import numpy as np
import math
from scipy.signal import convolve2d
import cv2

__all__ = [
    "set_requires_grad", "compute_psnr", 'compute_ssim'

]

# 反向传播开关
def set_requires_grad(parameters, requires_grad=False):
    # 注意这里默认是False
    if not isinstance(parameters, list):
        parameters = [parameters]  # 包装一下（对传入的单个参数）
    for param in parameters:
        if param is not None:
            param.requires_grad = requires_grad

def compute_psnr(img_src, img_guess):
    img1_array = np.array(img_src)
    img2_array = np.array(img_guess)
    print(img1_array.dtype)
    print(img2_array.dtype)
    mse = np.mean((img1_array - img2_array)**2)
    PIXEL_MAX = 255.0
    if mse == 0:
        return float('inf')
    else:
        return 20 * math.log10(PIXEL_MAX / math.sqrt(mse))



def matlab_style_gauss2D(shape=(3,3),sigma=0.5):
    """
    2D gaussian mask - should give the same result as MATLAB's
    fspecial('gaussian',[shape],[sigma])
    """
    m,n = [(ss-1.)/2. for ss in shape]
    y,x = np.ogrid[-m:m+1,-n:n+1]
    h = np.exp( -(x*x + y*y) / (2.*sigma*sigma) )
    h[ h < np.finfo(h.dtype).eps*h.max() ] = 0
    sumh = h.sum()
    if sumh != 0:
        h /= sumh
    return h

def filter2(x, kernel, mode='same'):
    return convolve2d(x, np.rot90(kernel, 2), mode=mode)

def compute_ssim(img1, img2, k1=0.01, k2=0.03, win_size=11, L=255):
    # img转array
    img1_array = np.array(img1, dtype=np.uint8)
    img2_array = np.array(img2, dtype=np.uint8)
    print(img1_array.shape)
    # 转灰度图
    gray_img1 = cv2.cvtColor(img1_array, cv2.COLOR_BGR2GRAY)
    gray_img2 = cv2.cvtColor(img2_array, cv2.COLOR_BGR2GRAY)

    print(gray_img1.shape)

    if not gray_img1.shape == gray_img2.shape:
        raise ValueError("Input Imagees must have the same dimensions")
    if len(gray_img1.shape) > 2:
        raise ValueError("Please input the images with 1 channel")

    M, N = gray_img1.shape
    C1 = (k1 * L) ** 2
    C2 = (k2 * L) ** 2
    window = matlab_style_gauss2D(shape=(win_size, win_size), sigma=1.5)
    window = window / np.sum(np.sum(window))

    if gray_img1.dtype == np.uint8:
        gray_img1 = np.double(gray_img1)
    if gray_img2.dtype == np.uint8:
        gray_img2 = np.double(gray_img2)

    mu1 = filter2(gray_img1, window, 'valid')
    mu2 = filter2(gray_img2, window, 'valid')
    mu1_sq = mu1 * mu1
    mu2_sq = mu2 * mu2
    mu1_mu2 = mu1 * mu2
    sigma1_sq = filter2(gray_img1 * gray_img1, window, 'valid') - mu1_sq
    sigma2_sq = filter2(gray_img2 * gray_img2, window, 'valid') - mu2_sq
    sigmal2 = filter2(gray_img1 * gray_img2, window, 'valid') - mu1_mu2

    ssim_map = ((2 * mu1_mu2 + C1) * (2 * sigmal2 + C2)) / ((mu1_sq + mu2_sq + C1) * (sigma1_sq + sigma2_sq + C2))

    return np.mean(np.mean(ssim_map))