Apr 14, 2019
Human-Body Skeleton Detection using OpenCV
Objective : A model is developed to detect the major key points of the body in the image/video using the image/video as the input stream and track the movements of the key–points using OpenCV to reduce the overheads of installation of Motion Sensing Technology which uses Infrared Sensors..
Project Link - Human-Body Skeleton Detection using OpenCV
Motivation - Why do we need to detect the key-points of the human body?
Motion sensing technology is increasingly deployed in many applications where human detection is reqired such as gaming, security and military.
Motion sensing input device such as Microsoft's Xbox 360 Kinect provides this applicability using infrared sensors, IR sensors are much more expensive comapred to optical cameras.
Installation procedure is hectic and inconvenient to be widely used.
Datasets used for training the model - COCO dataset and MPII dataset.
Applications - This model can be integrated into the systems for various purposes. Few of them are listed below:
Game Strategy - It can be used to understand the gameplay of a player and his next moves to improve the game strategy.
Orthopaedic patient diagnosis - It can be used by the doctors to understand the key points while analysing the body postures of the patients.
Injury Analysis of Sportsperson - It can be used by the sports authorities to live track the player's movements and detect the key-points of the body to understand the injury reasons if occured any.
Gymnasium - It can be used by the gymnastic person to improve their movement's in order to prevent future injuries.
Surveillance Activities - Action and behaviour analysis by gestures, body movements, and detection of abnormal activities.
This project was selected for Industrial Academic Meet (IAM) 2019 held at IIIT Raipur under the track - Data Science and Machine Learning.
