https://github.com/starkdmi/swiftorientcrop

Working with oriented images in Swift
https://github.com/starkdmi/swiftorientcrop

cgimage ciimage crop-image image-manipulation image-processing swift vimage

Last synced: about 8 hours ago
JSON representation

Working with oriented images in Swift

• Host: GitHub
• URL: https://github.com/starkdmi/swiftorientcrop
• Owner: starkdmi
• License: mit
• Created: 2023-08-24T16:36:44.000Z (about 1 year ago)
• Default Branch: main
• Last Pushed: 2023-08-25T14:32:41.000Z (about 1 year ago)
• Last Synced: 2023-08-25T19:05:58.975Z (about 1 year ago)
• Topics: cgimage, ciimage, crop-image, image-manipulation, image-processing, swift, vimage
• Language: Swift
• Homepage:
• Size: 35.6 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# SwiftOrientCrop

Project content:

- Oriented image generator

- Load and orient CGImage and CIImage

- Orient CGRect based on image orientation

- Crop CGImage, CIImage and vImage

- Benchmarks

Code located in [Tests](SwiftOrientCropTests/SwiftOrientCropTests.swift).

## Oriented image generator

Those images looks the same when correctly displayed, but under the hood they pretty different.

|  |  |  |

| --- | --- | --- |

|  |  |  |

|  | | |

### Usage

```Swift

try OrientedGenerator.generateFrom(

  source: sourceImageFile,

  destination: destinationDirectory,

  format: .jpeg,

  size: CGSize(width: 1280, height: 1280),

  quality: 0.75

)

```

## Image Crop Benchmarks

One call equals to loading 8 images, crop them and write image data to 

file. Full execution is 10 single calls - 80 images in summary.

| Method | Full Execution Time (seconds) | Average Execution Time per call (seconds) | Average CPU Time (seconds) | Memory Usage (kB) | Memory Peak Physical (MB) | Disk Writes (MB) | CPU Usage (% relative to CGImage) |

|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|

| CGImage | 38.719 | 6.372 | 5.365 | 101.6 | __11.6__ | 9.0 | 100% |

| CIImage | __14.432__ | __2.352__ | __1.344__ | 62.3 | 15.0 | 10.9 | 29.3% |

| vImage (CG)\* | 31.458 | 5.186 | 5.162 | __55.7__ | __11.7__ | 9.0 | __23.8%__ |

| vImage (CI)\* | 30.431 | 4.933 | 4.445 | - | 15.4 | 11.0 | 28.8% |

\* vImage doesn't have API to load and save image files, so the CGImage and CIImage was used for it.