From the images of the tools on a piece of paper I will ask you to find their outline contour to vectorize them.
I am new to computer vision problems and the only thing I was thinking about is OpenCV and border detection.
The result is better than I imagined, it is still very unreliable, especially if the original image is not "perfect".
I took 2 photos of the key that they gave me.
After playing with opencv binding for node , I get the following:
Then, I tried with a less good picture:
It is completely useless. I can get something a little better by modifying Canny thresold, but this should be automated (given that the image is relatively correct).
So, I have a few questions:
- ? GrabCut ? Grabcut Canny? , , GrabCut , .
- .
approxPolyDP, . - , -, ,
approxPolyDP. ? - , . OpenCL, , , , ? - , , ? , , , , ( ).
- Canny thresold, . ?
- , , . ? ( btw!)
:
const cv = require('opencv');
const lowThresh = 90;
const highThresh = 90;
const nIters = 1;
const GRAY = [120, 120, 120];
const WHITE = [255, 255, 255];
cv.readImage('./files/viv1.jpg', function(err, im) {
if (err) throw err;
width = im.width()
height = im.height()
if (width < 1 || height < 1) throw new Error('Image has no size');
const out = new cv.Matrix(height, width);
im.convertGrayscale();
im_canny = im.copy();
im_canny.canny(lowThresh, highThresh);
im_canny.dilate(nIters);
contours = im_canny.findContours();
let maxArea = 0;
let biggestContour;
for (i = 0; i < contours.size(); i++) {
const area = contours.area(i);
if (area > maxArea) {
maxArea = area;
biggestContour = i;
}
out.drawContour(contours, i, GRAY);
}
const arcLength = contours.arcLength(biggestContour, true);
contours.approxPolyDP(biggestContour, 0.001 * arcLength, true);
out.drawContour(contours, biggestContour, WHITE, 5);
out.save('./tmp/out.png');
console.log('Image saved to ./tmp/out.png');
});