Caracterización de movimientos a través de imágenes y sensores inerciales para la prevención de lesiones de miembro superior durante actividades repetitivas Thesis

short description

  • Master's thesis

Thesis author

  • Jasbón Mutis, Adriana Samira

external tutor

  • Perdomo Charry, Oscar Julián

abstract

  • The upper limb is a structure of great complexity and importance in our daily lives due to the variety of movements and degrees of freedom it has. As well as its usefulness in numerous daily tasks. The upper limb can be affected by injuries due to tasks that include: efforts, inadequate postures, postures out of range of motion, and sustained or repetitive postures; during long periods, such as work activities, generating a negative impact on daily life. For this reason, there is a need for studies on the complexity of upper limb movement for the prevention of this type of injury. This study focuses on the development of an upper limb kinematics simulation tool using resources such as images and inertial sensors for motion capture. For the development of a supporting tool or software capable of identifying movement patterns using sources of motion information in people, captured by sensors, in a work environment for further study. This project aims to design a method for the automatic identification of movement patterns associated with the upper limb by fusing information from video images and inertial markers. To this end, a review of programming and data processing methods obtained from images and inertial sensors, their programming, the acquisition of data related to posture, the study of movement, and the calibration of the software are proposed. Among the results obtained, a portable software for motion capture using images was created, which provides results of RMS angles similar to those given by the inertial sensors. However, the acceleration presents a greater difference, because the acceleration is affected by the moments in which the images oscillate. In conclusion, it was found a great potential for artificial intelligence algorithms for the identification, tracking, and differentiation of posture and the calculation of kinematics variables such as joint movement angle and angular acceleration.

publication date

  • August 9, 2023 6:23 PM

keywords

  • Biomechanics
  • Gestures
  • Inertial sensors
  • Mediapipe
  • Movement study,
  • Upper limb

Document Id

  • 5ca1d8e5-756e-4962-985c-19ae47f8391b