Application of image quality assessment module to motion-blurred wood images for wood species identification system

Heshalini Rajagopal,Anis Salwa Mohd Khairuddin,Norrima Mokhtar,Azlin Ahmad,Rubiyah Yusof
DOI: https://doi.org/10.1007/s00226-019-01110-2
2019-06-21
Wood Science and Technology
Abstract:Despite tighter conservation regulations, demand for timber products has continued to increase due to growing population. Normally, experts identify the wood species based on the pattern of the wood surface texture. However, manual inspection on wood texture is tedious, time-consuming, impractical and cost-ineffective for a human to analyze a large number of timber species. Therefore, a reliable automatic wood recognition system is needed in order to classify the wood species efficiently. The proposed system includes image acquisition, image quality assessment module (IQA), image deblurring, feature extraction and classification. In this research, the wood images are motion-blurred due to imperfections in the imaging and capturing process. Hence, an IQA module is proposed to monitor the quality of images before proceeding to the next stage which is the feature extraction process. The IQA module will determine whether the image has to undergo the image deblurring process based on the image quality value. If the image is of low quality based on the image quality value obtained, then the image will be deblurred before the feature extraction procedure. A reliable motion deblurring technique, which is based on Lucy–Richardson algorithm, is employed to enhance the motion-blurred images before proceeding to the next stage, which is the feature extraction process. Then, a statistical feature extraction technique is proposed to extract 24 features from each wood image. Finally, a support vector machine is used to classify the 20 tropical wood species.
materials science, paper & wood,forestry
What problem does this paper attempt to address?