Human Body Detection with Occlusion

Session Number

Project ID: CMPS 08

Advisor(s)

Rui Chen; Carnegie Mellon University, Robotics Institute / Intelligent Control Lab

Dr. Changliu Liu; Carnegie Mellon University, Robotics Institute / Intelligent Control Lab

Discipline

Computer Science

Start Date

19-4-2023 10:50 AM

End Date

19-4-2023 11:05 AM

Abstract

The purpose of this design experiment was to attempt to determine the accuracy/possibility of using Kalman filters to approximate a human’s location when blocked by an occlusion/obstacle. This could potentially be implemented in the field of human-robot interaction to allow humans and robots to collaborate without interference. The outcome of this research project could similarly result in more accurate measurements for human body detection.

This project consists of the following two different procedures: 1) design steps of the solution, 2) testing final design iteration. In order to construct the hypothetical solution to the project, the first procedure must be followed, while to test the solution, the second procedure must be followed. Each procedure is extremely descriptive, explaining the several steps to building and testing the experimental solution precisely.

Based on the results of the experiment, the problem statement was proved inconclusive due to both positive and negative results. Throughout all the data collected and graphs constructed, the Kalman filter’s average mean-squared errors were never concretely/always lower than those of the ZED depth camera. The project’s findings reveal the inconclusive and ambiguous behavior of the Kalman filter, indicating more trials and testing is needed for a concrete determination.

Share

COinS
 
Apr 19th, 10:50 AM Apr 19th, 11:05 AM

Human Body Detection with Occlusion

The purpose of this design experiment was to attempt to determine the accuracy/possibility of using Kalman filters to approximate a human’s location when blocked by an occlusion/obstacle. This could potentially be implemented in the field of human-robot interaction to allow humans and robots to collaborate without interference. The outcome of this research project could similarly result in more accurate measurements for human body detection.

This project consists of the following two different procedures: 1) design steps of the solution, 2) testing final design iteration. In order to construct the hypothetical solution to the project, the first procedure must be followed, while to test the solution, the second procedure must be followed. Each procedure is extremely descriptive, explaining the several steps to building and testing the experimental solution precisely.

Based on the results of the experiment, the problem statement was proved inconclusive due to both positive and negative results. Throughout all the data collected and graphs constructed, the Kalman filter’s average mean-squared errors were never concretely/always lower than those of the ZED depth camera. The project’s findings reveal the inconclusive and ambiguous behavior of the Kalman filter, indicating more trials and testing is needed for a concrete determination.