Low-cost Global Navigation Satellite System (GNSS) receivers and monocular cameras are widely used in daily activities. The complementary nature of these two devices is ideal for outdoor navigation. In this paper, we investigate the integration of GNSS and monocular camera measurements in a simultaneous localization and mapping system. The proposed system first aligns the coordinates between two sensors. Subsequently, the measurements are fused by an optimization-based scheme. Our system can function in real-time and obtain the absolute position, scale, and attitude of the vehicle. It achieves a high accuracy without a preset map and also has the capability to work with a preset map. The system can easily be extended to create other forms of maps or for other types of cameras. Experimental results on a popular public dataset are presented to validate the performance of the proposed system.

The calibration and correction of backplane aberration of a liquid-crystal spatial light modulator (LCSLM) are important for its proper functioning. To simplify the calibration procedurally, we study a random-illumination-phase-retrieval-based method. Our method improves the convergence of the phase-retrieval-based calibration and reduces the calibration complexity. However, the cross talk of the LCSLM deteriorates the calibration performance. We determined the relationship between the probability density function of the random phases and the light-intensity pattern and proposed an algorithm to compensate for the cross talk. We conducted a series of simulations to test the performance of the above-mentioned algorithms. The results show that our algorithms are effective and outperform other methods.

The description of local surface features is a critical step in surface matching and object recognition. We present a descriptor for three-dimensional shapes based on the bispectrum of spherical harmonics (BSH). First, points in a support region of a feature point are used to construct a local reference frame, and a histogram is formed by accumulating the points falling within each bin in the support region. Second, spherical harmonic coefficients of the histogram and its bispectrum are calculated. Finally, the feature descriptor is obtained via principal component analysis. We tested our BSH descriptor on public datasets and compared its performance with that of several existing methods. The results of our experiments show that the proposed descriptor outperforms other methods under various noise levels and mesh resolutions.

Using phase-only liquid crystal spatial light modulator (LC-SLM) to generate random illumination is always used in phase-retrieval and single pixel imaging as it have high energy efficiency. While the methods were only used to generate monochromatic illumination. This paper aims to use LC-SLM generating the multi-spectral random illumination. The method requires only one SLM; multiple laser beams are simultaneously used to illuminate the entire SLM. Random voltages designed with a probability density function (PDF) are applied to modulate the random phases in the multi-spectral emitted light beam, leading to random amplitude spots in the far field. By optimize the PDF, the random illumination spot can be close to uniform distributed with high efficiency. The average intensities of the random light spots are almost uniformly distributed within the range of the first diffraction order. And the zero-order beam is eliminated. The proposed method has a high energy efficiency and refresh rate and does not require the SLM to have an excessive modulation range. This paper provides a method to generate multi-spectral random illumination using a LC-SLM, which can be used in phase-retrieval and single pixel imaging.