From 4eb86764539a99b0d7c843d4aa81bfb2955db13c Mon Sep 17 00:00:00 2001
From: Nick White <git@njw.name>
Date: Thu, 18 Apr 2019 13:24:02 +0100
Subject: Add basic cleanup tool; working, but more refinements planned.

This uses integral image calculations, so they have been exported in the binarization package
---
 binarize/integralimg.go |   2 +-
 binarize/sauvola.go     |   2 +-
 cleanup/main.go         | 129 ++++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 131 insertions(+), 2 deletions(-)
 create mode 100644 cleanup/main.go

diff --git a/binarize/integralimg.go b/binarize/integralimg.go
index 75e9cce..382b495 100644
--- a/binarize/integralimg.go
+++ b/binarize/integralimg.go
@@ -14,7 +14,7 @@ type integralwindow struct {
 	height int
 }
 
-func integralimg(img *image.Gray) [][]uint64 {
+func Integralimg(img *image.Gray) [][]uint64 {
 	b := img.Bounds()
 	var oldy, oldx, oldxy uint64
 	var integral [][]uint64
diff --git a/binarize/sauvola.go b/binarize/sauvola.go
index ee773ba..6d9c1af 100644
--- a/binarize/sauvola.go
+++ b/binarize/sauvola.go
@@ -35,7 +35,7 @@ func IntegralSauvola(img *image.Gray, ksize float64, windowsize int) *image.Gray
 	b := img.Bounds()
 	new := image.NewGray(b)
 
-	integral := integralimg(img)
+	integral := Integralimg(img)
 	integralsq := integralimgsq(img)
 
 	for y := b.Min.Y; y < b.Max.Y; y++ {
diff --git a/cleanup/main.go b/cleanup/main.go
new file mode 100644
index 0000000..05ff5f2
--- /dev/null
+++ b/cleanup/main.go
@@ -0,0 +1,129 @@
+package main
+
+import (
+	"flag"
+	"fmt"
+	"image"
+	"image/color"
+	"image/draw"
+	_ "image/jpeg"
+	"image/png"
+	"log"
+	"os"
+
+	"rescribe.xyz/go.git/binarize"
+)
+
+type windowslice struct {
+	topleft uint64
+	topright uint64
+	bottomleft uint64
+	bottomright uint64
+}
+
+func getwindowslice(i [][]uint64, x int, size int) windowslice {
+	maxy := len(i)-1
+	maxx := x+size
+	if maxx > len(i[0])-1 {
+		maxx = len(i[0])-1
+	}
+
+	return windowslice { i[0][x], i[0][maxx], i[maxy][x], i[maxy][maxx] }
+}
+
+// checkwindow checks the window from x to see whether more than
+// thresh proportion of the pixels are white, if so it returns true.
+func checkwindow(integral [][]uint64, x int, size int, thresh float64) bool {
+	height := len(integral)
+	window := getwindowslice(integral, x, size)
+	// divide by 255 as each on pixel has the value of 255
+	sum := (window.bottomright + window.topleft - window.topright - window.bottomleft) / 255
+	area := size * height
+	proportion := float64(float64(area) / float64(sum)) - 1
+	return proportion <= thresh
+}
+
+// cleanimg fills the sections of image not within the boundaries
+// of lowedge and highedge with white 
+func cleanimg(img *image.Gray, lowedge int, highedge int) *image.Gray {
+	b := img.Bounds()
+	new := image.NewGray(b)
+
+	// set left edge white
+	for x := b.Min.X; x < lowedge; x++ {
+		for y := b.Min.Y; y < b.Max.Y; y++ {
+			new.SetGray(x, y, color.Gray{255})
+		}
+	}
+	// copy middle
+	for x := lowedge; x < highedge; x++ {
+		for y := b.Min.Y; y < b.Max.Y; y++ {
+			new.SetGray(x, y, img.GrayAt(x, y))
+		}
+	}
+	// set right edge white
+	for x := highedge; x < b.Max.X; x++ {
+		for y := b.Min.Y; y < b.Max.Y; y++ {
+			new.SetGray(x, y, color.Gray{255})
+		}
+	}
+
+	return new
+}
+
+func main() {
+	flag.Usage = func() {
+		fmt.Fprintf(os.Stderr, "Usage: cleanup [-t thresh] [-w winsize] inimg outimg\n")
+		flag.PrintDefaults()
+	}
+	wsize := flag.Int("w", 5, "Window size for mask finding algorithm.")
+	thresh := flag.Float64("t", 0.05, "Threshold for the proportion of black pixels below which a window is determined to be the edge.")
+	flag.Parse()
+	if flag.NArg() < 2 {
+		flag.Usage()
+		os.Exit(1)
+	}
+
+	f, err := os.Open(flag.Arg(0))
+	defer f.Close()
+	if err != nil {
+		log.Fatalf("Could not open file %s: %v\n", flag.Arg(0), err)
+	}
+	img, _, err := image.Decode(f)
+	if err != nil {
+		log.Fatalf("Could not decode image: %v\n", err)
+	}
+	b := img.Bounds()
+	gray := image.NewGray(image.Rect(0, 0, b.Dx(), b.Dy()))
+	draw.Draw(gray, b, img, b.Min, draw.Src)
+
+	integral := binarize.Integralimg(gray)
+	maxx := len(integral[0])-1
+	var lowedge, highedge int = 0, maxx
+	// find right edge
+	for x := maxx / 2; x < maxx - *wsize; x++ {
+		if checkwindow(integral, x, *wsize, *thresh) {
+			highedge = x
+			break
+		}
+	}
+	// find left edge
+	for x := maxx / 2; x > 0; x-- {
+		if checkwindow(integral, x, *wsize, *thresh) {
+			lowedge = x
+			break
+		}
+	}
+
+	clean := cleanimg(gray, lowedge, highedge)
+
+	f, err = os.Create(flag.Arg(1))
+	if err != nil {
+		log.Fatalf("Could not create file %s: %v\n", flag.Arg(1), err)
+	}
+	defer f.Close()
+	err = png.Encode(f, clean)
+	if err != nil {
+		log.Fatalf("Could not encode image: %v\n", err)
+	}
+}
-- 
cgit v1.2.1-24-ge1ad