Hi,
I please need a relatively simple application constructed for video-based golf club tracking and analysis. Previous attempts have been partially successful, so there is some work there that can be referenced and potentially utilized.
A standard 2D camera and video are all that are required for acceptable application performance. Simple marker tracking is planned at this time, comprising two visible markers placed at determined, measured locations on a club to be tracked.
Original videos input into the needed application will be relatively short, limited to perhaps a few seconds or less. The application will initially need to reformat videos to include frame numbers for user and application referencing. After careful examination of the reformatted video by a user as instructed, the user will select just two frames (so four tracking markers) for the application to further work with.
Tracking should be performed just manually and interactively for now. So a means should be provided for a user to view, manually place, and precisely adjust computer tracking indicators over each of the four markers in the two frames selected. With that information, combined with specified club measurements input into another area of the application by the user, an initial club tracking result can be graphically and mathematically displayed.
That tracking result will typically reveal a preliminary rotation point location along the length of the club (the needed solution). But due to the nature of the original club movement, a preliminarily calculated rotation point location will routinely be incorrect, commonly by a substantial amount. Thus, an additional process involving some fairly basic geometric and mathematic calculations will need to be further applied (will be provided). And a corrected, final rotation point location as calculated will be the result, graphically and mathematically displayed by the application for a user.
I will need to test and work with any efforts on my (Windows) desktop. Desktop and web versions are minimally planned for now. So a programming language(s) that could work effectively for both would be helpful. R and Python were largely tried previously (with various pluses and minuses about each), and I am open to any other language(s) as well.
Additional descriptions needed for both the video tracking and solution facets to accomplish functional success are much more thoroughly detailed in separate documentation. That is not limited to systematic breakdowns of previous failed attempts for reference, what worked well and not so well with them, and the processes and equations to be applied to calculate final solutions.
Various elements of intellectual property are involved. So while I can additionally discuss some matters beforehand, an NDA will be strictly required before I will be able to disclose and/or discuss certain information included in the more detailed documentation. I can provide the needed NDA after reviewing any proposals I might be fortunate enough to receive.
While there is no set budget, past failures have prevented me from better validating certain critical elements concerning the needed application on an extremely foundational level. And until such better proof is obtained first, my additional spending on this particular project will be as minimal as possible to say the very least. If and when better validated, I can reevaluate where to go from there.
I am open to either project or hourly rate offers, but I cannot seriously consider any hourly rate offer without a guaranteed maximum number of hours stated up front to successfully complete the needed application. Thank you very much.
... Show more