Other Parts Discussed in Thread: IWR6843
Tool/software:
Hi,
After signing an NDA with TI, I obtained the source code for the Vital Signs with People Tracking Demo from you. Recently, while studying this source code, I've noticed an issue with the data being used within the cardiopulmonary calculation logic. I've captured the real-time data required for cardiopulmonary calculations through recording, and upon simulation analysis, confirmed that the recorded data is correct. However, when comparing the captured data with the cardiopulmonary data used internally by the Demo, theoretically, these two sets of data should match, but in reality, the internal data has issues. As a result, the final detection results of the Demo exhibit significant errors.
Below, I've provided a comparison of the two sets of data from a single channel and a range bin. The first graph represents the correct data, while the second graph displays the data used in the Demo. It's evident that while the general trend is similar, the data in the second graph contains numerous zero values and some other erroneous segments, with only partial data being accurate. The current logic does not reveal the source of the problem, and I would appreciate it if you could provide technical support to help me understand the issue.
I'm looking forward to your reply. Thank you very much!
Additional clarification: For the captured data, the first graph's data was recorded within the function MmwDemo_vitalsignPickupPoint, where the data storage logic resides, and an additional output copy was saved. The second graph's data was captured from within the cardiopulmonary processing function MmwDemo_vitalsignProcessing, where an output copy of the input data was saved separately.
