Cherry-picked from develop branch commit 4dc3836449
ignoring changes in all `CMakeLists.txt` files but this one
`test/extension/toolbox/CMakeLists.txt` which somehow missed the changes.
Possibly due to my mistake during previous cherry-picks to compose
the emergency fix#394 for docs in `master` branch, that also
brought the restructuring of the tests directories.
Missing changes in `test/extension/toolbox/CMakeLists.txt` file
are currently causing failures of AzP builds using CMake
for `master` Boost 1.72 merge e8a04f403e
This is attempt to fix it in the `master`.
* Implement Sobel and Scharr operators
This commit adds Sobel and Scharr
operators with support for 0th and 1st
degrees with other degrees planned for
later
* Migrate and fix Harris example
Generate Harris entries now uses
signed image view.
The Harris corner detector example
now uses the Scharr filter generator
and convolve_2d to reduce amount
of code needed.
* Fix and migrate Hessian example
The Hessian example now uses signed
image views and uses newly added kernel
generators to compute gradients
* Fix Harris and Hessian tests
The tests broke due to migration to
signed views in algorithms, but tests
were not adjusted
* Fix Jamfile for example/sobel_scharr.cpp
* Cosmetic changes
* Commented out fail tests
* Fixed pixel16 used in image16s
In Harris and Hessian tests, unsigned
pixel values was used to construct
signed image, which was causing
appveyor to error out.
* Reenable failing targets
* Unify kernel generator interface
This commit makes all kernel
generator functions to return kernel_2d
and adapts dependant threshold
function to use the new interface
* Migrate Hessian and Harris tests
Migrate Hessian and Harris tests to new
interface for kernel generators
* Migrate Harris and Hessian examples
Harris and Hessian examples now use
new interface for kernel generation
* Migrate simple_kernels tests
simple_kernels are now using kernel_2d
interface
* Add missing return
Normalized mean generation had missing
return at the end of the function
* Adapt code to namespace move
This commit reacts to kernel_2d,
convolve_2d being moved to
namespace detail
There is no need for format-specific subfolders.
There is no code in place that checks if all those folders exist and
creates them.
This change simplifies the structure to simplify running of I/O tests.
Rename images/jpg/ to images/jpeg/ in tests
The `jpeg` is name of the format used in the I/O extension and
the folders inside `images/` are named after those formats.
* Hide convolve_1d and convolve_2d in namespace detail
* Hide kernel_2d in namespace detail.
* Tidy up naming of horizontal/vertical to y/x for kernel center.
Following @lpranam comments on planned changes to convolve_2d,
which will not be ready for Boost 1.72,
@mloskot proposed to hide the function as implementation detail,
https://lists.boost.org/boost-gil/2019/10/0316.php,
until it is ready and @stefanseefeld agreed on the proposal,
https://lists.boost.org/boost-gil/2019/10/0320.php.
Split header tests accordingly i.e. test core headers as part of
core tests, numeric extension headers as part of numeric tests, etc.
It extends the convention of sub-directories already established in
`include/boost/gil` directory. It is sensible to follow it in other
areas of the source tree (i.e. `test/`, `doc/` and `benchmark/`).
Another important reason to move the tests is to enable removal of
the top-level `Jamfile` with all its definitions of test-specific
requirements.
The top-level `Jamfile` is not advised, especially if it specifies
build requirements like C++ language version.
Those affect non-tests builds e.g. documentation, causing failures
during generation of HTML documentation (leads to missing docs).
(cherry picked from develop branch commit 4ed7701b47)
* Hessian plain determinant
This commit partially implements
Hessian corner detector, but only
uses determinant as Hessian response
* Implement full Hessian corner detector
This commit complements last one
by summing in a window and applying
det - k * trace * trace formula to
final Hessian response
* Add docs and make code align with docs
This commit adds docs to new functions,
and makes function and variable names
align with docs to not confuse readers
* Use determinant as response function
A-KAZE uses only determinant in it's
response, and since for now Hessian
is only a mean to advance A-KAZE
implementation, response function is
adjusted to use only determinant
* Create simple test for Hessian detector
This commit adds an s at the end of
function name to make it uniform with
another detector, and adds a simple
test for sanity check
* Improve documentations for d params
dx and similar naming seems to be
confusing, improved documentation
to explain the naming convention
* Address minor comments about style
* Address type based issues
Mostly changes to constness and
integral types, with small cosmetic
changes mixed in
* Fix typo and address review comment
Fixes typo in call for Hessian and
addresses a review comment about
replacing multiple exact indexing
calls to one with a reference
* Reorder includes in Hessian example
* Address review comments
Add literature reference to luminosity
computation, and perform some cosmetic
changes.
* Restore previous state of Harris
This is a fresh start of branch
Harris with manual copying and
placement of code from previous
versions
* Use supplied weights in Harris
Weights passed as arguments were not
used prior to this commit
* Address review comments
Mostly moving brackets and adding
const, but also address MSVC's
max macro
* Make namespace qualification consistent
In file numeric.hpp, there is
full qualification, but harris.hpp
didn't have full qualification,
thus full qualification is added to
harris.hpp
* Add copyright and license notice
The tests cover issue #117 and pull request #118, extending the coverage
for all color types that may have tRNS chunks:
- color type 0 (grayscale), the tRNS chunk contains a single gray level
- color type 2 (truecolor), the tRNS chunk contains a single RGB color
- color type 3 (indexed color), the tRNS chunk contains a series of
one-byte alpha values
Add sample PNG files from the "official" test-suite for PNG
created by Willem van Schaik.
Add <define>BOOST_GIL_IO_TEST_ALLOW_READING_IMAGES to Jamfile for PNG tests.
NOTE: Deliberately not defined BOOST_GIL_IO_TEST_ALLOW_WRITING_IMAGES
as it seems to require some extra configuration, currently it causes
some tests to fail. So, Jamfile likely needs more updates and clean-up.
Closes#120
Add static assertions to verify channel types are convertible to
expected result type.
Add new and update existing Doxygen comments describing interfaces in
`extension/numeric/channel_numeric_operations.hpp`.
* Implement generation of simple kernels
This commits implements mean and
Gaussian kernel generators,
roughly documents them, and has
tests. But tests for Gaussian have
accuracy problem
* Use difference for testing Gaussian
This commit applies percentage based
difference check on generation of
Gaussian kernels, as exact values are
dependant on platform. The error margin
is set to 5% in this commit
* Remove debugging code
In simple_kernels.cpp, there was
some debugging code, mostly streams
related stuff and commented out code,
which is now removed.
* Address review comments
This commit adds inline to kernel
generator functions, uses
std::ptrdiff_t instead of long int,
resolves some conversion warnings
and does small cosmetic changes
* Remove useless iostream include
There was stray include from debugging
code, now removed
* Use \code and \endcode for formula
This commit applies formatting to
formula of normalized mean
* Change are_equal to is_equal
Change a function name in
simple_kernels.cpp test
* Remove redundant function
In simple_kernels.cpp test there was
a function that was not used at all,
now removed
* Cosmetic changes
Mostly opening brackets moved
Align with `subsampled_view` overloads accepting `point_t` and `coord_t`
Both types are based on `std::ptrdiff_t`. Fixes compilation warnings.
Tidy up `subimage_view` and `subsampled_view` with trailing return.
Add tests for `subimage_view` in core and dynamic_image extension.
(First tests for the dynamic_image, hurray! :))
Add explicit cast to double in division operator for point<T> to
fix compilation warnings.
First stab at convolution tests includes cases with the identity kernel.
Move `pixel`-s streaming facility used by Boost.Test (required by
`BOOST_TEST` macro) to the common header for re-use in other tests.
Add basic run-time test of `planar_rgba_view` function.
Add missing includes of `pixel.hpp` and `planar_pixel_reference.hpp`
also required for successful compilation.
Fixes#331
Since use of `gil_function_requires` check is optional, it can be
disabled by not defining `BOOST_GIL_USE_CONCEPT_CHECK`, we need
a mandatory form of the check.
Add compile-time test verifying the static assertion.
Closes#323
Use of `std::extent` was introduced in PR #200 but it turns out as
not applicable for `std::array` or derived types (e.g. `kernel_1d_fixed`).
This led to obtaining invalid size of `kernel_1d_fixed` and erroneous
results of the rows and columns convolution.
This change replaces `std::extent` with new public constant member
`kernel_1d_fixed::static_size`.
Since `kernel_1d_fixed` is derived from `std::array`, the Alternative
could be to use `std::tuple_size` specialization for `std::array`.
Add static assertion to require that kernel size must be odd to
ensure validity at the center.
The test also illustrates and verifies that the `static_transform` is
constrained by the size of destination color base. That is, it is
applicable if the source color base has number of elements equal-to or
greater-than the destination color base.
* Simple implementation of lanczos scaling
* Simple implementation of lanczos scaling
* Refactor lanczos into separate header
This commit moves the algorithm
into its own header. It also provides
templated interface now, utilizing
a little bit of decltype and declval
to resolve types for lambdas and stuff
* Zero pixel at start of lanczos_at
This commit fixes a possible bug and
fixes some style incosistencies
* Improve documentation for Lanczos
This commit adds documentation which
describes when to use the algorithm,
a brief description of how it is
supposed to work, and some caution
on the quality of the output
* Address style issues and fix warnings
This commit fixes a style issue realted
to namespace declaration and
fixes a warning in added zeroing of
a pixel at start of lanczos
* text-realign function arguments
* fix formatting issues
* Implement handful of sanity tests
This commit introduces a couple of
sanity tests such as black image
scaling to black image, and lanczos
response being 0 at x = 0
* bracket on newline for for loops
* add lanczos scaling to tests
* more precision in lanczos calculation
This commit migrates integral values
to double precision and uses PI
provided by boost.Math. These changes
solve downscaled image being darker
* Simple implementation of lanczos scaling
* Refactor lanczos into separate header
This commit moves the algorithm
into its own header. It also provides
templated interface now, utilizing
a little bit of decltype and declval
to resolve types for lambdas and stuff
* Zero pixel at start of lanczos_at
This commit fixes a possible bug and
fixes some style incosistencies
* Improve documentation for Lanczos
This commit adds documentation which
describes when to use the algorithm,
a brief description of how it is
supposed to work, and some caution
on the quality of the output
* Address style issues and fix warnings
This commit fixes a style issue realted
to namespace declaration and
fixes a warning in added zeroing of
a pixel at start of lanczos
* text-realign function arguments
* fix formatting issues
* Implement handful of sanity tests
This commit introduces a couple of
sanity tests such as black image
scaling to black image, and lanczos
response being 0 at x = 0
* bracket on newline for for loops
* add lanczos scaling to tests
* more precision in lanczos calculation
This commit migrates integral values
to double precision and uses PI
provided by boost.Math. These changes
solve downscaled image being darker
* Add Jamfile for ip test directory
Add Jamfile for image_processing
test directory and build-project
from outer test directory
* Add IP test directory to ci build
This commit adds a line at the end
of .ci/build-and-test.sh to include
image processing tests in CI builds
* Remove redundant lines from Jamfile
Simplify Jamfile at
test/image_processing
* Rewrite range condition
x > -a && x < a exchange with
-a < x && x < a in lanczos
* Add newline at the end of files
* Add math and lexical_cast to get-boost
As math and lexical_cast are used
in lanczos scaling, both were added
to get-boost
* Revert "Add newline at the end of files"
Since Boost.Math.Constants is
overengineered for use case by
requiring Boost.lexical_cast or
a global define, it is removed
This reverts commit 0743ab072ff3455421853697f7f46aee7d22382c.
* Define pi and use in lanczos
Value of pi is moved out into detail
and used by lanczos
* Move image_processing to test/core
* formatting fixes
* Adjust CMakeLists for moved IP tests
This commit removes image_processing as
subdirectory from test/ and adds to
test/core's CMakeLists
* Remove unused from get-boost
Since lanczos scaling no longer uses
boost.math.constants, lexical_cast
has been removed as well
* Downgrade math to transitive dep
This is a stray change left from
my incorrect resetting of the HEAD
on lanczos branch
* Remove unnecessary includes
io includes are not used in lanczos
scaling test, and they break build
* Fix ambiguous overload issue for min
Since width() and height() now return
std::ptrdiff_t, call to std::min needs
a cast. There was also shadowing of
pi declared in detail/math.hpp,
which is also fixed
* Apply mloskot's patch
The patch provided by Mateusz changes
all usages of long int into ptrdiff_t,
which is returned by width() and
height() functions of image_view
* Apply .editorconfig rules
This commit is a simple reformat of
affected files
* Use aliases x_coord_t and y_coord_t
The change converges integer handling
in arguments, using view' type aliases.
View arguments have to come first to
avoid non-deduced context problem.
Also replaced long int with ptrdiff_t
in lanczos, in numeric.hpp file
* Apply alias usage for rest of the code
Some places with ptrdiff_t were
left out from previous commit, so they
are changed in this one. Plus a fix for
max call being ambiguous.
* Replace all literals with casted vars
This commit takes extreme stance of
never using an integer literal due to
problems in ambiguity of deduction of
min and max functions
* Fix unenclosed foreach
During rebase of gsoc2019 onto develop,
I didn't enclose first foreach with
matching endforeach, hence tests failed
Avoid Clang flooding with non-disableable warnings like:
T does not declare any constructor to initialize its non-modifiable members
when compiling with concepts check enabled.
Bug 41759: `warn_no_constructor_for_refconst` can not be disabled (Boost.ConceptCheck)
https://bugs.llvm.org/show_bug.cgi?id=41759
Split header tests accordingly i.e. test core headers as part of
core tests, numeric extension headers as part of numeric tests, etc.
It extends the convention of sub-directories already established in
`include/boost/gil` directory. It is sensible to follow it in other
areas of the source tree (i.e. `test/`, `doc/` and `benchmark/`).
Another important reason to move the tests is to enable removal of
the top-level `Jamfile` with all its definitions of test-specific
requirements.
The top-level `Jamfile` is not advised, especially if it specifies
build requirements like C++ language version.
Those affect non-tests builds e.g. documentation, causing failures
during generation of HTML documentation (leads to missing docs).
