Thumbnail Image

Investigation of jitter on full-field amplitude modulated continuous wave time-of-flight range imaging cameras

The time-of-flight (ToF) range imaging cameras indirectly measure the time taken from the modulation light source to the scene and back to the camera and it is this principle that is used in depth cameras to perform depth measurements. This thesis is focused on ToF cameras that are based on the amplitude modulated continuous wave (AMCW) lidar techniques which measure the phase difference between the emitted and reflected light signals. Due to their portable size, feasible design, low weight and low energy consumption, these cameras have high demand in many applications. Commercially available AMCW ToF cameras have relatively high noise levels due to electronic sources such as shot noise, reset noise, amplifier noise, crosstalk, analogue to digital converters quantization and multipath light interference. Many noise sources in these cameras such as harmonic contamination, non-linearity, multipath interferences and light scattering are well investigated. In contrast, the effect of electronic jitter as a noise source in ranging cameras is barely studied. Jitter is defined to be any timing movement with reference to an ideal signal. An investigation of the effect of jitter on range imaging is important because timing errors potentially could cause errors in measuring phase, thus in range. The purpose of this research is to investigate the effect of jitter on range measurement in AMCW ToF range imaging. This is achieved by three main contributions: a development of a common algorithm for measurement of the jitter present in signals from depth cameras, secondly the proposal of a cost effective alternative method to measure jitter by using a software defined radio receiver, and finally an analysis of the influence of jitter on range measurement. Among the three contributions of this thesis, first, an algorithm for jitter extraction of a signal without access to a reference clock signal is proposed. The proposed algorithm is based upon Fourier analysis with signal processing techniques and it can be used for real time jitter extraction on a modulated signal with any kind of shape (sinusoidal, triangular, rectangular). The method is used to measure the amount of jitter in the light signals of two AMCW ToF range imaging cameras, namely, MESA Imaging SwissRanger 4000 and SoftKinetic DepthSense 325. Periodic and random jitter were found to be present in the light sources of both cameras with the MESA camera notably worse with random jitter of (159.6 +/- 0.1) ps RMS in amplitude. Next, in a novel approach, an inexpensive software defined radio (SDR) USB dongle is used with the proposed algorithm to extract the jitter in the light signal of the above two ToF cameras. This is a cost effective alternative to the expensive real-time medium speed digital oscilloscope. However, it is shown that this method has some significant limitations, (1) it can measure the jitter only up to half of the intermediate-frequency obtained from the down shift of the amplified radio frequency with the local oscillator which is less than the Nyquist frequency of the dongle and (2) if the number of samples per cycle captured from this dongle is not sufficient then the jitter extraction does not succeed since the signal is not properly (smoothly) represented. Finally, the periodic and random jitter influence on range measurements made with AMCW range imaging cameras are studied. An analytical model for the periodic jitter on the range measurements under the heterodyne and homodyne operations in AMCW ToF range imaging cameras is obtained in the frequency domain. The analytical model is tested through simulated data with various parameters in the system. The product of angular modulation frequency of the camera and the amplitude of the periodic jitter is a characteristic parameter for the phase error due to the presence of periodic jitter. We found that for currently available AMCW cameras (modulation frequency less than 100 MHz), neither periodic nor random jitter has a measurable effect on range measurement. But with modulation frequency increases and integration period decreases likely in the near future, periodic jitter may have a measurable detection affect on ranging. The influence of random jitter is also investigated by obtaining an analytical model based on stochastic calculus by using fundamental statistics and Fourier analysis. It is assumed that the random jitter follows the Gaussian distribution. Monte Carlo simulation is performed on the model obtained for a 1 ms integration period. We found increasing the modulation frequency above approximately 400 MHz with random jitter of 140 ps has a measurable affect on ranging.
Type of thesis
Anthonys, G. (2021). Investigation of jitter on full-field amplitude modulated continuous wave time-of-flight range imaging cameras (Thesis, Doctor of Philosophy (PhD)). The University of Waikato, Hamilton, New Zealand. Retrieved from https://hdl.handle.net/10289/14129
The University of Waikato
All items in Research Commons are provided for private study and research purposes and are protected by copyright with all rights reserved unless otherwise indicated.