Investigation of TLS Intensity Data and Distance Measurement Errors from Target Specular Reflections

Kai Tan, Weiguo Zhang, Fang Shen, Xiaojun Cheng
2018 Remote Sensing  
Terrestrial laser scanners (TLSs) can provide accurate and high-resolution data by measuring the distances (ranges) between the scanned points and the scanner center using time-of-flight or phase-shift-based methods. Distance measurement accuracy is of vital importance in TLSs and mainly influenced by instrument mechanism, atmospheric conditions, scanning geometry, and target surface properties. In general, existing commercial TLSs can achieve millimeter precision. However, significant errors
more » ... entimeter and even decimeter levels) beyond the instruments' nominal accuracy exist in distance observations for targets with highly reflective surfaces whose specular reflections are dominant because these reflections can increase the backscattered laser signal power considerably and cause further disorder in the echo detection and recognition by TLS photodetectors. Apart from distance, the intensity value derived from the backscattered signal and influenced by the same factors as that of the distance measurement errors is recorded by TLSs. A certain link exists between the two instrumental observations. In this study, the anomalous distance measurement errors caused by target specular reflections are explored. The different planar reflective targets scanned by a Faro Focus 3D 120 terrestrial scanner are used to experimentally investigate the relationship between the original intensity values and the distance measurement errors. Results imply that the distance measurement errors caused by specular reflections are not as erratic as they ostensibly seem. On the contrary, distance measurement errors are strongly related to the original intensity values. A polynomial can be established to empirically model the relationship between the original intensity data and the distance measurement errors. With use of the original intensity to compensate for the measured distance observations, the point cloud data accuracy can be improved by approximately 55.52%. instrument. Consequently, distance measurement unit is a core component of TLSs. The investigation of distance measurement error sources and solutions is of vital importance to the improvement of TLS data accuracy and quality and has attracted widespread attention from manufacturers, users, and researchers in the past two decades [2] [3] [4] [5] [6] [7] [8] [9] . Theoretically, the distance measurement accuracy of TLS is influenced by instrument mechanism, atmospheric conditions, scanning geometry (i.e., distance and incidence angle), and target surface properties [2] . Only the acquisition configuration and the target surface properties must be considered for the compensation of distance measurement errors because the atmospheric conditions near the surface of the Earth are relatively stable and the instrument mechanism is usually unchanged during one campaign. The effects of distance and incidence angle on distance measurement accuracy have been extensively studied, whereas research about distance measurement errors caused by target properties is relatively rare. The reflectorless nature of TLS rangefinders increases the possibility of distance measurement errors due to the pulse attenuation by the reflecting surfaces [1, 6] . The distance errors caused by the properties of rough and dull surfaces usually measure a few millimeters. However, these errors can significantly increase to centimeter and even decimeter levels [8,10] for smooth surfaces because diffuse and specular reflections exist in all-natural surfaces and the type of reflection affects the direction and strength of backscattered light. For targets with a relatively smooth or glossy surface (e.g., water [11] [12] [13] , fresh ice [14-16], foliage/leaves [17,18], metal [19], porcelain, and plastic [20, 21] ), specular reflections are dominant over diffuse reflections. For a generally smooth surface of multiple facets with different orientations, specular reflections are distributed in the surroundings of the reflection angle direction [22] . The target specular reflections can significantly change the amplitude, width, or shape of the backscattered signal, especially in the case of small incidence angles that are close to zero due to the coincidence of the emitter and receiver [19, 23] . A deformed signal influences the scanner system's determination of the arrival time of the backscattered echo or the phase comparison between the emitted and received signals. Thus, target specular reflections can lead to errors in distance observations and even in the saturation effect [10, [19] [20] [21] 23, 24] of photodetectors due to the extremely high amplitude of returned pulses. As stated above, distance measurement errors for a certain TLS are influenced by the target surface properties, range, and incidence angle. A direct solution is to consider these factors individually. While this is feasible for the distance, the surface properties for individual points are usually unknown [1]. The incidence angles for single points can be computed using neighborhood points, but they are vulnerable to local point density and noise [1]. Therefore, considering distance, incidence angle, and target surface properties separately is not practical for modeling distance errors because most of them are unavailable or unreliable. In addition to discrete topography measurements, nearly all current TLS instruments simultaneously measure the power of the backscattered laser signal of each point and record it as an intensity value [25] [26] [27] . Backscattered optical power is internally converted to voltage, amplified in the system, and finally transformed into a digital number, i.e., a scaled integer value called "intensity" [28] . Intensity, which is insensitive to ambient light and shadowing, is initially used to improve point cloud separability [29, 30] . Apart from visualization purposes, various object-based studies can adopt intensity data as a major or complementary data source [31] [32] [33] [34] [35] . Intensity and distance are two types of data in TLS, one for physical and the other for geometry. No connection seems to exist between these data. However, the intensity detected by a TLS system is also mainly affected by target scattering characteristics, range, and incidence angle [36] [37] [38] [39] [40] , which are nearly the same as those of the distance measurement errors [1] . Additionally, when the backscattered signal reaches the TLS, the receiver distance measurement unit calculates the distance based on the features of the backscattered signal. Synchronously, the TLS system obtains the intensity value according to the amplitude of the backscattered signal. Therefore, both distance and intensity are derived from the features of the backscattered signal. Thus, a certain link exists between the intensity value and the distance error, and this connection is the theoretical basis for this study. Instead of investigating the
doi:10.3390/rs10071077 fatcat:gydqfpkq3rax7azvb35exnjada