Automated Sweep Measurement Project
Forest Growers Research and Interpine have joined to help harvesting operations to face some big challenges now and in the future. These include labour shortages, rising costs reducing competitive advantages, increasing barriers to long term sustainability, and poor profitability of small forest holdings.
The project will deliver an automated and objective system for assessing sweep during log processing. Detection of log sweep has relied on manual measurement which is very inefficient, or the experience of machine operators. This will replace the manual and subjective assessment of sweep currently made at high speeds by the log processor operator. It should result in improved sweep measurement accuracy and improved value recovery for both the buyers and suppliers of logs.
This project shows how computer vision and deep learning can build a model that automatically identifies and calculates log sweep. The results of the project will be delivered into two main phases.
The first is to make a pre-trained log recognition model. The second is the extraction and analysis of data based on the trained recognition model, using pixel point coordinates as the primary tool for calculating the sweep ratio of the detected logs in real-time.
The project aims to develop and test a prototype system. It is not intended that the system developed will be at a commercial ready stage by the end of the project. The prototype system will require further customisation and development to integrate it with the operating and control systems of the many log processing brands and models currently used in New Zealand.