dir = '/home/xiuying.chen/jingpu/data/panoramas'
csv_path = '/home/xiuying.chen/jingpu/data/exist_pano.csv'
headings = {
'north': '0',
'east': '90',
'south': '180',
'west': '270',
}
def get_street_view_image(pano_id):
params = {
'size': '640x640',
'pano': pano_id,
'key': API_KEY,
'pitch': '0',
'return_error_code': 'true'
}
images = []
error_panoids = []
# Download images facing east, south, west, and north
for direction, heading in headings.items():
params['heading'] = heading
sleep(random.randint(1, 5))
response = requests.get(base_url, params=params)
print(f'{direction} request status:', response.status_code)
if response.status_code == 200:
images.append(Image.open(io.BytesIO(response.content)))
image_filename = f'{pano_id}_{direction}.jpg'
save_path = os.path.join(save_dir, image_filename)
with open(save_path, 'wb') as f:
f.write(response.content)
print(f'Saved {image_filename}')
elif response.status_code == 404:
if pano_id not in error_panoids:
error_panoids.append(pano_id)
print(f'Failed to get the {direction} image, panoid: {pano_id} does not exist')
# If there are panoids causing a 404, write them to 404panoid.csv
if error_panoids:
error_df = pd.DataFrame(error_panoids, columns=['panoid'])
error_df.to_csv('404panoid.csv', mode='a', header=not os.path.exists('404panoid.csv'), index=False)
# Stitch images together to form a panorama
try:
# Calculate the dimensions of the final panorama image
width = sum(img.width for img in images)
height = images[0].height
# Create a new image and stitch the downloaded images together
panorama = Image.new('RGB', (width, height))
x_offset = 0
for img in images:
panorama.paste(img, (x_offset, 0))
x_offset += img.width
# Save the panorama image
output_path = os.path.join(save_dir, f'{pano_id}_panoramic.jpg')
panorama.save(output_path, quality=95)
# Update exist_pano.csv with the new panoid information
new_row = {'panoid': pano_id, 'address': save_dir}
existing_data = pd.read_csv(csv_path) if os.path.exists(csv_path) else pd.DataFrame(columns=['panoid', 'address'])
existing_data = existing_data._append(new_row, ignore_index=True)
existing_data.to_csv(csv_path, index=False)
return output_path
except Exception as e:
print(f"Error downloading panorama: {str(e)}")
return None
def check_panoramas_exist(panoid):
"""Check if the given panoid has already been crawled"""
csv_path = '/home/xiuying.chen/jingpu/data/exist_pano.csv'
# Read existing panoid data
if not os.path.exists(csv_path):
return False # If the file does not exist, return False
existing_data = pd.read_csv(csv_path)
# Check if the panoid exists in the 'panoid' column
return panoid in existing_data['panoid'].values
def read_pano_ids(file_path):
pano_ids = []
with open(file_path, 'r', encoding='utf-8') as f:
reader = csv.DictReader(f) # Use DictReader to read the CSV file
for row in reader:
pano_id = row['panoID']
pano_ids.append((pano_id)) # Add the panoid to the list
return pano_ids
def load_exist_pano(scaned_dir):
"""Read .jpg files in the specified directory and update exist_pano.csv"""
# Read existing panoid data
csv_path = '/home/xiuying.chen/jingpu/data/exist_pano.csv'
if os.path.exists(csv_path):
existing_data = pd.read_csv(csv_path)
else:
print("exist_pano.csv not found, quit")
return
# Iterate over all .jpg files in the specified directory
for filename in os.listdir(scaned_dir):
if filename.endswith('.jpg'):
# Use regex to extract the panoid by removing the directional suffix
match = re.match(r'^(.*?)(?:_east|_north|_south|_west|_panoramic)\.jpg$', filename)
if match:
panoid = match.group(1) # Extract the panoid part
# Check whether the panoid already exists
if panoid not in existing_data['panoid'].values:
# Create a new row and save the file address
new_row = {'panoid': panoid, 'address': scaned_dir}
existing_data = existing_data._append(new_row, ignore_index=True)
# Write the updated data back to the CSV file
existing_data.to_csv(csv_path, index=False)
if __name__ == '__main__':
pano_ids = read_pano_ids("GT.csv")
for pano_id in tqdm(pano_ids, desc="Downloading street view images"):
# Check whether the image has already been downloaded
if check_panoramas_exist(pano_id):
# print(f'{pano_id} already exists')
continue
else:
# print(f'{pano_id} does not exist')
sleep(5)
get_street_view_image(pano_id)
print(f'{pano_id} downloaded')
|