segmentator

segmentator

This is a simple library for image segmentation written in Golang. I need it for my CG synopsis (it is in synopsis folder), but hope it'll be usefull for You too :)

structure of the library

1. Loading and Saving Images:

   • Load Image: func LoadImage(path, name string) (img Image, err error)
   • Save Image: func SaveImage(path, name string, img Image) (err error)

2. Grayscale algorithms:

   • Averaging ("Quick and Dirty"): func GSAveraging(img Image)
   • Correcting the human eye ("Luma"): func GSLuma(img Image)
   • Desaturation: func GSDesaturation(img Image)
   • Decomposition (Maximal and Minimal): func GSDecomposition(img Image, do int) (err error)
   • Single Color Channel (Red, Green or Blue): func GSDecomposition(img Image, do int) (err error)

3. First generation algorithms:

   • Based on Edge-Detection:
     • Roberts operator: func FGEDRoberts(img Image, do int) (err error)
     • Previtt operator: func FGEDPrevitt(img Image, do int) (err error)
     • Sobel operator: func FGEDSobel(img Image, do int) (err error)
     • Scharr operator: func FGEDScharr(img Image, do int) (err error)
     • Custom operator: func FGEDCustomOperators(img Image, GxOp, GyOp [][]int, do int) (err error)
   • Based on Pixel Classification:
     • Simple Iterative: func FGPCIterative(img Image) (threshold int)
     • Otsu: func FGPCOtsuThresholding2(img Image) (threshold int)
     • Custom: FGPCThreshold(img Image, thresholds []int, colors []Pixel) (err error)

4. Helpers:

   • abs: func abs(value int) int
   • min: func min(values ...int) int
   • max: func max(values ...int) int

important notes

1. Disputes on which coordinate system to use when processing images continues to the day. I prefer Gonsales-Woods varinant: