I have bought a Lidar Lite V3 (not HP) and noticed the same issue as you have, that the reading is incorrect up to 90 cm. I suspect this is because the signal is too strong and thus impacting noise/peak level.
On my desk, I have plotted a line of 100 cm, with a mark every 5 cm. The Lidar at the start (0cm) and a box to reflect back the Lidar Laser signal moving every 5 centimeters. I have then adjusted the sketch to show the average of 100 readings and then make an average of that.
Then made a correction table. When the measured distance returned is below 100, you can apply the correction that belongs to the range.
E.g. if you get a distance reading of 55 cm, that is in between 53.55 and 57.38 thus the correction factor is (0.22 + 0.25) / 2 = 0.23 to apply. You get 42 cm , 55 *( 1- 0.23) , real distance.
Real measured Correction
100 100,54 0,01
95 96,76 0.02
90 95,2 0.05
85 83,37 -0.02
80 81,12 0.01
75 77,4 0.03
70 75,41 0.07
65 72,46 0.10
60 69,14 0.13
55 65,12 0.16
50 61,92 0.19
45 57,38 0.22
40 53,55 0.25
35 43,25 0.19
30 40,16 0.37
25 39,72 0.40
20 33,24 0.44
15 26,73 0.55
10 22,2 0.55
5 18,33 0.73
While this is not 100% accurate, there are remarks to add :
-
When looking at the average of 100 readings, without moving the box the number was fluctuating every 100 readings… WHY ?
-
If the 100 cm line was not 100% straight in front of the sensor the readings would differ a lot.
This makes me believe that applying a correction table is providing good enough information and no need for an extra sensor.
Just an alternative thought. 