sangbijaksana / SIFT_stitching

Implementation of image stitching and panorama creation from two images using SIFT.

SIFT Image Stitching

This project implements image stitching and panorama creation using the Scale-Invariant Feature Transform (SIFT) algorithm. It combines two images into a single seamless panorama using Python.

Installation

Required: Python 3.x, OpenCV, NumPy

pip3 install opencv-python numpy

Usage

import cv2
from utils.combine import stitch_image 
from utils.panorama import make_panorama

image_pair_list = [
    ['./images/original/pairs_01_01.jpg', './images/original/pairs_01_02.jpg'],
    ['./images/original/pairs_02_01.png', './images/original/pairs_02_02.png'],
    ['./images/original/pairs_03_01.jpg', './images/original/pairs_03_02.jpg'],
    ['./images/original/pairs_04_01.jpg', './images/original/pairs_04_02.jpg'],
]

idx = 0
for image_pair in image_pair_list:
    idx += 1
    img1_path = image_pair[0]
    img2_path = image_pair[1]

    stitch_image_result = stitch_image(img1_path, img2_path)
    stich_image_name = "./images/combined/pairs_{}.jpg".format(idx)

    cv2.imwrite(stich_image_name, stitch_image_result)
    print("Created stich image:" + stich_image_name)

    panorama_image_result = make_panorama(stitch_image_result)
    panorama_image_name = "./images/panorama/pairs_{}.jpg".format(idx)
    
    cv2.imwrite(panorama_image_name, panorama_image_result)
    print("Created panorama image:" + panorama_image_name)

Example

Example of image striching combining two images and creating a panorama.

Original Image

Image 1	Image 2

Combined Image

Panorama Image

License

BSD-3-Clause License. See LICENSE.

Acknowledgments

References: D. G. Lowe, "Object recognition from local scale-invariant features," Proceedings of the Seventh IEEE International Conference on Computer Vision, Kerkyra, Greece, 1999, pp. 1150-1157 vol.2, doi: 10.1109/ICCV.1999.790410.