Google-Image-Query-Optimizer
This program takes in 2 image queries. The first one is the baseline image query. The second one is a query which is used to compare the first one against. The difference between simply querying Google Images and using this program is this program uses structural similarity of images using image manipulation in SciKit to figure out the best IMAGE REPRESENTATION of your query compared to all the responses retrieved, while Google Images uses PageRank or essentially popularity indexes to give you the most relevant images to your query. That is why this program is an optimized version of Google Image Querying.
Getting Started
You can install the package with pip3 install google-image-query-optimizer. You might want to prepend sudo to that line. You can now run the script with “google-image-query-optimizer”.
If you do not install using pip, you can also download this repo and do the following: You might have to install the following: pip3 install numpy pip3 install Pillow pip3 install -U scikit-image
To run: python3 genart.py
Prerequisites
pip3 install numpy pip3 install Pillow pip3 install -U scikit-image
Example
Input
Enter Google Image Query (GIQ): Iron Man
Image Identifier (second GIQ): professional classy
Output
Expected Output or Image that will open is collage of top 3 most representative responses to what you queried:
Built With
- SciKit - Structural similarity
- Hardikvasa - Downloading images from Google
- PIL - helps with dealing with images, opening them, resizing etc.
Authors
- Abhishek Mangla