add documenation for image read and write

docs/Project.toml

@@ -1,3 +1,4 @@
 [deps]
 Documenter = "e30172f5-a6a5-5a46-863b-614d45cd2de4"
-OpenCV = "f878e3a2-a245-4720-8660-60795d644f2a"
+FileIO = "5789e2e9-d7fb-5bc7-8068-2c6fae9b9549"
+OpenCV = "f878e3a2-a245-4720-8660-60795d644f2a"

docs/make.jl

@@ -7,6 +7,10 @@ DocMeta.setdocmeta!(OpenCV, :DocTestSetup, :(using OpenCV); recursive=true)
 makedocs(;
     modules=[OpenCV],
     sitename="OpenCV.jl",
+    pages = [
+        "OpenCV.jl" => "index.md"
+        "Getting started with Images" => "Getting started with Images.md"
+    ]
 )
 
 deploydocs(;

docs/src/Getting started with Images.md

@@ -0,0 +1,106 @@
+# Reading and Writing of Images
+
+OpenCV provides a wide range of functions for image processing, including reading and writing images.
+
+OpenCV.jl provides 2 ways to read and write images
+1. OpenCV native api
+2. FileIO.jl api
+
+OpenCV provides `imread` and `imwrite`, while FileIO privides `load` and `save` for similar functionalities
+
+
+!!! note
+
+    FileIO api not suppport all image formats currently but it support major formats.
+    If FileIO api not works then use OpenCV native api
+
+## Reading Images
+
+```julia
+using OpenCV
+
+img_path = "/path/to/image"
+img = OpenCV.imread(img_path)
+```
+with FileIO
+```julia
+using OpenCV, FileIO
+
+img_path = "/path/to/image"
+img = load(img_path)
+```
+
+in both cases `img` has same data. Both methods accept same reading flags[^imreadflags].
+
+```julia
+using OpenCV
+
+img_path = "/path/to/image"
+flag = OpenCV.IMREAD_UNCHANGED 
+img = OpenCV.imread(img_path, flag)
+```
+with FileIO
+```julia
+using OpenCV, FileIO
+
+img_path = "/path/to/image"
+flag = OpenCV.IMREAD_UNCHANGED 
+img = load(img_path, flag)
+```
+
+## Writing Images
+```julia
+using OpenCV
+
+img = rand(UInt8, 1000, 1000) |> OpenCV.Mat
+img_path = "/path/to/image"
+OpenCV.imwrite(img_path, img)
+```
+with FileIO
+```julia
+using OpenCV, FileIO
+
+img = rand(UInt8, 1000, 1000) |> OpenCV.Mat
+img_path = "/path/to/image"
+save(img_path, img)
+```
+
+Both methods accept same writing flags[^imwriteflags].
+
+```julia
+using OpenCV
+
+img = rand(UInt8, 1000, 1000) |> OpenCV.Mat
+img_path = "/path/to/image"
+flag = Int32[OpenCV.IMWRITE_JPEG_QUALITY, 100]
+OpenCV.imwrite(img_path, img, flag)
+```
+with FileIO
+```julia
+using OpenCV, FileIO
+
+img = rand(UInt8, 1000, 1000) |> OpenCV.Mat
+img_path = "/path/to/image"
+flag = Int32[OpenCV.IMWRITE_JPEG_QUALITY, 100]
+save(img_path, img, flag)
+```
+
+## Displaying Images
+
+When working with images, it's obviously helpful to be able to look at them. If you use Julia through Pluto, VSCode, or IJulia, images should display automatically.
+
+Preview from Pluto.jl
+
+![pluto preview](assets/pluto.png)
+
+`OpenCV.imshow` method display image on Qt window.
+
+![qt preview](assets/qt.png)
+
+`Pluto.jl` To-do
+
+`Makie.jl` To-do
+
+## Notes
+[^imreadflags]: https://docs.opencv.org/4.x/d8/d6a/group__imgcodecs__flags.html#ga61d9b0126a3e57d9277ac48327799c80
+[^imwriteflags]: https://docs.opencv.org/4.x/d8/d6a/group__imgcodecs__flags.html#ga292d81be8d76901bff7988d18d2b42ac

src/fileio.jl

@@ -22,7 +22,13 @@ function load(f::File{T}, flags::Int) where {T<:_IMAGE_DATA_FORMATS}
     return data
 end
 
-function load(s::Stream{T}, flags::Int=IMREAD_UNCHANGED) where {T<:_IMAGE_DATA_FORMATS}
+function load(s::Stream{T}) where {T<:_IMAGE_DATA_FORMATS}
+    data = read(stream(s))
+    img = imdecode(reshape(data, 1, 1, :))
+    return img
+end
+
+function load(s::Stream{T}, flags::Int) where {T<:_IMAGE_DATA_FORMATS}
     data = read(stream(s))
     img = imdecode(reshape(data, 1, 1, :), flags)
     return img