jeresig/ukiyoe-web

A rewrite of the Ukiyo-e.org website.

ukiyoe-web's Issues

seek some advice converting fig to docker-compose

Hi there!

I'm trying to 'refresh' things a bit. Working on making a docker-compose file instead of fig.yml

Where can I mirror some of this stuff:

     - MONGODB_URL=mongodb://localhost:27017/ukiyoe
     - ELASTICSEARCH_URL=http://localhost:9200
     - BASE_DATA_DIR=/data
     - BASE_DATA_URL=http://data.ukiyo-e.org/
     - ME_PASSWORD=USERNAME
     - ME_USERNAME=PASSWORD
     - ME_SERVER=matchengine.tineye.com
     - S3_BUCKET=data.ukiyo-e.org
     - S3_KEY=KEY
     - S3_SECRET=SECRET
     - S3_STATIC_BUCKET=static.ukiyo-e.org

I'm uncertain which have data one would need for the site to have a little bit of content.

Use Fast and Brief algorithms for feature detection and description on images

Recently I watched a video about some project you were indexing images based on similarity and I was wondering if some computer vision algorithms may fit into what you need. One of them is Fast for feature detection and the other is Brief for feature description.

Those utilities are available on http://github.com/eduardolundgren/tracking.js implemented in JavaScript and work with Node, you can see more on http://trackingjs.com/docs.html#utilities.

A good example of the feature matching is available here http://trackingjs.com/examples/brief.html.