diff options
| author |  Chris Xiong <chirs241097@gmail.com>
| 2022-08-30 20:20:39 -0400 |
|---|---|---|
| committer | Chris Xiong <chirs241097@gmail.com>
| 2022-08-30 20:20:39 -0400 |
| commit | 6e389e62d381f65602a890f03b32bcd41ad983d9 (patch) | |
| tree | 0cbf05a7b2d4852c13a8ff2f6f6414b5cbe97a56 /README.md | |
| parent | 8cbd8ff531bf2c1c98ee8141981c95d6950a8175 (diff) | |
| download | deduper-6e389e62d381f65602a890f03b32bcd41ad983d9.tar.xz | |
Add proper readme and license.
Diffstat (limited to 'README.md')
| -rw-r--r-- | README.md | 24 |
1 files changed, 21 insertions, 3 deletions
@@ -1,7 +1,25 @@ # Deduper -(take two) +Deduper finds you similar images on your local filesystem. -still a work in progress +Deduper ... -license tbd. all rights reserved atm. +- finds similar images, regardless of differences in dimensions and compression artifects. +- does pretty much the same thing as [libpuzzle](https://github.com/jedisct1/libpuzzle). +- is an implementation of the same article as libpuzzle. +- uses fast routines provided by OpenCV whenever possible. +- is up to 3 times faster than libpuzzle *. +- provides a tool to manage duplicate images on your local filesystem efficiently. +- has an image indexer to provide fast reverse image searching locally **. +- has a C++ API that provides low level and high level functionalities. + +*: When using a sub-sliced, signature hash assisted implementation to search for +duplicate images in a dataset containing 4000 images, against a plain, distance +comparing implementation. Both implementations are multi-threaded, run on a dual +core mobile processor with hyperthreading. + +**: Planned. + +## License + +Mozilla Public License 2.0 |