I am looking for someone to write a piece of software for me.
The purpose of the software is to be able to detect objects through a webcam, then write the position of it's parts.
A good comparison would be the xbox kinect software which can detect the pose of a person's body. It should be able to detect the pose of a person's body, but also similar information on other objects. For instance, it should be able to :
- detect which way a vehicle is facing
- whether an animal is sitting or standing
- Whether a book is open or shut
Obviously, developing a massive database of every type of object is the world would be an impossible task. So, what I'm expecting is a software that reduces all objects to their basic shapes.
For example, when it seems a human, it divides them up into a sphere/circle for their head, an oblong for their torso, and smaller oblongs for their limbs. It then writes the coordinates of the center of each object, their size, and rotation.
Similarly, for the animal, the software would divide it up into basic geometric shapes then write the coordinates and information for each part. Similar things can be done with a book.
This is only one example of how I could see the software working however. If you know of a more practical way of doing this, I am still interested in hearing your proposal.
The software will need to run on Windows.
16 freelance hanno fatto un'offerta media di $2436 per questo lavoro
Hi, How are you ? I am a Senior Full stack developer. I have a strong skills for OpenCV. Before I have experience for Pose Detection by OpenCV. I like to discuss the project with you. Thanks Lee
I can develop this in NodeJs or Python language. The software will have all the features you have mentioned. I will have to implement neural networks and train the algorithm.