在图像对齐之前首先进行亮度/对比度校正，反之亦然？(Brightness/contrast correction first before image alignment or vice versa?)

 需要你的帮助以下。 我有3个相同的静止物体图像，每隔30分钟捕获一次。 我已将相机和物体锁定在适当的位置，房间仍然是黑暗的，但我仍然得到了3种不同的曝光/亮度/伽玛图像，物体也移动了一点。 Image1 Image2 
 
 我试图做的是调整对准，亮度/伽马/对比度第2和第3张图像参考第1张图像。 我有如何使用ECCtransform方法对齐图像的解决方案： 
 
import os, sys
import cv2
from PIL import Image
import numpy as np

path = "C:\\Users\\xxxx\\Desktop\\"
path1 = "C:\\Users\\xxxx\\Desktop\\aligned\\"

def alignment():
    for i in range(1,4):
        # Read the images to be aligned
        im1 =  cv2.imread(path + '1.png')
        im2 =  cv2.imread(path + '%d.png' %(i))
        print(i)

        # Convert images to grayscale
        im1_gray = cv2.cvtColor(im1,cv2.COLOR_BGR2GRAY)
        im2_gray = cv2.cvtColor(im2,cv2.COLOR_BGR2GRAY)

        # Find size of image1
        sz = im1.shape

        # Define the motion model
        warp_mode = cv2.MOTION_TRANSLATION

        # Define 2x3 or 3x3 matrices and initialize the matrix to identity
        if warp_mode == cv2.MOTION_HOMOGRAPHY :
            warp_matrix = np.eye(3, 3, dtype=np.float32)
        else :
            warp_matrix = np.eye(2, 3, dtype=np.float32)
        # Specify the number of iterations.
        number_of_iterations = 5000;

        # Specify the threshold of the increment
        # in the correlation coefficient between two iterations
        termination_eps = 1e-10;

        # Define termination criteria
        criteria = (cv2.TERM_CRITERIA_EPS | cv2.TERM_CRITERIA_COUNT, number_of_iterations,  termination_eps)

        # Run the ECC algorithm. The results are stored in warp_matrix.
        (cc, warp_matrix) = cv2.findTransformECC(im1_gray, im2_gray, warp_matrix, warp_mode, criteria)


        if warp_mode == cv2.MOTION_HOMOGRAPHY :
            # Use warpPerspective for Homography 
            im2_aligned = cv2.warpPerspective (im2, warp_matrix, (sz[1],sz[0]), flags=cv2.INTER_LINEAR + cv2.WARP_INVERSE_MAP)
        else :
            # Use warpAffine for Translation, Euclidean and Affine
            im2_aligned = cv2.warpAffine(im2, warp_matrix, (sz[1],sz[0]), flags=cv2.INTER_LINEAR + cv2.WARP_INVERSE_MAP);

        cv2.imwrite(path1 + "%d.png" % (i), im2_aligned )

alignment()
 
 我的问题是，哪种方式更好？ 序列是否重要？ 以第一张图片为标准参考： 
 
 我应该首先执行transformECC图像对齐，以便我可以准确地调整图像的亮度/曝光吗？ 
 
 要么 
 
 我应该先调整亮度/曝光，以便准确对齐照片？ 
 
 我仍在考虑参考第一张图像来调整我的第二和第三图像亮度/曝光。 任何想法都受到欢迎和赞赏!!!! 

need your help with the following. I have 3 same stationary object image that were captured per 30mins interval. I have lock the camera and object in position and the room was left dark but still I ended up with 3 different exposure/brightness/gamma images and the object move a little as well. Image1 Image2
 
What I was trying to do is to adjust alignment, brightness/gamma/contrast 2nd and 3rd images in reference to 1st images. I have solution on how to align the image using ECCtransform method below: 
 
import os, sys
import cv2
from PIL import Image
import numpy as np

path = "C:\\Users\\xxxx\\Desktop\\"
path1 = "C:\\Users\\xxxx\\Desktop\\aligned\\"

def alignment():
    for i in range(1,4):
        # Read the images to be aligned
        im1 =  cv2.imread(path + '1.png')
        im2 =  cv2.imread(path + '%d.png' %(i))
        print(i)

        # Convert images to grayscale
        im1_gray = cv2.cvtColor(im1,cv2.COLOR_BGR2GRAY)
        im2_gray = cv2.cvtColor(im2,cv2.COLOR_BGR2GRAY)

        # Find size of image1
        sz = im1.shape

        # Define the motion model
        warp_mode = cv2.MOTION_TRANSLATION

        # Define 2x3 or 3x3 matrices and initialize the matrix to identity
        if warp_mode == cv2.MOTION_HOMOGRAPHY :
            warp_matrix = np.eye(3, 3, dtype=np.float32)
        else :
            warp_matrix = np.eye(2, 3, dtype=np.float32)
        # Specify the number of iterations.
        number_of_iterations = 5000;

        # Specify the threshold of the increment
        # in the correlation coefficient between two iterations
        termination_eps = 1e-10;

        # Define termination criteria
        criteria = (cv2.TERM_CRITERIA_EPS | cv2.TERM_CRITERIA_COUNT, number_of_iterations,  termination_eps)

        # Run the ECC algorithm. The results are stored in warp_matrix.
        (cc, warp_matrix) = cv2.findTransformECC(im1_gray, im2_gray, warp_matrix, warp_mode, criteria)


        if warp_mode == cv2.MOTION_HOMOGRAPHY :
            # Use warpPerspective for Homography 
            im2_aligned = cv2.warpPerspective (im2, warp_matrix, (sz[1],sz[0]), flags=cv2.INTER_LINEAR + cv2.WARP_INVERSE_MAP)
        else :
            # Use warpAffine for Translation, Euclidean and Affine
            im2_aligned = cv2.warpAffine(im2, warp_matrix, (sz[1],sz[0]), flags=cv2.INTER_LINEAR + cv2.WARP_INVERSE_MAP);

        cv2.imwrite(path1 + "%d.png" % (i), im2_aligned )

alignment()
 
My question are, which is the better way to go? Is the sequence matter? By taking first image as standard reference:
 
Should I perform transformECC image alignment first so that I can adjust the brightness/exposure of my images accurately?
 
OR
 
I should adjust the brightness/exposure first so that I aligned the photo accurately?
 
I am still thinking way to adjust my 2nd and 3rd images brightness/exposure with reference to 1st images. Any ideas is welcomed and appreciated !!!! 

原文：https://stackoverflow.com/questions/45179304