AI Rd-03E - 10 Considerations

Rd-03E Gesture Recognition User Manual V1.0.0
Copyright © 2023 Shenzhen Ai-Thinker Technology Co., Ltd All Rights Reserved
Page 15 of
18
reference value
PARAM_SLIDEFRAMECNT
-4
Negative Doppler
Reference Value
ARAM_DOPPLERNUM_P
1
number of positive
doppler
PARAM_DOPPLERNUM_N
1
number of negative
doppler
PARAM_HIGHSPEED_DOPPLERVALUE
9
high-speed Doppler value
PARAM_INVALID_FRAME
14
Invalid Frame Count
PARAM_TARGET_RNGBIN
3
Gesture Distance Unit
PARAM_NOISE_COEF
2.2
bottom noise coefficient
PARAM_HIGHSPEED_DOPPLERNUM
0
high speed doppler
quantity
PARAM_WAVECNT_THD
3
Swing Statistics
PARAM_PNNUM_CNT
2
Number of positive and
negative alternations
PARAM_WAVENUM_THD
50
frame times threshold
PARAM_GESTURECNT_THD
2
Half swing threshold
10 Considerations
⚫ At present, the algorithm is debugged in three ways: bare module, silicon microshell and
silicon microshell + toilet cover. Since there may be differences in the structure of toilet
covers of different brands and models, when the customer places this module in the toilet,
he may encounter abnormal ranging or abnormal gesture. At this time, please print 1DFFT
data or rdmap data, and adjust the link gain or detection parameters according to the data.
⚫ At present, this firmware only supports continuous ranging and gesture output, and does
not enable discharge level trigger or PWM mode. Customers can independently develop it
in the released SDK according to specific requirements.