Golang : Find and draw contours with OpenCV example




Ability to detect shapes is often the most important steps in allowing a computer to make decision...such as deciding if the shape is a circle or rectangle. There are a couple of ways to detect shapes in an image. One of them is to find contours on the image.

Basically, contours are a set of points connected to each other to form an outline of an object(kinda like drawing along the dotted line in children coloring books)

find and draw contours implementation in Golang

The Golang code example below illustrates how to extract contours with FindContours() function and draw contours back on the original image with DrawContours() function. Bear in mind that Golang implementation is slightly different from C/C++ or Python code examples found on OpenCV website, but the end result is the same.

Here you go!


 package main

 import (
 "fmt"
 "github.com/lazywei/go-opencv/opencv"
 "os"
 )

 var (
 originalWindow = new(opencv.Window)
 image = new(opencv.IplImage)
 seq *opencv.Seq
 redColor = opencv.NewScalar(0, 0, 255, 255) // (blue, green, red, alpha)
 blackColor = opencv.NewScalar(0, 0, 0, 255) // (blue, green, red, alpha)
 blueColor = opencv.NewScalar(255, 0, 0, 255) // (blue, green, red, alpha)
 slider = 15
 )

 func trackBar(position int, param ...interface{}) {

 width := image.Width()
 height := image.Height()

 // Convert to grayscale
 gray := opencv.CreateImage(width, height, opencv.IPL_DEPTH_8U, 1)
 defer gray.Release()

 opencv.CvtColor(image, gray, opencv.CV_BGR2GRAY)

 // for edge detection
 cannyImage := opencv.CreateImage(width, height, opencv.IPL_DEPTH_8U, 1)
 defer cannyImage.Release()

 // Run the edge detector on grayscale
 //opencv.Canny(gray, cannyImage, float64(position), float64(position)*2, 3)

 // ** For better result, use 50 for the canny threshold instead of tying the value
 // to the track bar position
 opencv.Canny(gray, cannyImage, float64(50), float64(50)*2, 3)

 // Find contours sequence from canny edge processed image
 // see http://docs.opencv.org/2.4/modules/imgproc/doc/structural_analysis_and_shape_descriptors.html
 // for mode and method

 seq = cannyImage.FindContours(opencv.CV_RETR_TREE, opencv.CV_CHAIN_APPROX_NONE, opencv.Point{0, 0})
 defer seq.Release()

 // based on the sequence, draw the contours
 // back on the original image
 finalImage := image.Clone()

 maxLevel := position
 opencv.DrawContours(finalImage, seq, redColor, blueColor, maxLevel, 2, 8, opencv.Point{0, 0})
 originalWindow.ShowImage(finalImage)

 fmt.Printf("Levels  = %d\n", position)
 }

 func main() {

 if len(os.Args) != 2 {
 fmt.Printf("Usage : %s <image filename>\n", os.Args[0])
 os.Exit(0)
 }

 imageFileName := os.Args[1]

 fmt.Println("Loading image from ", imageFileName)
 fmt.Println("Press ESC key to quit")

 image = opencv.LoadImage(imageFileName)
 if image == nil {
 panic("LoadImage failed")
 }
 defer image.Release()

 originalWindow = opencv.NewWindow("Find contours in image")
 defer originalWindow.Destroy()

 originalWindow.CreateTrackbar("Levels : ", 0, slider, trackBar)

 // initialize by drawing the top level contours(level = 0)
 trackBar(0, 0)

 for {
 key := opencv.WaitKey(20)
 if key == 27 {
 os.Exit(0)
 }
 }

 os.Exit(0)
 }

References:

https://socketloop.com/tutorials/golang-eroding-and-dilating-image-with-opencv-example

http://docs.opencv.org/3.2.0/d4/d73/tutorialpycontours_begin.html

  See also : Golang : Eroding and dilating image with OpenCV example





By Adam Ng

IF you gain some knowledge or the information here solved your programming problem. Please consider donating to the less fortunate or some charities that you like. Apart from donation, planting trees, volunteering or reducing your carbon footprint will be great too.


Advertisement