meitarc / FinalProject

1 stars 3 forks source link

Compare Two Images and determine if they're similar #6

Open GuyOhayon opened 5 years ago

GuyOhayon commented 5 years ago
GuyOhayon commented 5 years ago

We need to go over the algorithms:

  1. SIFT - compares between 2 images, while considering scale-space.
  2. BRUTE FORCE
  3. ORB
  4. SURF
  5. FAST

Relevant pdfs: an image Features - Meaning.pdf Harris Corner Detection.pdf Introduction to SIFT - Scale Invariant Feature Transform.pdf Introduction to SURF (Speeded-Up Robust Features).pdf OpenCV panorama stitching - PyImageSearch.pdf Shi-Tomasi Corner Detector & Good Features to Track.pdf

GuyOhayon commented 5 years ago

using SIFT algorithm to find Scaleable-Space corners. this is relevant because moving cars might have to consider far-away corners coming closer as the car moves forward.

image

GuyOhayon commented 5 years ago

finished extracting the probability of two images being similar.

created the function similarityProbability(img1, img2) in the project's python library.

attached the updated library. project2018Draft.py

GuyOhayon commented 5 years ago

updated project's python library

project2018Draft.py compare(1).py

GuyOhayon commented 5 years ago

after sitting with Shay, we need to compare two images not only based on their pixel count, but also by the similarity of their points of interest.

see the Meeting Summery #3. [סיכום פגישה שלישית- 6.1.19.docx]