Wednesday, March 31, 2021

Contour Detection : The Definitive Guide


Hello trend,

Today, we are sharing the definitive guide for Contour Detection in OpenCV. We will also talk about OpenCV Weekly Webinar.

An image is simply an array of numbers (pixel values). To make sense of this array of numbers, we have to extract information from an image so that we can reason at a higher level of abstraction. So, we often pre-process the image using operations like thresholding, or find interesting features like edges. With detected edges, you can do simple shape analysis like finding circles. But if you want to do more interesting shape analysis, you need to contours which are nothing but a chain of connected edges.

In OpenCV, findContours provides a powerful method for finding contours and drawContours help you draw the contours on the image. For beginners, it is not easy to interpret the various inputs to the function. That will why we have written a comprehensive guide that explains how to use findContour in details.

https://learnopencv.com/contour-detection-using-opencv-python-c/

You can find the C++ and Python code at the link below.

https://github.com/spmallick/learnopencv/tree/master/Contour-Detection-using-OpenCV

Beginners in OpenCV often do not recognize that it has a rich collection of over 2500+ algorithms. That's a double edged sword. On one hand, we have a large variety of algorithms to choose from. On the other hand, it is easy to get lost if you do not have a background in computer vision.

To help beginners explore the richness of OpenCV and to gain mastery over important algorithms, we created a series of three Official OpenCV Courses. In these courses, we focus heavily on practical problem solving without drowning students in theoretical nitty gritty. If you are interested in getting started, I recommend beginning with Computer Vision I : Introduction. On the other hand, if you are convinced you want a career in Artificial Intelligence, and want to go all in, I recommend Bundle 4 which is a collection of all three courses.

The courses are usually overbooked, and so we have a waitlist. We always give priority to people on the waitlist and you get a one time chance of getting a 20% discount when the enrollment opens. Click on the button below to instantly join the waitlist.

JOIN THE WAITLIST

OPENCV WEEKLY WEBINAR : MODELPLACE.AI

We started OpenCV Weekly Webinar last week, and without much promotion we got 250+ people on the live webinar. You can checkout last week's episode at

https://youtu.be/80kEhGSlFE0

This week on OpenCV Weekly Webinar, we will discuss an exciting new product - ModelPlace.ai

Modelplace.ai is a marketplace for AI models. We created it to help companies struggling to integrate AI to their processes, and developers aiming to monetize their models.
Modelplace contains many models for a variety of tasks including object detection, pose estimation, text detection, emotion recognition. It is currently in beta and access is by invitation only.
We will invite all the participants of the webinar so you can try out the models right in the browser without checking out any github repo. We have 20 models designed for OpenCV AI Kit - we provide models and code to easily run the models on the device. But the most exciting part of modelplace.ai is a marketplace where users will be able to upload their own models for other people to try out.
Join the webinar to know more and get early access.

Time : 9 AM Pacific Time, April 1, 2021
Hosts : Anna Petrovicheva, Grigory Serebryakov and Satya Mallick.
Link : https://us02web.zoom.us/j/83231402214

Hope to see you there!

Satya

Code for ALL blog posts

Facebook / LinkedIn / Twitter / Instagram

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.

Generate a catchy title for a collection of newfangled music by making it your own

Write a newfangled code fragment at an earlier stage to use it. Then call another method and make sure their input is the correct one. The s...