Blue skin in images

[redtoad]

Originally reported by Anonymous

---

I've successfully pulled images through the api package, however the skin color of people in the images are blue. Any ideas as to why this is and what can be done to rectify it?

---

• Bitbucket: https://bitbucket.org/basti/python-amazon-product-api/issue/69

[redtoad]

Original comment by Sebastian Rahlf (Bitbucket: basti, GitHub: basti)

---

Please provide code to reproduce problem.

[redtoad]

Original comment by berd5 (Bitbucket: berd5, GitHub: Unknown)

---

Here is what I'm working with:

#!python

#!/usr/bin/python

# Import package
from amazonproduct import API
from lxml import etree
import cv2
from skimage import io
import os
from itertools import izip
import httplib2
from BeautifulSoup import BeautifulSoup, SoupStrainer
import re 
import requests


# Initiate api wrapper 
api = API(locale="us")


# Api request
item_img = api.item_search('FashionMen', BrowseNode=1045558, Sort='popularity-rank', paginate=False, ResponseGroup='Images')


# Extract complete results link from API response
results_url = etree.tostring(item_img.Items.MoreSearchResultsUrl, pretty_print=True, method='text')
print results_url


# print results_url
results_pages = item_img.Items.TotalResults
print results_pages
asin = []


# Extract search result link
def extract_results(response):
    soup = BeautifulSoup(response)
    for div in soup.findAll('div', {'class' : 's-item-container'}):
        a = div.findAll('a')[0]
        if a.attrs[1][1].startswith('https://'):
            link = str(a.attrs[1][1])
            asin.append(re.findall(r'dp/(\w+)', link))
            
    
# Initiate connection and obtain response from results link
http = httplib2.Http()
headers = {'User-Agent': 'Safari//5.33.4'}
status, response = http.request(results_url, headers = headers)
extract_results(response)
soup = BeautifulSoup(response)
# Extract link to next page
for a in soup.findAll('a', {'id' : 'pagnNextLink'}):
    next_page = 'https://www.amazon.com'+a['href']
    pg_no = re.findall(r'ref=lp_1045558_pg_(\w+)', next_page)
    print(pg_no)
    pg_no = int(pg_no[0].encode("ascii"))
    print type(pg_no)
    while pg_no != 397:
        status, response = http.request(next_page)
        extract_results(response)
        next_soup = BeautifulSoup(response)
        for a in next_soup.findAll('a', {'id' : 'pagnNextLink'}, href = True):
            next_page = 'https://www.amazon.com'+a['href']
            pg_no = re.findall(r'ref=sr_pg_(\w+)', next_page)
            pg_no = int(pg_no[0].encode("ascii"))
            print next_page

print len(asin)


# Save images
os.chdir("/men")
for i in xrange(len(asin)):
    try:
        number = str(asin[i]).strip("['']")
        image = api.item_lookup(number, ResponseGroup='Images')
        url = etree.tostring(image.Items.Item.LargeImage.URL, pretty_print=True, method='text')
        img = io.imread(url)
        cv2.imwrite('{}.jpg'.format(number), img)
    except:
        continue