Advancements in markerless motion capture have been raising a lot of questions about its impact on the future of biomechanics. Will markerless allow for significant increases in the scale of data collection and the diversity of its applications? How will the overall quality of data be affected? What barriers must we overcome to achieve widespread adoption? What does the future hold for biomechanical research and motion capture?
When an industry builds up around a particular technology, it can be challenging to accept any alternatives. Marker-based systems have been the industry standard because, until recently, no other viable alternatives existed. What would markerless motion capture need to do to become the new industry standard?
Cost, hardware limitations, and hesitance of the unfamiliar will all slow adoption of this emerging technology. Purchasing several high-quality cameras is costly, and markerless systems require significantly more computing power and storage than marker-based systems. However, as technology advances and more computing shifts to the cloud, these problems will resolve over time.
How concerned should we be about the accuracy and reliability of markerless? We know markers have several inherent issues, dependency on an individual’s marker placements can sometimes complicate comparisons of data collected at different labs, or even on different days. Even when correctly placed, soft tissue artifacts (the movements of muscle and skin independent from bone) lead to more potential inaccuracies. Finally, marker based equipment will often affect a subject’s performance, limiting range of motion or inhibiting natural movement. Removing markers would mean trials collected on different days or in different labs would all be consistently comparable, and we can ensure that participant performance remains unaffected.
We know markerless systems boast improved ease of use and global consistency, but how does the raw data compare? Rob Kanko et al conducted a study “Concurrent assessment of gait kinematics using marker-based and markerless motion capture” testing the two systems by running simultaneous trials with both, finding comparable results from marker based and markerless when measuring joint centers and angles across thirty subjects, joint center positions differed by less than 3 cm for all joints except the hip, which differed by 3.6 cm Rob noted that these differences are similar in magnitude to the differences found when comparing different marker based joint center location techniques.
These findings show promise for the future of markerless systems, but they also call into question the value of comparing markerless with marker based. Researchers have not conducted any studies comparing a markerless motion capture system to Biplanar Fluoroscopy, the established gold standard in pose estimation. Until researchers conduct such studies, it is hard to be certain whether the deviations found when comparing markerless and marker-based systems result from inaccuracies in the markerless system or the marker-based system.
Regardless of the lack of meaningful comparisons there is enough research to indicate that markerless is an effective tool for measurement, so what new possibilities will emerge that were once out of reach? Researchers could conduct trials in places previously impossible, such as lowlight environments, crowded spaces, and even in the middle of a sporting event. This also opens the door for people who lack the technical skills needed to apply a markerset, a clinician could run a motion trial at each appointment to track progress or assess any impairments, or monitor mobility over the course of someone’s life. This widespread use will result in data-sets far larger than any possibly collected via marker based systems; so what trends will come to light when we number participants in the thousands? What new insights will novel applications provide?
Ultimately, beyond speculations we don’t know where this technology will take us. However, HAS-Motion is eager and excited to offer software solutions to the expanding applications of motion capture.
