Skip to content

io_image read_image bugfix & add feature read_image_16 (bit) #681

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 3 commits into
base: master
Choose a base branch
from
+33 −5
Open

io_image read_image bugfix & add feature read_image_16 (bit) #681

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
56bbdef
Fix: bugfix in reshape
voidpunk Apr 8, 2025
fe2a20d
Refactor: default enum & docs
voidpunk Apr 8, 2025
9722f42
Feat: add read_image proc for 16-bit
voidpunk Apr 8, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
38 changes: 33 additions & 5 deletions src/arraymancer/io/io_image.nim
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,7 +18,7 @@ func chw_to_hwc[T](img: Tensor[T]): Tensor[T] {.inline.}=
img.permute(1, 2, 0)

proc read_image*(filepath: string): Tensor[uint8] =
## Read an image file and loads it into a Tensor[uint8] of shape
## Read an 8-bit image file and loads it into a Tensor[uint8] of shape
## Channel x Height x Width. Channel is 1 for greyscale, 3 for RGB.
##
## Supports JPEG, PNG, TGA, BMP, PSD, GIF, HDR, PIC, PNM
Expand All @@ -31,13 +31,13 @@ proc read_image*(filepath: string): Tensor[uint8] =
## x.float32 / 255.0

var width, height, channels: int
let desired_channels = 0 # Channel autodetection
let desired_channels = Default # Channel autodetection

let raw_pixels = load(filepath, width, height, channels, desired_channels)
result = raw_pixels.toTensor.reshape(height, width, channels).hwc_to_chw

proc read_image*(buffer: seq[byte]): Tensor[uint8] =
## Read an image from a buffer and loads it into a Tensor[uint8] of shape
## Read an 8-bit image from a buffer and loads it into a Tensor[uint8] of shape
## Channel x Height x Width. Channel is 1 for greyscale, 3 for RGB.
##
## Supports JPEG, PNG, TGA, BMP, PSD, GIF, HDR, PIC, PNM
Expand All @@ -48,11 +48,39 @@ proc read_image*(buffer: seq[byte]): Tensor[uint8] =
# but nim-stb_image only accept seq[bytes] (and convert it to pointer + length internally)

var width, height, channels: int
let desired_channels = 0 # Channel autodetection
let desired_channels = Default # Channel autodetection

let raw_pixels = loadFromMemory(buffer, width, height, channels, desired_channels)
result = raw_pixels.toTensor.reshape(width, height, channels).hwc_to_chw
result = raw_pixels.toTensor.reshape(height, width, channels).hwc_to_chw

proc read_image_16*(filepath: string): Tensor[uint16] =
## Read a 16-bit image file and loads it into a Tensor[uint16] of shape
## Channel x Height x Width. Channel is 1 for greyscale, 3 for RGB.
##
## Supports JPEG, PNG, TGA, BMP, PSD, GIF, HDR, PIC, PNM
## See stb_image https://github.com/nothings/stb/blob/master/stb_image.h
var width, height, channels: int
let desired_channels = Default # Channel autodetection

let raw_pixels = load16(filepath, width, height, channels, desired_channels)
result = raw_pixels.toTensor.reshape(height, width, channels).hwc_to_chw

proc read_image_16*(buffer: seq[byte]): Tensor[uint16] =
## Read a 16-bit image from a buffer and loads it into a Tensor[uint16] of shape
## Channel x Height x Width. Channel is 1 for greyscale, 3 for RGB.
##
## Supports JPEG, PNG, TGA, BMP, PSD, GIF, HDR, PIC, PNM
## See stb_image https://github.com/nothings/stb/blob/master/stb_image.h
##

# TODO: ideally this should also accept pointer + length
# but nim-stb_image only accept seq[bytes] (and convert it to pointer + length internally)

var width, height, channels: int
let desired_channels = Default # Channel autodetection

let raw_pixels = load16FromMemory(buffer, width, height, channels, desired_channels)
result = raw_pixels.toTensor.reshape(height, width, channels).hwc_to_chw

template gen_write_image(proc_name: untyped): untyped {.dirty.}=

Expand Down