Gait Analysis

Human locomotion is achieved through a complex interaction of muscles, skeleton and the nervous system. Possible impairments of one of these anatomical systems can be detected early on by means of a gait analysis. For example, step length, walking speed, step frequency, joint angle, forces acting on the joints, muscle activity and energy consumption are evaluated. This information can be recorded using sensors.

In cooperation with Prof. Kerstin Lüdtke (Physiotherapy, University of Lübeck) and the company Future-Shape GmbH, we are developing artificial intelligence methods to automate and objectify gait analysis. For the labelling of the required data we use, among others, the SensFloor, RGB cameras, depth sensors and wearables (IMUs). The SensFloor records all places where the foot touches the ground while walking. From this we can then calculate the geometric and temporal information of steps. IMUs can be attached directly to the body and provide data on acceleration, rotation rates and patient alignment. In addition, we extract the joint angles and positions from the camera images. We use all these data to distinguish between pathological and healthy gait patterns.

From an AI methodology point of view, we are exploring deep learning methods in this context, but also an algorithm developed by us, which calculates the spatio-temporal motion characteristics of the human gait for individual identification without the need for traditional silhouette extraction from video sequences. The extracted local motion data are encoded in high-level descriptors using the so-called codebook approach based on the Gaussian mixed model and classified with a simple linear support vector machine. We have already tested our pattern recognition platform on five widely used gait databases and obtained a recognition rate of up to 98%.


Khan, Muhammad Hassan and Farid, Muhammad Shahid and Grzegorzek, Marcin. A Non-linear View Transformations Model for Cross-view Gait Recognition. Elsevier Neurocomputing, April 2020.


Khan, Muhammad Hassan, Muhammad Shahid Farid, and Marcin Grzegorzek. Spatiotemporal features of human motion for gait recognition. Signal, Image and Video Processing 13.2 (2019): 369-377.