import gradio as gr
import pickle
import numpy as np
import cv2
import imutils
from PIL import Image
from cv2 import CascadeClassifier
from fastai.vision.all import *
import os

os.system('wget https://www.dropbox.com/s/q2498c2ww9q2yx0/noticiasFotos.zip')
os.system('unzip noticiasFotos.zip')

def hamming(a, b):
	return bin(int(a) ^ int(b)).count("1")

# Cargamos el árbol y los hashes
tree = pickle.loads(open("treeHaar.pickle", "rb").read())
hashes = pickle.loads(open('hashesHaar.pickle', "rb").read())
detector = CascadeClassifier(cv2.data.haarcascades + 'haarcascade_frontalface_alt.xml')

def convert_hash(h):
	return int(np.array(h, dtype="float64"))
	
def dhash(image, hashSize=8):
    gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
    resized=cv2.resize(gray, (hashSize+1, hashSize))
    diff = resized[:, 1:] > resized[:, :-1]
    return sum([2 ** i for (i, v) in enumerate(diff.flatten()) if v])

def face_detector(image):
    image = np.array(PILImage.create(image))
    image = imutils.resize(image, width=500)
    gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
    l=[]
    rects = detector.detectMultiScale(gray, scaleFactor=1.05,
        minNeighbors=5, minSize=(30, 30),
        flags=cv2.CASCADE_SCALE_IMAGE)
    for (x, y, w, h) in rects:
        l.append(image[y:y+h, x:x+w])
    return l

def searchAll(img):
    sol = []
    for (i, face) in enumerate(face_detector(img)):
        sol = sol + search(face)
    return sol

# Definimos una función que se encarga de llevar a cabo cada búsqueda
def search(img, dist=10):
    queryHash = convert_hash(dhash(img))
    treeResults = sorted(tree.get_all_in_range(queryHash, dist))
    return [Image.open(hashes.get(h, [])[9:]) for (d, h) in treeResults]
    
# Creamos la interfaz y la lanzamos. 
gr.Interface(fn=searchAll, inputs=gr.inputs.Image(), outputs=gr.Gallery()).launch(share=False)