Accurate forest inventory is critical for sustainable management, ecological assessment, and biomass estimation. Combining near-proximal and proximal Light Detection and Ranging (LiDAR) data produces point clouds that fully capture the forest structure. This study presents a unified framework for processing multi-resolution LiDAR data to extract key forest attributes, including tree location, height, and diameter at breast height (DBH). The proposed methodology integrates LiDAR data captured by Unmanned Aerial Vehicles (UAVs) and BackPack systems for canopy structure delineation and fine-scale understory mapping, enhancing the accuracy of tree segmentation and biometrics estimation. A multi-stage processing pipeline is developed, incorporating adaptive ground removal, intensity/geometry-based filtering for woody part separation, and layered density-based spatial clustering of applications with noise (DBSCAN) to mitigate over-segmentation. Additionally, an image-LiDAR linking strategy is introduced as a precursor for tree species identification by associating segmented trees with UAV and BackPack imagery. The proposed approach is evaluated in a plantation, demonstrating an F1-score of 100% for tree detection and a 3.1 cm root mean square error (RMSE) for DBH estimation. The results highlight the reliability of the proposed framework for accurately detecting trees and estimating their DBH. Furthermore, by combining geometric information from LiDAR with the rich semantic information in captured imagery, the proposed image-LiDAR linking strategy shows its potential for tree species identification. The effectiveness of multi-source LiDAR integration for forest inventory applications, offers a scalable solution for large-scale forest monitoring. Future work will focus on improving tree segmentation in complex forest environments and leveraging machine learning models for automated species classification.