A Comprehensive Review of Vision-Based Sensor Systems for Human Gait Analysis

Area of Science:

• Biomechanics
• Clinical Research
• Interdisciplinary Studies

Background:

• Human gait analysis is crucial in biomechanics and clinical research.
• Advancements in visual sensors and machine learning have driven progress in gait analysis systems.
• This review focuses on vision-based systems, examining recent developments over the last five years.

Purpose of the Study:

• To provide a comprehensive review of advancements in vision-based human gait analysis systems.
• To emphasize the roles of vision sensors, machine learning algorithms, and technological innovations.
• To identify trends and establish foundations for future gait analysis tool development.

Main Methods:

• Systematic literature review using the PRISMA method.
• Analysis of 72 selected research articles.
• Detailed examination of sensor systems, algorithms, parameters, and methods.

Main Results:

• Non-invasive depth cameras are emerging as a popular choice for gait analysis.
• Deep learning algorithms, including Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM) networks, are increasingly utilized.
• Key aspects reviewed include sensor types, algorithms, gait parameters, camera placement, and extraction techniques.

Conclusions:

• The field is trending towards non-invasive depth sensing and advanced deep learning.
• Future innovations aim for more effective, versatile, and user-friendly gait analysis tools.
• Enhanced gait analysis holds significant potential to improve human mobility, health, and quality of life.