Principles And Key Techniques Of Multispectral LiDAR For Earth Observation

LiDAR has been well developed since 60th 20 century, its types and functions have been continuously improved, and has comprehensive applications. As one of its greate applications, LiDAR has become a well-established technique for the remote sensing of surface topography. However, due to its single wavelength detection, the use of intensity in the retrieval of target characteristics is limited. Most work in objects detection especially vegetation canopies by LiDAR systems has not investigated the potential of laser backscatter reflectance.In order to investigate the great potential of LiDAR applications in objects detection, a multi-designed canopy LiDAR system, which can transmit multi-wavelength lasers within and adjacent to the most sensitive wavelengths of vegetation spectral characteristic regions, was designed. It can join the advantageous capabilities of LiDAR detection with passive multispectral sensors, providing a large number of laser backscatter features for vegetation classification and biochemical recognition. Similar to trend of passive sensors which developed from Panchromatic—Multispectral—Hyperspectral, LiDAR technique for earth observation would evaluate from single wavelength to Multiwavelength. Our research would promote this developing process of LiDAR. This thesis concentrates on the principle of a new multiwavelength LiDAR system and its key techniques. The studies are related to the following four topics:(1) According to the characteristics of LiDAR, a back-reflection model of laser detection based on objects reflection characteristics was developed. Meanwhile, considering the function, complexity and stability of the whole system, the standard of wavelength selection which is "to win the best ability of objects characteristic detection with the least laser wavelengths" was proposed. Based on the hyperspectral reflectance measurements of representative objects, hyperspectral analysis was made to select the most sensitive wavelength to objects, then evaluate the performance of each selected wavelength, and finally confirm the optimal wavelength combination. Wavelength selection research is the foundation of multispectral LiDAR for objects detection and recognition, also for the laser backscatter data processing.(2) Analyse and summarize the theory and key technique, complete the unit technique design and development of system. The unit technique was devided into three parts:multi-wavelength laser transmitting system; signal receive system and scan & range system. The laser transmitting system research focues on the special wavelength laser design and the optical design in order to make the multiwavelength lasers synthesize to a beam; in contrast to the transmitting system, the backscattered light is collected by the Schmidt-Cassegrain telescope and split by dichroic filters into four separate receiving paths, and then researth on the weak signal receive and detection; scan & range system studies the design and deveop of ranging, scanning, and the control system.(3) Based on the unit technique design and development, the whole demonstration system design of multispectral LiDAR system was made, including signal receiving control, synchronous scanning, system control, and signal collection and process. By study and achieve these functions, real time control and process from laser transmitting to detection data process for multispectral LiDAR detection have been made.(4) According to the studies above, a new demonstration system of Multi-Wavelength LiDAR for earth observation remote sense was made, and its measuring performance has been investigated. In our designed ground demonstration experiments, the Multi-Wavelength LiDAR system was moveable to scan and detect objects with multi-wavelength lasers. The results showed that this Multi-Wavelength LiDAR system has greater capability to discriminate living vegetations from non-living sources, even to discriminate living vegetations with various growing statues, in addtion to the capability of spatial detection. The demonstration experiment manifests the feasibility of this new Multi-Wavelength LiDAR.