Skip to content
Switch branches/tags
Go to file

Latest commit


Git stats


Failed to load latest commit information.
Latest commit message
Commit time


imgrep is a command-line utility in Go to search for keywords found within images.

coverage goreport


imgrep depends on Tesseract.

  • On Fedora: sudo dnf install tesseract-devel
  • On Debian: sudo apt-get install libtesseract-dev
  • On macOS: brew install tesseract



# Fedora
sudo dnf install tesseract tesseract-devel leptonica-devel golang

# Debian
sudo apt-get install tesseract-ocr libtesseract-dev libleptonica-dev golang


  1. Install Go
  2. Install homebrew
brew install tesseract

Get imgrep

Make sure your $GOPATH is set, then run:

# fetch src and install binary
go get -u
go install


imgrep like grep, searches file contents for text. imgrep works exclusively on images however, using text extracted using OCR as the search haystack.

imgrep comes with two interfaces; a CLI as one might expect, and a web-UI graphical front-end.


Because OCR is expensive to the CPU, imgrep can pre-process and index files by keywords stored to a database. This database is queried unless specified otherwise. To preindex an entire directory (including subdirectories):

imgrep init # pre-process and index image files in working directory

By default, imgrep uses this database of pre-indexed files to perform simple queries. Because of the nature of OCR, picked up keywords may not be accurate, so a sort of "fuzzy-search" is employed here. To mimic the original usage of grep, imgrep queries are case-sensitive. The -i option is provided to ignore case specifiers of query strings — it is nearly always recommended to run your queries with this option.

To check preindexed directories:

imgrep search -i QUERY

To use imgrep without checking against the database of preindexed files, simply call

imgrep search -n -i QUERY

Like the grep family of functions, imgrep is useful with Unix-pipes:

# Example: Count the number of images that contain the first line of a
#          plain-text file
head -n1 myfile | imgrep search -n -i - | wc -l

# Example: Open the first file that matches a search
xdg-open "$(imgrep s learn | head -n1)"

Web UI

See imgrep-web.


imgrep is free software licensed under the MIT license.

Copyright (c) 2017 Keefer Rourke and Ivan Zhang.

See LICENSE for details.