GmAPD data normalization using Bernoulli trials
Lidar systems may use highly sensitive GmAPD detectors to track obstacles in the environment of an autonomous vehicle. Data sensed by a lidar GmAPD detector can be pre-conditioned to facilitate differentiating low intensity signals from background noise. By sampling raw avalanche counts accumulated by the detector as Bernoulli trials, binomial statistics can be leveraged to transform raw data to a probability distribution, and then to a normalized data set suitable for signal processing.
1. A method, comprising:
detecting laser signals reflected from a target, by sensing an accumulation of single photons by a photon detector;
creating an avalanche histogram from the detected laser signals;
transforming the avalanche histogram to an avalanche probability histogram by framing raw data from the photon detector as a sequence of Bernoulli trials within a timestamp interval and applying a binomial confidence estimation;
transforming the avalanche probability histogram into a linearized intensity histogram by correcting waveform distortion; and
determining a photon intensity of the reflected laser signals from an average count rate and an average photon flux rate associated with the linearized intensity histogram.
2. The method of claim 1 , wherein detecting the laser signals comprises using a high sensitivity Geiger-mode avalanche photodiode (GmAPD) detector as the photon detector to sense the accumulation of the single photons.
3. The method of claim 1 , further comprising merging adjacent Bernoulli trials when a reflected laser signal spans multiple time bins of the avalanche histogram.
4. The method of claim 1 , wherein transforming the avalanche histogram to an avalanche probability histogram comprises normalizing lidar data in which signals are distinguished relative to background noise.
5. The method of claim 4 , wherein normalizing the lidar data suppresses multiple peaks caused by re-arming the photon detector.
6. The method of claim 1 , further comprising superimposing a tunable noise threshold curve on one or more of the avalanche histogram, the avalanche probability histogram, and the linearized intensity histogram.
7. The method of claim 6 , wherein superimposing the tunable noise threshold depends on a user-specified minimum confidence threshold.
8. A system, comprising:
a memory configured to store instructions; and
at least one processor coupled to the memory and configured to execute the instructions to perform operations comprising:
detecting laser signals reflected from a target, by sensing an accumulation of single photons by a photon detector;
creating an avalanche histogram from the detected laser signals;
transforming the avalanche histogram to an avalanche probability histogram by framing raw data from the photon detector as a sequence of Bernoulli trials within a timestamp interval and applying a binomial confidence estimation;
transforming the avalanche probability histogram into a linearized intensity histogram by correcting waveform distortion; and
determining a photon intensity of the reflected laser signals from an average count rate and an average photon flux rate associated with the linearized intensity histogram.
9. The system of claim 8 , wherein the at least one processor is configured to perform operations further comprising merging adjacent Bernoulli trials when a reflected laser signal spans multiple time bins of the avalanche histogram.
10. The system of claim 8 , wherein transforming the avalanche histogram to an avalanche probability histogram comprises normalizing lidar data in which signals are distinguished relative to background noise.
11. The system of claim 8 , wherein detecting the laser signals comprises using a high sensitivity Geiger-mode avalanche photodiode (GmAPD) detector as the photon detector to sense the accumulation of the single photons.
12. The system of claim 8 , wherein the at least one processor is configured to perform operations further comprising superimposing a tunable noise threshold curve on one or more of the avalanche histogram, the avalanche probability histogram, and the linearized intensity histogram.
13. The system of claim 12 , wherein superimposing the tunable noise threshold depends on a user-specified minimum confidence threshold.
14. A non-transitory computer-readable medium having instructions stored thereon that, when executed by at least one computing device, cause the at least one computing device to perform operations comprising:
detecting laser signals reflected from a target, by sensing an accumulation of single photons by a photon detector;
creating an avalanche histogram from the detected laser signals;
transforming the avalanche histogram to an avalanche probability histogram by framing raw data from the photon detector as a sequence of Bernoulli trials within a timestamp interval and applying a binomial confidence estimation;
transforming the avalanche probability histogram into a linearized intensity histogram by correcting waveform distortion; and
determining a photon intensity of the reflected laser signals from an average count rate and an average photon flux rate associated with the linearized intensity histogram.
15. The non-transitory computer-readable medium of claim 14 , wherein detecting the laser signals comprises using a high sensitivity Geiger-mode avalanche photodiode (GmAPD) detector as the photon detector to sense the accumulation of the single photons.
16. The non-transitory computer-readable medium of claim 14 , further comprising merging adjacent Bernoulli trials when a reflected laser signal spans multiple time bins of the avalanche histogram.
17. The non-transitory computer-readable medium of claim 14 , wherein transforming the avalanche histogram to an avalanche probability histogram comprises normalizing lidar data in which signals are distinguished relative to background noise.
18. The non-transitory computer-readable medium of claim 17 , wherein normalizing the lidar data suppresses multiple peaks caused by re-arming the photon detector.
19. The non-transitory computer-readable medium of claim 14 , further comprising superimposing a tunable noise threshold curve on one or more of the avalanche histogram, the avalanche probability histogram, and the linearized intensity histogram.
20. The non-transitory computer-readable medium of claim 19 , wherein superimposing the tunable noise threshold depends on a user-specified minimum confidence threshold.