assets/preprocessing/align_manhattan.py

from .utils_from_LGT_Net import *
import sys
import numpy as np
import cv2
import os
from PIL import Image
import glob
import json
from natsort import natsorted
from tqdm import tqdm

def config_setup():
	config = {}
	config["home_param"] = "<scene>/"
	return config

def main():
    config = config_setup()
    print(f"Now Processing: {config["home_param"]}...")
    input_folder = f"{config["home_param"]}/RGB"
    output_folder = f"{config["home_param"]}/RGB_mh_aligned"
    os.makedirs(output_folder, exist_ok=True)

    input_files = natsorted(glob.glob(f"{input_folder}/*_rgb.png"))
    mat_dict = {}
    mat_dict["data"] = []

    for input_file in tqdm(input_files):

        # disable OpenCV3's non thread safe OpenCL option
        cv2.ocl.setUseOpenCL(False)

        # Read image
        img_ori = np.array(Image.open(input_file))

        olines, vp, views, edges, panoEdge, score, angle = panoEdgeDetection(img_ori,
                                                                                qError=0.7,
                                                                                refineIter=3)

        img, R = rotatePanorama(img_ori / 255.0, vp[2::-1])
        
        file_name = input_file.split("/")[-1].split(".")[0]
        file_path = f"{output_folder}/{file_name}_aligned.png"
        Image.fromarray((img * 255).astype(np.uint8)).save(file_path)

        each_dict = {"input_file": input_file, "output_file": file_path, "rotation_matrix": R.tolist()}
        mat_dict["data"].append(each_dict)

    with open(f'{output_folder}/rotation_matrix.json', 'w') as f:
        json.dump(mat_dict, f, indent=2)

if __name__ == "__main__":
    main()