Due to the increasing use of the Terrestrial LiDAR Scanning (TLS also called T-LiDAR) technology in the forestry domain, many researchers and forest management organizations have developed several algorithms for the automatic measurement of forest inventory attributes. However, to the best of our knowledge not much has been done regarding single tree species recognition based on T-LiDAR data despite its importance for the assessment of the forestry resource. In this paper, we propose to put the light on the few works reported in the literature. The various algorithms presented in this paper uses the bark texture criteria and can be categorized into three families of approaches: those how combine T-LiDAR technology and photogrammetry, those based on depth images generated from T-LiDAR data and those based on raw 3D point cloud.
TopTree Species Identification From T-Lidar Data: An Overview
Remote sensing by laser or LiDAR (Light Detection And Ranging) has continuously evolved over the past decades. In the late 1990s, it has grown considerably in the industry due to associated development of relevant computer technologies and its ability to process and save the huge amount of information obtained. LiDAR (Wulder et al., 2008) technology has become increasingly important in many areas such as the study of the environment and air pollution, detection of contaminants or bacteria, 3D scanning of an archaeological site, quality control, etc. This technology has already demonstrated, through numerous studies, its potential for the characterization of forest resources, in particular with aerial LiDAR. This technology has provided improved 3D forest information and within the same forest stand. In this context, the research focused on the exploitation of LiDAR through different techniques to map, characterize and quantify the forest resource.
Remote sensing methods for airborne LiDAR (A-LiDAR or ALS) have the advantage of providing population variables according to clumps or stands. They provide comprehensive assessment of the resource over large scales and a population zoning; but they are often not sufficiently accurate to estimate dendrometric variables such as stem density and fine structure of the population, in the case of heterogeneous populations in particular. In order to accurately identify these dendrometric variables, it is necessary to conduct inventory plots in the field using a terrestrial laser scanner (T-LiDAR or TLS). Indeed, good accuracy and high-resolution T-LiDAR (accuracy of ± 2 mm between 10 and 25 m, and an angular step of 0.009° (for the Faro Photon 120, for example) allow for highly precise inventories of forest plots (location, diameter and volume of each stem). Earlier work has helped develop algorithms to isolate tree trunks of a scanned scene, estimate the diameter trunks at the height of 1.30 m or also called diameter at breast height (DBH) and tree height. The paper of Dassot, Constant & Fournier (2011) can complete this paper to ensure that the scope of the content of this chapter is clear and well-presented. Many research teams and organizations of forest management have recently started to automate forest inventories due to such developments (Dassot, Constant & Fournier, 2011). Among them, The National Forests Office (French: Office national des forêts), or ONF, the public establishment of the National Government of France charged with the management of national forests, was one of the first to set up a processing platform of T-LiDAR data, called Computree (Othmani, Piboule, Krebs, Stolz & Lew Yan Voon, 2011). Other point cloud processing software has been specifically dedicated to forest measurements. This is the case of 
software developed by the TreeMetrics Company. A comparison of the uses and the limitations of some generic and specific software found in the literature for forest structure assessment are presented in the paper of Dassot, Constant and Fournier (2011).
The species of trees is one of the primordial parameters of the forest inventory especially for the case of heterogeneous mixed forests. It is very important to assign to each detected tree the correct species. Most studies have used data A-LiDAR and species of trees are determined across the forest. However, at the scale of a plot, this data must be measured manually in the field, which complicates the measurement phase and causes problems with processing and corresponding stems detected from T-LiDAR data. For now, despite its importance, this area has been little explored in the literature. Many reasons can explain the lack of literature and research works in this area:
- 1.
The complexity of the processing of the data (a noisy, randomly distributed, disordered and unstructured point cloud that must be reconstructed by dedicated programs to provide information).
- 2.
The research teams and organizations started recently the automation of forest inventories and they are more focused on the extraction of tree trunk, the estimation of the DBH and the estimation of tree height.
- 3.
The confidentiality of 3D point cloud of forest plots and their non-open source availability.