rtesseract

rtesseract is an R package providing a fast and flexible interface to the Tesseract OCR engine via Rcpp. It allows you to extract text from images and convert scanned PDFs into searchable PDFs.

✨ Features

OCR on single or multiple images
Convert scanned PDFs to searchable PDFs
Control DPI and rendering options
Cross-platform support (Windows, Linux, macOS)
Direct integration with Tesseract C++ API
Uses pdftoppm (Poppler) for PDF rasterization

📦 Installation

1. Install system dependencies

📌 Dependencies

Tesseract OCR
Leptonica
Poppler (pdftoppm)

Installation Tesseract

run_cmd <- function(cmd, args = character(), error_msg = NULL, echo = TRUE) {
  if (echo) {
    message(">> ", paste(c(cmd, args), collapse = " "))
  }
  status <- system2(cmd, args = args)
  if (!identical(status, 0L)) {
    stop(if (is.null(error_msg)) paste("Command failed:", cmd) else error_msg, call. = FALSE)
  }
  invisible(TRUE)
}

check_cmd <- function(cmd) {
  nzchar(Sys.which(cmd))
}

write_tesseract_env <- function(include_dir,
                                lib_dir,
                                bin_dir = NULL,
                                scope = c("session", "renviron")) {
  scope <- match.arg(scope)

  Sys.setenv(
    TESSERACT_INCLUDE = include_dir,
    TESSERACT_LIB = lib_dir
  )

  if (!is.null(bin_dir)) {
    Sys.setenv(TESSERACT_BIN = bin_dir)
  }

  if (scope == "renviron") {
    renv <- path.expand("~/.Renviron")
    old <- if (file.exists(renv)) readLines(renv, warn = FALSE) else character()

    # retire anciennes lignes
    old <- old[!grepl("^TESSERACT_INCLUDE=", old)]
    old <- old[!grepl("^TESSERACT_LIB=", old)]
    old <- old[!grepl("^TESSERACT_BIN=", old)]

    new_lines <- c(
      old,
      sprintf('TESSERACT_INCLUDE="%s"', normalizePath(include_dir, winslash = "/", mustWork = FALSE)),
      sprintf('TESSERACT_LIB="%s"', normalizePath(lib_dir, winslash = "/", mustWork = FALSE))
    )

    if (!is.null(bin_dir)) {
      new_lines <- c(
        new_lines,
        sprintf('TESSERACT_BIN="%s"', normalizePath(bin_dir, winslash = "/", mustWork = FALSE))
      )
    }

    writeLines(new_lines, renv)
    message("Variables to add in ~/.Renviron")
  }

  message("TESSERACT_INCLUDE = ", Sys.getenv("TESSERACT_INCLUDE"))
  message("TESSERACT_LIB     = ", Sys.getenv("TESSERACT_LIB"))
  message("TESSERACT_BIN     = ", Sys.getenv("TESSERACT_BIN"))

  invisible(TRUE)
}

install_tesseract_linux <- function(use_sudo = TRUE, persist = TRUE) {
  if (!check_cmd("apt-get")) {
    stop("This Linux function is intended for Debian/Ubuntu (apt-get).", call. = FALSE)
  }

  apt_pkgs <- c("tesseract-ocr", "libtesseract-dev", "libleptonica-dev")

  if (use_sudo) {
    run_cmd("sudo", c("apt-get", "update"), "apt-get update failed.")
    run_cmd("sudo", c("apt-get", "install", "-y", apt_pkgs),
            "APT installation of Tesseract failed.")
  } else {
    run_cmd("apt-get", c("update"), "apt-get update failed.")
    run_cmd("apt-get", c("install", "-y", apt_pkgs),
            "APT installation of Tesseract failed")
  }

  include_candidates <- c("/usr/include", "/usr/local/include")
  lib_candidates <- c("/usr/lib/x86_64-linux-gnu", "/usr/lib", "/usr/local/lib")
  bin_candidates <- c("/usr/bin", "/usr/local/bin")

  include_dir <- include_candidates[
    file.exists(file.path(include_candidates, "tesseract", "baseapi.h"))
  ][1]

  lib_dir <- lib_candidates[
    file.exists(file.path(lib_candidates, "libtesseract.so")) |
      file.exists(file.path(lib_candidates, "libtesseract.a")) |
      file.exists(file.path(lib_candidates, "libtesseract.dll.a"))
  ][1]

  bin_dir <- bin_candidates[
    file.exists(file.path(bin_candidates, "tesseract"))
  ][1]

  if (is.na(include_dir) || is.na(lib_dir)) {
    warning("Tesseract is installed, but the include/lib paths were not detected automatically")
    include_dir <- "/usr/include"
    lib_dir <- "/usr/lib"
  }

  if (is.na(bin_dir)) {
    bin_dir <- "/usr/bin"
  }

  write_tesseract_env(
    include_dir = include_dir,
    lib_dir = lib_dir,
    bin_dir = bin_dir,
    scope = if (persist) "renviron" else "session"
  )

  invisible(list(include = include_dir, lib = lib_dir, bin = bin_dir))
}

install_tesseract_macos <- function(persist = TRUE) {
  if (!check_cmd("brew")) {
    stop("Homebrew is not available.", call. = FALSE)
  }

  run_cmd("brew", c("install", "tesseract"),
          "Homebrew installation of Tesseract failed.")

  brew_prefix <- trimws(system2("brew", "--prefix", stdout = TRUE))
  tess_prefix <- trimws(system2("brew", c("--prefix", "tesseract"), stdout = TRUE))

  include_candidates <- c(
    file.path(tess_prefix, "include"),
    file.path(brew_prefix, "include")
  )

  lib_candidates <- c(
    file.path(tess_prefix, "lib"),
    file.path(brew_prefix, "lib")
  )

  bin_candidates <- c(
    file.path(tess_prefix, "bin"),
    file.path(brew_prefix, "bin")
  )

  include_dir <- include_candidates[
    file.exists(file.path(include_candidates, "tesseract", "baseapi.h"))
  ][1]

  lib_dir <- lib_candidates[
    file.exists(file.path(lib_candidates, "libtesseract.dylib")) |
      file.exists(file.path(lib_candidates, "libtesseract.a"))
  ][1]

  bin_dir <- bin_candidates[
    file.exists(file.path(bin_candidates, "tesseract"))
  ][1]

  if (is.na(include_dir) || is.na(lib_dir)) {
    warning("Tesseract is installed, but the include/lib paths were not detected automatically.")
    include_dir <- file.path(brew_prefix, "include")
    lib_dir <- file.path(brew_prefix, "lib")
  }

  if (is.na(bin_dir)) {
    bin_dir <- file.path(brew_prefix, "bin")
  }

  write_tesseract_env(
    include_dir = include_dir,
    lib_dir = lib_dir,
    bin_dir = bin_dir,
    scope = if (persist) "renviron" else "session"
  )

  invisible(list(prefix = tess_prefix, include = include_dir, lib = lib_dir, bin = bin_dir))
}

install_tesseract_windows_msys2 <- function(rtools_root = "C:/rtools45",
                                            repo = "ucrt64",
                                            persist = TRUE) {
  bash <- file.path(rtools_root, "usr", "bin", "bash.exe")
  if (!file.exists(bash)) {
    stop("bash.exe not found in Rtools/MSYS2 ", bash, call. = FALSE)
  }

  pkgs <- switch(
    repo,
    ucrt64 = c("mingw-w64-ucrt-x86_64-tesseract-ocr"),
    clang64 = c("mingw-w64-clang-x86_64-tesseract-ocr"),
    mingw64 = c("mingw-w64-x86_64-tesseract-ocr"),
    stop("repo has to be 'ucrt64', 'clang64' ou 'mingw64'")
  )

  msys_cmd <- sprintf("pacman -S --needed --noconfirm %s", paste(pkgs, collapse = " "))
  run_cmd(bash, c("-lc", shQuote(msys_cmd)),
          "MSYS2 installation of Tesseract failed.")

  prefix <- switch(
    repo,
    ucrt64  = file.path(rtools_root, "ucrt64"),
    clang64 = file.path(rtools_root, "clang64"),
    mingw64 = file.path(rtools_root, "mingw64")
  )

  include_dir <- file.path(prefix, "include")
  lib_dir <- file.path(prefix, "lib")
  bin_dir <- file.path(prefix, "bin")

  hdr <- file.path(include_dir, "tesseract", "baseapi.h")
  if (!file.exists(hdr)) {
    warning("Tesseract header not found at the expected location: ", hdr)
  }

  libs_found <- any(file.exists(file.path(lib_dir, c(
    "libtesseract.dll.a", "libtesseract.a", "tesseract.lib"
  ))))
  if (!libs_found) {
    warning("Tesseract library not found at the expected location: ", lib_dir)
  }

  exe <- file.path(bin_dir, "tesseract.exe")
  if (!file.exists(exe)) {
    warning("Tesseract executable not found at the expected location : ", exe)
  }

  write_tesseract_env(
    include_dir = include_dir,
    lib_dir = lib_dir,
    bin_dir = bin_dir,
    scope = if (persist) "renviron" else "session"
  )

  invisible(list(
    prefix = prefix,
    include = include_dir,
    lib = lib_dir,
    bin = bin_dir,
    packages = pkgs
  ))
}

install_tesseract <- function(...) {
  sys <- Sys.info()[["sysname"]]

  if (identical(sys, "Linux")) {
    return(install_tesseract_linux(...))
  }
  if (identical(sys, "Darwin")) {
    return(install_tesseract_macos(...))
  }
  if (identical(sys, "Windows")) {
    return(install_tesseract_windows_msys2(...))
  }

  stop("OS not supported: ", sys, call. = FALSE)
}

install_tesseract(rtools_root = "C:/rtools45", repo = "ucrt64")

Download tesseract training data

install_tesseract_languages <- function(
    langs = c("eng", "fra", "osd"),
    dest = "C:/rtools45/ucrt64/share/tessdata/",
    overwrite = FALSE
) {
  base_url <- "https://github.com/tesseract-ocr/tessdata_best/raw/main"

  # create dir if necessary
  if (!dir.exists(dest)) {
    dir.create(dest, recursive = TRUE)
  }

  message("Installation in : ", dest)

  for (lang in langs) {
    file <- paste0(lang, ".traineddata")
    url <- paste0(base_url, "/", file)
    destfile <- file.path(dest, file)

    if (file.exists(destfile) && !overwrite) {
      message("✔ Already there : ", file)
      next
    }

    message("⬇ Download : ", file)

    tryCatch({
      download.file(url, destfile, mode = "wb", quiet = TRUE)
      message("✔ OK : ", file)
    }, error = function(e) {
      message("✖ Error for ", file, " : ", e$message)
    })
  }

  # config environnement
  Sys.setenv(TESSDATA_PREFIX = dest)

  message("\nTESSDATA_PREFIX defined at : ", dest)

  # vérification
  message("\nVérification :")
  print(file.exists(file.path(dest, paste0(langs, ".traineddata"))))

  invisible(dest)
}

install_tesseract_languages()

Installation Poppler

run_cmd <- function(cmd, args = character(), error_msg = NULL, echo = TRUE) {
  if (echo) {
    message(">> ", paste(c(cmd, args), collapse = " "))
  }
  status <- system2(cmd, args = args)
  if (!identical(status, 0L)) {
    stop(if (is.null(error_msg)) paste("Command failed:", cmd) else error_msg, call. = FALSE)
  }
  invisible(TRUE)
}

check_cmd <- function(cmd) {
  nzchar(Sys.which(cmd))
}

write_pdftoppm_env <- function(include_dir = NULL,
                               lib_dir = NULL,
                               bin_dir = NULL,
                               scope = c("session", "renviron")) {
  scope <- match.arg(scope)

  if (!is.null(include_dir)) {
    Sys.setenv(POPPLER_INCLUDE = include_dir)
  }
  if (!is.null(lib_dir)) {
    Sys.setenv(POPPLER_LIB = lib_dir)
  }
  if (!is.null(bin_dir)) {
    Sys.setenv(PDFTOPPM_BIN = bin_dir)
  }

  if (scope == "renviron") {
    renv <- path.expand("~/.Renviron")
    old <- if (file.exists(renv)) readLines(renv, warn = FALSE) else character()

    old <- old[!grepl("^POPPLER_INCLUDE=", old)]
    old <- old[!grepl("^POPPLER_LIB=", old)]
    old <- old[!grepl("^PDFTOPPM_BIN=", old)]

    new_lines <- old

    if (!is.null(include_dir)) {
      new_lines <- c(
        new_lines,
        sprintf('POPPLER_INCLUDE="%s"',
                normalizePath(include_dir, winslash = "/", mustWork = FALSE))
      )
    }

    if (!is.null(lib_dir)) {
      new_lines <- c(
        new_lines,
        sprintf('POPPLER_LIB="%s"',
                normalizePath(lib_dir, winslash = "/", mustWork = FALSE))
      )
    }

    if (!is.null(bin_dir)) {
      new_lines <- c(
        new_lines,
        sprintf('PDFTOPPM_BIN="%s"',
                normalizePath(bin_dir, winslash = "/", mustWork = FALSE))
      )
    }

    writeLines(new_lines, renv)
    message("Added variables to ~/.Renviron")
  }

  message("POPPLER_INCLUDE = ", Sys.getenv("POPPLER_INCLUDE"))
  message("POPPLER_LIB     = ", Sys.getenv("POPPLER_LIB"))
  message("PDFTOPPM_BIN    = ", Sys.getenv("PDFTOPPM_BIN"))

  invisible(TRUE)
}

install_pdftoppm_linux <- function(use_sudo = TRUE, persist = TRUE) {
  if (!check_cmd("apt-get")) {
    stop("This Linux function is intended for Debian/Ubuntu (apt-get).", call. = FALSE)
  }

  apt_pkgs <- c("poppler-utils", "libpoppler-cpp-dev")

  if (use_sudo) {
    run_cmd("sudo", c("apt-get", "update"), "apt-get update failed.")
    run_cmd("sudo", c("apt-get", "install", "-y", apt_pkgs),
            "Installation APT of Poppler/pdftoppm failed.")
  } else {
    run_cmd("apt-get", c("update"), "apt-get update failed.")
    run_cmd("apt-get", c("install", "-y", apt_pkgs),
            "APT installation of Poppler/pdftoppm failed")
  }

  include_candidates <- c("/usr/include", "/usr/local/include")
  lib_candidates <- c("/usr/lib/x86_64-linux-gnu", "/usr/lib", "/usr/local/lib")
  bin_candidates <- c("/usr/bin", "/usr/local/bin")

  include_dir <- include_candidates[
    file.exists(file.path(include_candidates, "poppler", "cpp", "poppler-document.h")) |
      file.exists(file.path(include_candidates, "poppler", "poppler-config.h"))
  ][1]

  lib_dir <- lib_candidates[
    file.exists(file.path(lib_candidates, "libpoppler.so")) |
      file.exists(file.path(lib_candidates, "libpoppler-cpp.so")) |
      file.exists(file.path(lib_candidates, "libpoppler.a")) |
      file.exists(file.path(lib_candidates, "libpoppler-cpp.a"))
  ][1]

  bin_dir <- bin_candidates[
    file.exists(file.path(bin_candidates, "pdftoppm"))
  ][1]

  if (is.na(include_dir)) {
    warning("Poppler headers not found automatically")
    include_dir <- "/usr/include"
  }

  if (is.na(lib_dir)) {
    warning("Poppler libraries not found automatically")
    lib_dir <- "/usr/lib"
  }

  if (is.na(bin_dir)) {
    warning("pdftoppm executable not found automatically")
    bin_dir <- "/usr/bin"
  }

  write_pdftoppm_env(
    include_dir = include_dir,
    lib_dir = lib_dir,
    bin_dir = bin_dir,
    scope = if (persist) "renviron" else "session"
  )

  invisible(list(include = include_dir, lib = lib_dir, bin = bin_dir))
}

install_pdftoppm_macos <- function(persist = TRUE) {
  if (!check_cmd("brew")) {
    stop("Homebrew is not available", call. = FALSE)
  }

  run_cmd("brew", c("install", "poppler"),
          "Homebrew installation of Poppler/pdftoppm failed")

  brew_prefix <- trimws(system2("brew", "--prefix", stdout = TRUE))
  poppler_prefix <- trimws(system2("brew", c("--prefix", "poppler"), stdout = TRUE))

  include_candidates <- c(
    file.path(poppler_prefix, "include"),
    file.path(brew_prefix, "include")
  )

  lib_candidates <- c(
    file.path(poppler_prefix, "lib"),
    file.path(brew_prefix, "lib")
  )

  bin_candidates <- c(
    file.path(poppler_prefix, "bin"),
    file.path(brew_prefix, "bin")
  )

  include_dir <- include_candidates[
    file.exists(file.path(include_candidates, "poppler", "cpp", "poppler-document.h")) |
      file.exists(file.path(include_candidates, "poppler", "poppler-config.h"))
  ][1]

  lib_dir <- lib_candidates[
    file.exists(file.path(lib_candidates, "libpoppler.dylib")) |
      file.exists(file.path(lib_candidates, "libpoppler-cpp.dylib")) |
      file.exists(file.path(lib_candidates, "libpoppler.a"))
  ][1]

  bin_dir <- bin_candidates[
    file.exists(file.path(bin_candidates, "pdftoppm"))
  ][1]

  if (is.na(include_dir)) {
    warning("Poppler headers not found automatically")
    include_dir <- file.path(poppler_prefix, "include")
  }

  if (is.na(lib_dir)) {
    warning("Poppler libraries not found automatically")
    lib_dir <- file.path(poppler_prefix, "lib")
  }

  if (is.na(bin_dir)) {
    warning("pdftoppm executable not found automatically")
    bin_dir <- file.path(poppler_prefix, "bin")
  }

  write_pdftoppm_env(
    include_dir = include_dir,
    lib_dir = lib_dir,
    bin_dir = bin_dir,
    scope = if (persist) "renviron" else "session"
  )

  invisible(list(prefix = poppler_prefix, include = include_dir, lib = lib_dir, bin = bin_dir))
}

install_pdftoppm_windows_msys2 <- function(rtools_root = "C:/rtools45",
                                           repo = "ucrt64",
                                           persist = TRUE) {
  bash <- file.path(rtools_root, "usr", "bin", "bash.exe")
  if (!file.exists(bash)) {
    stop("bash.exe not found in Rtools/MSYS2 ", bash, call. = FALSE)
  }

  pkgs <- switch(
    repo,
    ucrt64  = c("mingw-w64-ucrt-x86_64-poppler"),
    clang64 = c("mingw-w64-clang-x86_64-poppler"),
    mingw64 = c("mingw-w64-x86_64-poppler"),
    stop("repo must be 'ucrt64', 'clang64', or 'mingw64'")
  )

  msys_cmd <- sprintf("pacman -S --needed --noconfirm %s", paste(pkgs, collapse = " "))
  run_cmd(bash, c("-lc", shQuote(msys_cmd)),
          "MSYS2 installation of Poppler/pdftoppm failed.")

  prefix <- switch(
    repo,
    ucrt64  = file.path(rtools_root, "ucrt64"),
    clang64 = file.path(rtools_root, "clang64"),
    mingw64 = file.path(rtools_root, "mingw64")
  )

  include_dir <- file.path(prefix, "include")
  lib_dir <- file.path(prefix, "lib")
  bin_dir <- file.path(prefix, "bin")

  hdr_found <- any(file.exists(c(
    file.path(include_dir, "poppler", "cpp", "poppler-document.h"),
    file.path(include_dir, "poppler", "poppler-config.h")
  )))
  if (!hdr_found) {
    warning("Poppler headers not found at the expected location : ", include_dir)
  }

  libs_found <- any(file.exists(file.path(lib_dir, c(
    "libpoppler.dll.a",
    "libpoppler-cpp.dll.a",
    "libpoppler.a",
    "poppler.lib"
  ))))
  if (!libs_found) {
    warning("Poppler libraries not found at the expected location : ", lib_dir)
  }

  exe <- file.path(bin_dir, "pdftoppm.exe")
  if (!file.exists(exe)) {
    warning("pdftoppm executable not found at the expected location : ", exe)
  }

  write_pdftoppm_env(
    include_dir = include_dir,
    lib_dir = lib_dir,
    bin_dir = bin_dir,
    scope = if (persist) "renviron" else "session"
  )

  invisible(list(
    prefix = prefix,
    include = include_dir,
    lib = lib_dir,
    bin = bin_dir,
    packages = pkgs
  ))
}

install_pdftoppm <- function(...) {
  sys <- Sys.info()[["sysname"]]

  if (identical(sys, "Linux")) {
    return(install_pdftoppm_linux(...))
  }
  if (identical(sys, "Darwin")) {
    return(install_pdftoppm_macos(...))
  }
  if (identical(sys, "Windows")) {
    return(install_pdftoppm_windows_msys2(...))
  }

  stop("OS not supported: ", sys, call. = FALSE)
}

# Exemple Windows / Rtools45 / UCRT64
install_pdftoppm(rtools_root = "C:/rtools45", repo = "ucrt64")

Example

library(rtesseract)

#### OCR english
eng <- tess_open(
  language = "fra",
  datapath = "C:/rtools45/ucrt64/share/tessdata/"
)

setwd("D:/Dropbox/test_Tesseract/")

#### Infos
tess_info(eng)

#### Image file
txt <- tess_ocr_file(eng, "lnCHO.png")
cat(txt)

#### Precise zone ####
rect <- c(1, 20, 30, 15)
txt <- tess_ocr_file(eng, "lnCHO.png", rect)
cat(txt)

#### TSV format ####
tsv <- tess_tsv_file(eng, "lnCHO.png")
cat(substr(tsv, 1, 500))

#### Structured HTML ####
hocr <- tess_hocr_file(eng, "lnCHO.png")
cat(substr(hocr, 1, 500))

#### Boxes ####
words <- tess_boxes_file(eng, "lnCHO.png", level = "word")
print(words)

#### Lines ####
words <- tess_boxes_file(eng, "lnCHO.png", level = "word")
print(words)

#### OCR multiple zones ####
rects <- matrix(
  c(10, 20, 50, 50,
    10, 80, 50, 60),
  ncol = 4,
  byrow = TRUE
)

res <- tess_ocr_rects_file(eng, "lnCHO.png", rects)
print(res)


#### Confidence
conf <- tess_confidence_file(eng, "lnCHO.png")
print(conf)


#### Mode
tess_set_page_seg_mode(eng, 6)
tess_set_variable(eng, "tessedit_char_whitelist", "0123456789")


#### Image to PDF searchable
tess_searchable_pdf_from_images(
  eng,
  image_paths = c("lnCHO.png"),
  output_pdf = "output.pdf",
  dpi = 300
)


setwd("D:/test_Tesseract/")

eng <- tess_open(
  language = "eng",
  datapath = "C:/rtools45/ucrt64/share/tessdata/"
)

#### Scanned PDF to PDF searchable
tess_searchable_pdf_from_scanned_pdf(
  eng,
  input_pdf = "D:\\test_Tesseract\\scan.pdf",
  output_pdf = "D:\\test_Tesseract\\scan_ocr.pdf",
  dpi = 300
)

📄 License

MIT License

👤 Author

Emmanuel Hamel

🤝 Contributing

Pull requests are welcome! Feel free to open issues for bugs or feature requests.

Name		Name	Last commit message	Last commit date
Latest commit History 5 Commits
R		R
src		src
.gitattributes		.gitattributes
DESCRIPTION		DESCRIPTION
NAMESPACE		NAMESPACE
readMe.md		readMe.md
rtesseract.Rproj		rtesseract.Rproj

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Repository files navigation

rtesseract

✨ Features

📦 Installation

1. Install system dependencies

📌 Dependencies

Installation Tesseract

Download tesseract training data

Installation Poppler

Example

📄 License

👤 Author

🤝 Contributing

About

Uh oh!

Releases

Packages

Uh oh!

Contributors

Uh oh!

Languages

Folders and files

Latest commit

History

Repository files navigation

rtesseract

✨ Features

📦 Installation

1. Install system dependencies

📌 Dependencies

Installation Tesseract

Download tesseract training data

Installation Poppler

Example

📄 License

👤 Author

🤝 Contributing

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Packages