constraining recognition to numbers

l3dlp-sandbox · Jan 12, 2015 · 843ab6c · 843ab6c
1 parent 4f1e5a4
commit 843ab6c
Show file tree

Hide file tree

Showing 2 changed files with 91 additions and 0 deletions.
diff --git a/examples/numbers.html b/examples/numbers.html
@@ -0,0 +1,91 @@
+<!doctype html>
+<html>
+	<head>
+	</head>
+	<body>
+		<script src="../ocrad.js"></script>
+		<style>
+			.highlight {
+				position: absolute;
+				background: yellow;
+				opacity: 0.5;
+				width: 10px;
+				height: 10px;
+			}
+		</style>
+		<script>
+
+			function OCRImage(image){
+				var canvas = document.createElement('canvas')
+				canvas.width  = image.naturalWidth;
+				canvas.height = image.naturalHeight;
+				canvas.getContext('2d').drawImage(image, 0, 0)
+				var raw = []
+				var text = OCRAD(canvas, 
+							false, // don't invert the colors (i.e. set this to true if you have white on black text)
+							function(line){ // this is for giving the Ocrad raw format which includes letter bounding boxes
+								raw.push(line); // its a function that gets called for each line of stdout
+							});
+
+				var letters = parseOcrad(raw);
+				console.log(letters) // here's a list of recognized letters and their corresponding coordinates and confidences
+
+				letters.map(function(letter){
+					// select the letters which have vanna white's approval
+					return letter.matches.filter(function(match){
+						return /\d/.test(match.letter)
+					})
+				}).forEach(function(candidates){
+					if(candidates.length > 0){
+						console.log(candidates[0].letter)
+					}else{
+						console.log('_')
+					}
+				})
+			}
+
+
+			function parseOcrad(raw){
+				var raw = raw.map(function(e){
+					return e.match(/^\s*(\d+)\s+(\d+)\s+(\d+)\s+(\d+)\s*;\s*(\d+)(\,?.+)?$/)
+				}).filter(function(e){
+					return e
+				}).map(function(e){
+					var x = parseInt(e[1]),
+						y = parseInt(e[2]),
+						w = parseInt(e[3]),
+						h = parseInt(e[4]),
+						g = parseInt(e[5]);
+
+					var matches = [];
+					if(g > 0){
+						var etc = e[6].trim();
+						while(etc[0] == ',' && etc[1] == ' '){
+							etc = etc.slice(2)
+							var m = etc.match(/^\'(.+?)\'(\d+)/)
+							matches.push({
+								letter: m[1], 
+								confidence: parseInt(m[2])
+							})
+							etc = etc.slice(m[0].length)
+						}
+					}
+					console.log(matches)
+					if(matches.length != g) console.error('recognition count mismatch', g, matches);
+					return {
+						x: x,
+						y: y,
+						width: w,
+						height: h,
+						matches: matches
+					}
+				});
+
+				return raw;
+			}
+		</script>
+		<img src="numbers.png" onload="OCRImage(this)">
+	</body>	
+</html>
+
+
diff --git a/examples/numbers.png b/examples/numbers.png