A Local Adaptive Thresholding framework for image binarization written in C++, with JS and Python bindings. Implementing: Otsu, Bernsen, Niblack, Sauvola, Wolf, Gatos, NICK, Su, T.R. Singh, WAN, ISauvola, Bataineh, Chan and Shafait.
Home Page: https://brandonmpetty.github.io/Doxa/WebAssembly
License: Creative Commons Zero v1.0 Universal
Sauvola-alike methods is implemented using integral images in Doxa. The main drawback of the approach is that the two integral images are quite memory consuming. In fact, a recent paper https://arxiv.org/abs/1905.13038 pointed out a simple way to greatly reduce memory footprint, while keeping the time complexity linear and independent of window size.
Should be "bataineh" and not "trsingh"
Hello,
I'm testing Doxa for converting color or grayscale book scans to bilevel images. I would say that 95% of time, it works, but 5% of time, Doxa produces malformed output with “zebra stripes” for lack of a better term. Here is how this output looks:
The code I’m using is very basic:
Doxa::Image doxaGsImage = Doxa::PNM::Read(input);
Doxa::ISauvola::UpdateToBinary(doxaGsImage);
Doxa::PNM::Write(doxaGsImage, output);
Thank you in advance for looking into this issue.
Please find attached, as ZIP file, a sample showing the problem. The ZIP file contains:
Doxa::PNM::ReadHI,
I try to run BinaryImageConverter.cpp on Clion but it have trouble:
Binarization/Image.hpp:127:47: error: expected ‘,’ or ‘...’ before ‘default’
inline Pixel32 Pixel(int x, int y, Pixel32 default) const { return (x < 0 || x >= width || y < 0 || y >= height) ? default: Pixel(x, y); }
Binarization/Image.hpp:127:119: error: expected ‘:’ before ‘default’
Binarization/Image.hpp:127:119: error: expected primary-expression before ‘default’
Binarization/Image.hpp:127:119: error: expected ‘;’ before ‘default’
Binarization/Image.hpp:127:119: error: case label not within a switch statement
I am using c++ 11: set(CMAKE_CXX_STANDARD 11)
Thanks
I test Doxa::Wan::UpdateToBinary(doxaGsImage, parameters); release
Wan algorithm is slowly if window size more than 17?
time , size k=0.1
0.136 , 6
0.133 , 7
0.178 , 8
0.171 , 9
0.227 , 10
0.231 , 11
0.286 , 12
0.287 , 13
0.343 , 14
0.339 , 15
0.402 , 16
0.601 , 17
0.601 , 18
0.620 , 19
0.620 , 20
0.621 , 21
0.617 , 22
0.635 , 23
0.638 , 24
0.633 , 25
0.641 , 26
0.641 , 27
0.637 , 28
0.653 , 29
0.653 , 30
0.657 , 31
0.647 , 32
0.661 , 33
0.672 , 34
0.680 , 35
0.666 , 36
Just wanted to let you know that I really appreciate your work. I like the new additions, and the general code structure. Performance is now really good and ISauvola give me some nice results (as a low hanging fruit I tried IWolf and IWan they work great too).
I am currently testing Doxa with via a javascript frontend (Doxa is compiled into Webassembly with emscripten) and everything is running great. If you are interested I will be happy to show you more of that or setup a small demo.
Unfortunately as I cannot find a way to send you a Private Message I am abusing the Github issues system, I hope that ok.
Hi again, @brandonmpetty!
You helped me get the 'R' version going back in August, and then recently kindly chimed in on an old issue I raised on @paulbauriegel's cython doxa (https://github.com/paulbauriegel/doxa_cython). (Ty!)
I am trying to build doxa with pybind11 per the readme in your repo.
When i get to the cmake .. step (in Windows 10, Visual Studio Code), I get
The term 'cmake' is not recognized as the name of a cmdlet, function, script file, or operable program
I tried adding the cmake extension. But no difference. (Disclaimer: any time I've tried to do anything with c-anything, I've had little success -- typically more complicated than I really know how to do, hours spent, and very little results. I'm more at the level of "pip install" and the like. :) )
We need to document the parameter defaults for each algorithm, and other algorithms.
Create a Wiki that describes how to use the API and its options, along with how each Algorithm works.
Otsu
Bataineh
No parameters
Niblack
Sauvola
ISauvola
TRSingh
Wan
Wolf
name: "window", default: 75
name: "k", default: 0.2
NICK
name: "window", default: 75
name: "k", default: -0.2
Bernsen
name: "window", default: 75
name: "threshold", default: 100
name: "contrast-limit", default: 25
Gatos
name: "glyph", default: 60
Su *
name: "window", default: 0 = Auto-Detect
name: "minN", default: value of "window"
* Su - A "window" value of 0 will cause the algorithm to auto-detect based on stroke width. This is not currently implemented so a default value of 9 is currently being used.
This is for C++ only
PNM::Read *
name: "grayscale", default: Mean
* "grayscale" algorithm options include: Qt, Mean, BT601, BT709, BT2100, Value, Luster, Lightness
Hello there
I've made an R wrapper around this package as I needed it in some preprocessing for handwritten text recognition. Package was published today on CRAN at https://cran.r-project.org/web/packages/image.binarization/index.html, source code has been put at https://github.com/DIGI-VUB/image.binarization
Putting a package on CRAN also means I have to maintain it and fixing possible issues. One of the issues appearing is that CRAN tests on various platforms (windows / debian / solaris / mac) with different compiler settings.
When testing on Mac OS (version 10.13), I'm getting the following compilation issue indicating std::get is only available in Mac OS 10.14. That's unfortunate given that I have no option to update Mac OS on CRAN to something higher. Hence my question, what would be a good way to solve this, such that it works on Mac OS 10.13 which does not have std::get. See below the compilation trace.
Thanks for the help!
* installing *source* package ‘image.binarization’ ...
** package ‘image.binarization’ successfully unpacked and MD5 sums checked
** using staged installation
** libs
clang++ -mmacosx-version-min=10.13 -std=gnu++17 -I"/Library/Frameworks/R.framework/Resources/include" -DNDEBUG -I'/Users/user9yOn5yDN/R/Rcpp/include' -I/usr/local/include -fPIC -Wall -g -O2 -c RcppExports.cpp -o RcppExports.o
clang++ -mmacosx-version-min=10.13 -std=gnu++17 -I"/Library/Frameworks/R.framework/Resources/include" -DNDEBUG -I'/Users/user9yOn5yDN/R/Rcpp/include' -I/usr/local/include -fPIC -Wall -g -O2 -c _rcpp_doxa.cpp -o _rcpp_doxa.o
In file included from _rcpp_doxa.cpp:4:
In file included from ./Doxa/Algorithm.hpp:7:
./Doxa/Parameters.hpp:36:12: error: call to unavailable function 'get': introduced in macOS 10.14
return std::get<Type>(pos->second);
^~~~~~~~~~~~~~
./Doxa/Bernsen.hpp:30:38: note: in instantiation of function template specialization 'Doxa::Parameters::Get<int>' requested here
const int windowSize = parameters.Get("window", 75);
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/variant:1393:22: note: candidate function [with _Tp = int, _Types = <int, double>] has been explicitly made unavailable
constexpr const _Tp& get(const variant<_Types...>& __v) {
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/variant:1401:23: note: candidate function [with _Tp = int, _Types = <int, double>] not viable: no known conversion from 'const std::__1::variant<int, double>' to 'const variant<int, double> &&' for 1st argument
constexpr const _Tp&& get(const variant<_Types...>&& __v) {
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:780:17: note: candidate template ignored: could not match 'pair' against 'variant'
constexpr _T1 & get(pair<_T1, _T2>& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:787:23: note: candidate template ignored: could not match 'pair' against 'variant'
constexpr _T1 const & get(pair<_T1, _T2> const& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:794:18: note: candidate template ignored: could not match 'pair' against 'variant'
constexpr _T1 && get(pair<_T1, _T2>&& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:801:24: note: candidate template ignored: could not match 'pair' against 'variant'
constexpr _T1 const && get(pair<_T1, _T2> const&& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:808:17: note: candidate template ignored: could not match 'pair' against 'variant'
constexpr _T1 & get(pair<_T2, _T1>& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:815:23: note: candidate template ignored: could not match 'pair' against 'variant'
constexpr _T1 const & get(pair<_T2, _T1> const& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:822:18: note: candidate template ignored: could not match 'pair' against 'variant'
constexpr _T1 && get(pair<_T2, _T1>&& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:829:24: note: candidate template ignored: could not match 'pair' against 'variant'
constexpr _T1 const && get(pair<_T2, _T1> const&& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/tuple:1035:16: note: candidate template ignored: could not match 'tuple' against 'variant'
constexpr _T1& get(tuple<_Args...>& __tup) noexcept
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/tuple:1042:22: note: candidate template ignored: could not match 'tuple' against 'variant'
constexpr _T1 const& get(tuple<_Args...> const& __tup) noexcept
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/tuple:1049:17: note: candidate template ignored: could not match 'tuple' against 'variant'
constexpr _T1&& get(tuple<_Args...>&& __tup) noexcept
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/tuple:1056:23: note: candidate template ignored: could not match 'tuple' against 'variant'
constexpr _T1 const&& get(tuple<_Args...> const&& __tup) noexcept
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/variant:1376:16: note: candidate template ignored: deduced type 'variant<...>' of 1st parameter does not match adjusted type 'const variant<...>' of argument [with _Tp = int, _Types = <int, double>]
constexpr _Tp& get(variant<_Types...>& __v) {
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/variant:1384:17: note: candidate template ignored: deduced type 'variant<...>' of 1st parameter does not match adjusted type 'const variant<...>' of argument [with _Tp = int, _Types = <int, double>]
constexpr _Tp&& get(variant<_Types...>&& __v) {
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:746:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(pair<_T1, _T2>& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:754:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(const pair<_T1, _T2>& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:763:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(pair<_T1, _T2>&& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:771:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(const pair<_T1, _T2>&& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/tuple:958:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(tuple<_Tp...>& __t) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/tuple:967:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(const tuple<_Tp...>& __t) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/tuple:976:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(tuple<_Tp...>&& __t) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/tuple:986:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(const tuple<_Tp...>&& __t) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/array:307:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(array<_Tp, _Size>& __a) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/array:316:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(const array<_Tp, _Size>& __a) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/array:327:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(array<_Tp, _Size>&& __a) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/array:336:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(const array<_Tp, _Size>&& __a) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/variant:1336:59: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
constexpr variant_alternative_t<_Ip, variant<_Types...>>& get(
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/variant:1346:60: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
constexpr variant_alternative_t<_Ip, variant<_Types...>>&& get(
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/variant:1356:65: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
constexpr const variant_alternative_t<_Ip, variant<_Types...>>& get(
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/variant:1366:66: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
constexpr const variant_alternative_t<_Ip, variant<_Types...>>&& get(
^
In file included from _rcpp_doxa.cpp:4:
In file included from ./Doxa/Algorithm.hpp:7:
./Doxa/Parameters.hpp:36:12: error: call to unavailable function 'get': introduced in macOS 10.14
return std::get<Type>(pos->second);
^~~~~~~~~~~~~~
./Doxa/Sauvola.hpp:32:32: note: in instantiation of function template specialization 'Doxa::Parameters::Get<double>' requested here
const double k = parameters.Get("k", 0.2);
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/variant:1393:22: note: candidate function [with _Tp = double, _Types = <int, double>] has been explicitly made unavailable
constexpr const _Tp& get(const variant<_Types...>& __v) {
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/variant:1401:23: note: candidate function [with _Tp = double, _Types = <int, double>] not viable: no known conversion from 'const std::__1::variant<int, double>' to 'const variant<int, double> &&' for 1st argument
constexpr const _Tp&& get(const variant<_Types...>&& __v) {
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:780:17: note: candidate template ignored: could not match 'pair' against 'variant'
constexpr _T1 & get(pair<_T1, _T2>& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:787:23: note: candidate template ignored: could not match 'pair' against 'variant'
constexpr _T1 const & get(pair<_T1, _T2> const& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:794:18: note: candidate template ignored: could not match 'pair' against 'variant'
constexpr _T1 && get(pair<_T1, _T2>&& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:801:24: note: candidate template ignored: could not match 'pair' against 'variant'
constexpr _T1 const && get(pair<_T1, _T2> const&& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:808:17: note: candidate template ignored: could not match 'pair' against 'variant'
constexpr _T1 & get(pair<_T2, _T1>& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:815:23: note: candidate template ignored: could not match 'pair' against 'variant'
constexpr _T1 const & get(pair<_T2, _T1> const& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:822:18: note: candidate template ignored: could not match 'pair' against 'variant'
constexpr _T1 && get(pair<_T2, _T1>&& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:829:24: note: candidate template ignored: could not match 'pair' against 'variant'
constexpr _T1 const && get(pair<_T2, _T1> const&& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/tuple:1035:16: note: candidate template ignored: could not match 'tuple' against 'variant'
constexpr _T1& get(tuple<_Args...>& __tup) noexcept
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/tuple:1042:22: note: candidate template ignored: could not match 'tuple' against 'variant'
constexpr _T1 const& get(tuple<_Args...> const& __tup) noexcept
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/tuple:1049:17: note: candidate template ignored: could not match 'tuple' against 'variant'
constexpr _T1&& get(tuple<_Args...>&& __tup) noexcept
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/tuple:1056:23: note: candidate template ignored: could not match 'tuple' against 'variant'
constexpr _T1 const&& get(tuple<_Args...> const&& __tup) noexcept
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/variant:1376:16: note: candidate template ignored: deduced type 'variant<...>' of 1st parameter does not match adjusted type 'const variant<...>' of argument [with _Tp = double, _Types = <int, double>]
constexpr _Tp& get(variant<_Types...>& __v) {
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/variant:1384:17: note: candidate template ignored: deduced type 'variant<...>' of 1st parameter does not match adjusted type 'const variant<...>' of argument [with _Tp = double, _Types = <int, double>]
constexpr _Tp&& get(variant<_Types...>&& __v) {
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:746:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(pair<_T1, _T2>& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:754:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(const pair<_T1, _T2>& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:763:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(pair<_T1, _T2>&& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/utility:771:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(const pair<_T1, _T2>&& __p) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/tuple:958:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(tuple<_Tp...>& __t) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/tuple:967:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(const tuple<_Tp...>& __t) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/tuple:976:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(tuple<_Tp...>&& __t) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/tuple:986:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(const tuple<_Tp...>&& __t) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/array:307:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(array<_Tp, _Size>& __a) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/array:316:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(const array<_Tp, _Size>& __a) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/array:327:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(array<_Tp, _Size>&& __a) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/array:336:1: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
get(const array<_Tp, _Size>&& __a) _NOEXCEPT
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/variant:1336:59: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
constexpr variant_alternative_t<_Ip, variant<_Types...>>& get(
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/variant:1346:60: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
constexpr variant_alternative_t<_Ip, variant<_Types...>>&& get(
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/variant:1356:65: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
constexpr const variant_alternative_t<_Ip, variant<_Types...>>& get(
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/variant:1366:66: note: candidate template ignored: invalid explicitly-specified argument for template parameter '_Ip'
constexpr const variant_alternative_t<_Ip, variant<_Types...>>&& get(
^
2 errors generated.
make: *** [_rcpp_doxa.o] Error 1
ERROR: compilation failed for package ‘image.binarization’
* removing ‘/Users/user9yOn5yDN/R/image.binarization’
Do you have plan to support Nvidia CUDA?
Hello,
The Wan algorithm won't compile, because it makes uses of a morphology function:
Morphology::Dilate(maxImage, Algorithm::grayScaleImageIn, windowSize);
and yet, #include "Morphology.hpp" is missing from Wan.hpp.
Luckily, the fix is easy. The relevant include simply has to be added.
Hope this helps!
Has anyone had success using this on Windows?
If so, would you be willing to give simple instructions to a total beginner about how to get it up and running?
Any guidance would be much appreciated!
(Windows 10)
Hi @brandonmpetty,
Thanks for the awesome package. I have been experimenting with different neural networks architectures for document image binarization (DIBCO), and after trying multiple implementations of DIBCO metrics, yours are much faster, accelerating the training loop (using Hugging Face transformers' trainer API) by a few orders of magnitude.
Below I will leave a few open points or requests, but let me know if you want me to open separate issues:
calculate_performance function.import numpy as np
import numpy.typing as np_typing
import scipy
Bitmap = np_typing.NDArray[np.bool_]
G123_LUT = np.array([
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0,
0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 1, 0, 1, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 1, 0, 1, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
1, 0, 0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 1, 0, 1, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
1, 0, 0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 1, 0, 1, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
1, 1, 0, 0, 1, 0, 0, 0, 1, 1, 0, 0, 1, 0, 0, 0,
1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
1, 1, 0, 0, 1, 0, 0, 0, 1, 1, 0, 0, 1, 0, 0, 0
], dtype=bool)
G123P_LUT = np.array([
0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 1, 1, 1,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 1, 0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 1, 0, 1,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 1, 0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 1, 0, 1,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 1, 0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 1, 0, 1,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
1, 1, 0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
], dtype=bool)
def bwmorph_thin(bitmap: Bitmap, num_iters: int = -1) -> Bitmap:
"""The bwmorph thinning algorithm
For more information on the algorithm, see
https://it.mathworks.com/help/images/ref/bwmorph.html#bui7chk-1
"""
if bitmap.ndim not in [2, 3]:
raise ValueError("The bitmap must be a 2D array or a batched 3D tensor")
if not np.all(np.in1d(bitmap.flat, (0, 1))):
raise ValueError("The bitmap contains values other than 0 and 1")
if num_iters <= 0 and num_iters != -1:
raise ValueError("num_iters must be > 0 or equal to -1")
bitmap = np.array(bitmap).astype(np.uint8)
batched3d = bitmap.ndim == 3
if not batched3d:
bitmap = np.expand_dims(bitmap, 0)
# The neighborhood kernel
kernel = np.array([
[ 8, 4, 2],
[16, 0, 1],
[32, 64, 128]
], dtype=np.uint8)
finished = np.zeros(bitmap.shape[0], dtype=bool)
batch_size = bitmap.shape[0]
num_pixels_before = np.sum(bitmap, axis=(1, 2))
while num_iters != 0:
# The two subiterations
for lut in [G123_LUT, G123P_LUT]:
for idx in range(batch_size): # It is faster than the batched operation
if finished[idx]:
continue
N = scipy.ndimage.correlate(bitmap[idx], kernel, mode="constant")
D = np.take(lut, N)
bitmap[idx][D] = 0
num_pixels = np.sum(bitmap, axis=(1, 2))
finished = num_pixels == num_pixels_before
if np.all(finished):
break
num_pixels_before = num_pixels
num_iters -= 1
if not batched3d:
bitmap = np.squeeze(bitmap, axis=0)
return bitmap.astype(bool)
def pseudo_fmeasure(references: Bitmap, preds: Bitmap, eps: float = 1e-6, **kwargs) -> np_typing.NDArray[np.float_]:
"""The pseudo F-measure metric"""
neg_references = 1 - references
neg_preds = 1 - preds
skeletons = bwmorph_thin(neg_references, **kwargs).astype(np.uint8)
tpositives = neg_preds * neg_references
fpositives = neg_preds * references
num_tpositives = np.sum(tpositives, axis=(1, 2))
num_fpositives = np.sum(fpositives, axis=(1, 2))
precision = num_tpositives / (num_fpositives + num_tpositives + eps)
pseudo_tpositives = neg_preds * skeletons
pseudo_fnegatives = preds * skeletons
num_pseudo_tpositives = np.sum(pseudo_tpositives, axis=(1, 2))
num_pseudo_fnegatives = np.sum(pseudo_fnegatives, axis=(1, 2))
pseudo_recall = num_pseudo_tpositives / (num_pseudo_fnegatives + num_pseudo_tpositives + eps)
pseudo_nume = 2 * (precision * pseudo_recall)
pseudo_deno = precision + pseudo_recall + eps
pseudo_score = pseudo_nume / pseudo_deno
return pseudo_scoreThank you for this great library, it's very well put together and easy to use and understand. I'm doing a personal project, restarting after a very long time and this library will save me time so I can focus on what I'm doing.
When I use a large file, the binary result shows skewed lines thorough out the whole file. After I scaled the file in half it worked ok. The original file was 1800x1200 pixels.
Hello!
I'm trying to specify binarization methods and/or use all at once using demo.py as my starter template.
Here's the latest that I tried:
# Loop through all binarization algorithms
binarize_algos = [BERNSEN, NIBLACK, SAUVOLA, WOLF, NICK, SU, TRSINGH, BATAINEH, ISAUVOLA, WAN, GATOS]
for ba in binarize_algos:
algo = doxapy.Binarization(doxapy.Binarization.Algorithms.ba)
algo.initialize(grayscale_image)
algo.to_binary(binary_image, {"window": 75, "k": 0.2}) # algorithm params are optional
# Display our resulting image
Image.fromarray(binary_image).show()
I also tried with the list's methods in quotes, mixed in with 3 or so variations on this line:
algo = doxapy.Binarization(doxapy.Binarization.Algorithms.ba)
(str(ba), for example, pluses (for both str(ba) and just ba like you see above)
Any idea what I'm doing wrong?
Any plan to support pip install for the Python binding? Thank you
Hi, thank you for this great library.
There is another interesting method called AdOtsu which is an adaptive version inspired by the Otsu method. See paper AdOtsu: An adaptive and parameterless generalization of Otsu's method for document image binarization
Would it be possible to implement this one? I know this is a time consuming task, so feel free to ignore it.
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
Django
The Web framework for perfectionists with deadlines.
Laravel
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
Microsoft
Open source projects and samples from Microsoft.
Google
Google ❤️ Open Source for everyone.
Alibaba
Alibaba Open Source for everyone
D3
Data-Driven Documents codes.
Tencent
China tencent open source team.