group_photos_by_date_and_distance.py

import datetime
import exifread
import glob
import sys
import os
import shutil
import fractions
import pyexiv2

from geopy.distance import vincenty

# 500 sec 200 meters
# "*.JPG" grouped 500 0 200

max_gap_seconds = int(sys.argv[3])
min_meters = int(sys.argv[4])
max_meters = int(sys.argv[5])

print "Group photos by date and distance"
print "Max gap (seconds): %d, Min dist (meters): %d, Max dist (meters): %d" % (max_gap_seconds, min_meters, max_meters)

def process_file(filepath):

    # Open image file for reading (binary mode)
    f = open(filepath, 'rb')

    # Return Exif tags
    tags = exifread.process_file(f)

    

    
    if "GPS GPSLatitude" in tags and "GPS GPSLongitude" in tags and "EXIF DateTimeOriginal" in tags:

        date_str = tags["EXIF DateTimeOriginal"].values

        #print date_str
        date_taken = datetime.datetime.strptime(date_str, '%Y:%m:%d %H:%M:%S')
        lat_deg = tags["GPS GPSLatitude"].values
        lng_deg = tags["GPS GPSLongitude"].values


        lat = float(float(lat_deg[0].num)/lat_deg[0].den) + float(float(lat_deg[1].num)/lat_deg[1].den)/60 + (float(lat_deg[2].num)/lat_deg[2].den)/3600
        lng = float(float(lng_deg[0].num)/lng_deg[0].den) + float(float(lng_deg[1].num)/lng_deg[1].den)/60 + (float(lng_deg[2].num)/lng_deg[2].den)/3600


        location = (lat, lng)
    else:
        date_taken = None
        location = None

    #print date_taken, date_str
    return {
        "filepath": filepath,
        "date_taken": date_taken,
        "location": location
    }



filepaths = glob.glob(sys.argv[1])
destination_filepath = sys.argv[2]

photo_list = []
for filepath in filepaths:
    photo_details = process_file(filepath)
    if(photo_details["date_taken"]):
        photo_list.append(photo_details)

photo_list.sort(key=lambda r: r["date_taken"]) #

groups = {}

group_date = photo_list[0]["date_taken"]
groups[group_date] = [ photo_list[0] ]

distance_from_last_photo = 0.0

for i in range(1, len(photo_list)):
    photo = photo_list[i]
    previous_photo = photo_list[i-1]

    gap_seconds = (photo["date_taken"] - previous_photo["date_taken"]).total_seconds()

    dist = vincenty(photo["location"], previous_photo["location"]).meters
    distance_from_last_photo = distance_from_last_photo + dist

    print "File: %s Date: %s Seconds: %f  Meters: %f" % (photo["filepath"], photo["date_taken"], gap_seconds, dist)
    print "DIST %f" % distance_from_last_photo

    
    if distance_from_last_photo < min_meters:
        print "File %s is too close" % photo["filepath"]  
    elif gap_seconds < max_gap_seconds and dist < max_meters:
        groups[group_date].append(photo)
        distance_from_last_photo = 0
    else:
        group_date = photo["date_taken"]
        groups[group_date] = [ photo ]
        distance_from_last_photo = 0
        print "New group: %s" % photo["date_taken"]

for date in groups:
    dirname = date.strftime("%Y-%m-%d_%H%M%S")
    dirpath = "%s/%s" % (destination_filepath, dirname)

    if not os.path.exists(dirpath):
        os.makedirs(dirpath)

    print "GROUP", dirpath

    photos = groups[date]


    for photo in photos:
        head, tail = os.path.split(photo["filepath"])
        dest_filepath = "%s/%s" % (dirpath, tail)
        
        print dest_filepath
        shutil.copy2(photo["filepath"], dest_filepath)
        #print photo["filepath"], photo["date_taken"]


    #    destination_filepath

#print glob.glob(sys.argv[1])