Linear window correlation: new image processing based approach to strain distribution analysis of wood.

An in-depth understanding of the response of wood under compression/densification calls for characterization of strain distribution over the wood. Recently, an image processing technique has been used for strain evaluation on wood, as the traditional devices are not able to measure irregularly distributed strain field. Unfortunately, the classic image correlation algorithm is rather time consuming and could take up to days to compute the final result. In this paper, an improved image correlation algorithm with linearly changed searching region during the correlation seeking is presented. The processing speed is increased by over 1800 times while keeping the high accuracy of the displacement measurement. In addition, full feasibility tests were conducted on balsam fir ( (L.) Mill.) and eastern white pine ( L.) samples. The results demonstrate the high robustness, efficiency, and accuracy of the proposed method. [ABSTRACT FROM AUTHOR]