Timber Defect Identification: Enhanced Classification with Residual Networks

Within the furniture and forestry industries, there is a need for high value products to be developed for international markets utilising Australian hardwood timbers. This investigation has addressed this requirement, with a focus upon a particular timber species - Spotted Gum (corymbia citriodora subsp. variegata), a diversely eminent species of Queensland hardwood timber. The investigation was initiated by collaborating parties within the Co-operative Research Center for Wood Innovations (CRCWI) particularly the Department of Primary Industries – Forestry Research (DPI&F) in Queensland. It was decided by the DPI&F that an industrial design contribution, through the instigation of a design research led investigation would be a beneficial avenue for addressing prevalent issues in the forestry and furniture industries. Background research processes undertaken in both the forestry and furniture industries in a geographically specific area of Queensland were vital in establishing immediate investigation parameters. Following the establishment of these parameters and their accepted relevance to broader national industry concerns, the consequent development of an appropriate research method in this investigation was undertaken. The method generated needed to address two major issues. First, to address technical problems in the application of Spotted Gum timber to the production fine furniture, surpassing various initiatives to resolve these issues in the past, secondly, to address a lack of market knowledge, with regard to product design parameters for export markets within the participating Queensland furniture manufacturing industry. The method employed seeks to establish the degree of cultural difference that must be accounted for by manufacturers in developing products specifically for export market integration. This theory was tested by the development of two experimental indoor dining chairs, that were designed and prototyped, recognising to the best degree possible the exceptional technical requirements of Spotted Gum timber. Each of the two chairs were developed to the requirements of pre-determined market and user oriented needs of a separate case study destination, determined through qualitative and quantatative information generation. The specific niche market design parameters applied to design development, created a precursory theory that the products would have a greater degree of success in market integration if they were designed for specific niche market parameters.Each of the chairs was then exhibited in an appropriate market arena for the destination for which it was designed. A series of questions seeking preferences for each of the chairs, and the reasons for those preferences were solicited from those attending both of the exhibitions. The testing process resulted in a conclusion that there is little cultural difference that must be accounting for in approaching design development for the two international markets identified tested as case studies. The initial chair designs, developed and used in the testing role within the investigation, were according re-designed given the findings of the market testing process.

The decreasing availability of economic hardwood timber species throughout tropical countries has necessitated the need to adopt plantation grown exotic timber species into the mainstream of raw materials in the furniture and wood manufacturing industries. However, published research on the material properties of most of these exotic timber species grown in the Savannah Ecological Zone of Ghana is limited. The wood properties of these species when known could result in their optimal utilization and broad acceptance in the wood industries as an alternative for the extinct tropical timber species. This study determines the anatomical properties of plantation grown Anogeissus leiocarpa and Eucalyptus camaldulensis trees grown in the Savannah Ecological Zone of Ghana. Wood discs were obtained from three stem heights (butt, mid and top) from which cubes measuring 40 mm x 20 mm x 20 mm were produced for the microtome sections and macerated tissues, viewed under a microscope with Motic Image Plus software. Descriptions of wood anatomy followed IAWA's microscopic characteristics. It was observed that in the sapwood, there were more vessels than in heartwood, though the vessels were smaller in diameters, suggesting resistance to sap conduction which gives wood high aesthetic value and dimensional stability during drying. The species vessel diameter fell within the medium category and large vessels, demonstrating that wood will possess a greater mechanical strength since it will be denser, consequently possessing significant mechanical strength that can be compared favorably to other commercial timber species. This might be a major breakthrough in the production of quality and durable furniture as well as other wood products by using the exotic plantation grown timber species.

Biaxially oriented polyester film (BOPET) defect is an important factor affecting the quality of the film. In view of identification of defects in the conventional film production process, this mathod resulted in the identification of defects inaccurate. and low labor efficiency and machine vision recognition on identification of specific defect. This paper presents a LVQ neural network-based BOPET film of defects detection and identification methods. In this algorithm, the film images were processed and the outlines of the membrane defects were obtained. Through extracting the aspect ration, circularity, complexity and elongation , projection histogram central moment and so on, the characteristic values of membrane defects, which from the image of film images after image processing, and then input to the defects recognition system based on LVQ neural network that had been trained, in order to achieve the film defects identification, classification and localization. Through the study of features of the defects in BOPET and extracted some quantities as character input of the LVQ neural network, then input some characteristic values as training value into the LVQ neural network to achieve the learning and prediction purpose, and the LVQ neural network was designed. The experiments show that, the proposed method can meet the requirements analysis of air defects in transparent film.

The demand for forest-related goods and services has continued to grow in Ethiopia due to the increasing population, urbanization, and the boom of the construction sector. In the past, the demand was covered by indigenous timber species, which now have declined due to unwise exploitations. To narrow down the gap between supply and demand an alternative solution is promoting and utilizing fast-growing species such as Eucalyptus wood. To effectively utilize this species identifying its special characteristics, level of acceptance in wood industries, and utilization practices are very important. To achieve the goals of this review, a thorough search was carried out across multiple scholarly databases. Then, pertinent research, papers, and articles about the use of Eucalyptus in the wood industries were downloaded and reviewed. The study shows that Eucalyptus is widely cultivated in various regions of Ethiopia and plays a significant role in the socio-economic dimensions of many Ethiopians, providing livelihood opportunities, environmental services, and supporting sustainable development. Although it is being used as an alternative to traditional timber species in many places in the world, it is not being widely used in Ethiopia's wood industry for furniture manufacturing. This is because producers are unaware of the special properties of this species; its felling technique, sawing method, and drying behavior which demands proper attention and care. To ensure the sustainable use of Eucalyptus resources, better tree management techniques, innovative wood processing techniques, market development, and strengthening the capacity of relevant stakeholders are imperative.

At present, off-cut wood pieces are dumped by sawmills as they are considered to be wastes in the wood industry. Inadequate length of sawn timber material is also reported to be a limiting factor for fully utilization of timbers. Finger joint, a method which connects two small pieces of timber together is identified as a sound technique to minimize the wastage. As there is no classification system applicable for finger joint timber in Sri Lanka, the present study focused on developing a classification system for selected 32 timber species based on their strength properties. The relationship between the strength properties and density of selected timber species was also investigated. Strength properties of finger jointed timber species were evaluated by three-point bending and compression tests according to BS 373:1957 using Universal Testing Machine (UTM-100).Factors were identified through an analysis to determine the strength index for the selected samples. The strength index values were grouped into five strength classes as very low, low, medium, high and very high. Cluster analysis was used in grouping the species with similar strength properties. Regression analysis was performed to identify the strength index of compression parallel to grain, compression perpendicular to grain, modulus of elasticity and modulus of rupture varies on density. A significant correlation (p=0.05) between the strength index and timber density was observed. Keywords: Finger joint, timber classification, universal testing machine, strengthindex

Laser tweezers Raman spectroscopy combined with deep learning to classify marine bacteria

Researchers encounter substantial challenges in medical image classification, mainly due to limited image resolution and low signal-to-noise ratios. This situation makes it difficult for deep learning algorithms to identify abnormal regions based solely on image content accurately. This paper proposes ResGDANet (Residual Group Dual-Channel Attention Network), an enhanced architecture that builds upon ResGANet by incorporating two novel modules: a Dual-Channel Attention Fusion (DCAF) module and a Retention-Memory Transformer (RMT) module. The DCAF module utilizes a dual-path architecture that integrates global average pooling and max pooling operations, effectively enhancing local feature representation through the fusion of channel-wise and spatial attention mechanisms. The RMT module enhances rotation-invariant feature extraction by integrating the retention mechanism from Retentive Networks and the global context modeling capabilities of Vision Transformers. Extensive experiments on the COVID19-CT and ISIC2018 datasets demonstrate the superiority of ResGDANet, achieving classification accuracies of 83.74% and 81.73% respectively, outperforming state-of-the-art models including ResGANet, GvT, and SENet. Ablation studies and visualization analyses further validate the efficacy of the proposed attention module, showing notable enhancements in feature representation capability and classification accuracy. By introducing a more robust and precise classification framework, this research contributes importantly to the progress in medical image analysis.

The accuracy of the classification of user intentions is essential for motor imagery (MI)-based brain-computer interface (BCI). Effective and appropriate training for users could help us produce the high reliability of mind decision making related with MI tasks. In this study, we aimed to investigate the effects of visual guidance on the classification performance of MI-based BCI. In this study, leveraging both the single-subject and the multi-subject BCI paradigms, we train and classify MI tasks with three different scenarios in a 3D virtual environment, including non-object-directed scenario, static-object-directed scenario, and dynamic object-directed scenario. Subjects are required to imagine left-hand or right-hand movement with the visual guidance. We demonstrate that the classification performances of left-hand and right-hand MI task have differences on these three scenarios, and confirm that both static-object-directed and dynamic object-directed scenarios could provide better classification accuracy than the non-object-directed case. We further indicate that both static-object-directed and dynamic object-directed scenarios could shorten the response time as well as be suitable applied in the case of small training data. In addition, experiment results demonstrate that the multi-subject BCI paradigm could improve the classification performance comparing with the single-subject paradigm. These results suggest that it is possible to improve the classification performance with the appropriate visual guidance and better BCI paradigm. We believe that our findings would have the potential for improving classification performance of MI-based BCI and being applied in the practical applications.

In autonomous driving vehicles, the emergency braking system uses lidar or radar sensors to recognize the surrounding environment and prevent accidents. The conventional classifiers based on radar data using deep learning are single input structures using range–Doppler maps or micro-Doppler. Deep learning with a single input structure has limitations in improving classification performance. In this paper, we propose a multi-input classifier based on convolutional neural network (CNN) to reduce the amount of computation and improve the classification performance using the frequency modulated continuous wave (FMCW) radar. The proposed multi-input deep learning structure is a CNN-based structure using a distance Doppler map and a point cloud map as multiple inputs. The classification accuracy with the range–Doppler map or the point cloud map is 85% and 92%, respectively. It has been improved to 96% with both maps.

L 2 regularization for weights in neural networks is widely used as a standard training trick. In addition to weights, the use of batch normalization involves an additional trainable parameter γ, which acts as a scaling factor. However, L 2 regularization for γ remains an undiscussed mystery and is applied in different ways depending on the library and practitioner. In this article, we study whether L 2 regularization for γ is valid. To explore this issue, we consider two approaches: (1) variance control to make the residual network behave like an identity mapping and (2) stable optimization through the improvement of effective learning rate. Through two analyses, we specify the desirable and undesirable γ to apply L 2 regularization and propose four guidelines for managing them. In several experiments, we observed that applying L 2 regularization to applicable γ increased 1% to 4% classification accuracy, whereas applying L 2 regularization to inapplicable γ decreased 1% to 3% classification accuracy, which is consistent with our four guidelines. Our proposed guidelines were further validated through various tasks and architectures, including variants of residual networks and transformers.

Introduction: In order to assess how well conventional and hybrid pitch detection techniques perform in speech processing applications, a comparative analysis of the two types of methods is conducted. Method: A proposed hybrid approach, Proposed PEF+CEP, is examined alongside five traditional algorithms, namely Normalized Correlation Function (NCF), Pitch Estimation Filter (PEF), Log- Harmonic Summation (LHS), Summation of Residual Harmonics (SRH) and Cepstrum Pitch Determination (CEP). The effectiveness is evaluated using performance metrics like accuracy, specificity, sensitivity, and Gross Pitch Error (GPE). Results and Discussion: Our findings show that the accuracy and specificity of the traditional methods are impressive; the accuracy and sensitivity of the suggested hybrid method surpass their performance, with an astounding 98.8% accuracy and 99.2% sensitivity. Conclusion: Furthermore, the Proposed PEF+CEP method is a promising solution for accurate and dependable pitch detection in speech processing applications because it strikes a strong balance between computational efficiency and robustness. These results open up new avenues for research in the field of speech processing and demonstrate the potential of hybrid approaches.

The present study shows new findings that reveal the high association between emotional arousal and neuro-functional brain connectivity measures. For this purpose, contrasting discrete emotional states (happiness vs sadness, amusement vs disgust, calmness vs excitement, calmness vs anger, fear vs anger) are classified by using Support Vector Machines (SVMs) driven by Graph Theoretical segregation (clustering coefficients, transitivity, modularity) and integration (global efficiency, local efficiency) measures of the brain network. Emotional EEG data mediated by short duration video film clips is downloaded from publicly available database called DREAMER. Pearson Correlation (PC) and Spearman Correlation have been examined to estimate statistical dependencies between relatively shorter (6sec) and longer (12sec) non-overlapped EEG segments across the cortex. Then the corresponding brain connectivity encoded as a graph is transformed into binary numbers with respect to two different thresholds (60%max and mean). Statistical differences between contrasting emotions are obtained by using both one-way Anova tests and step-wise logistic regression modelling in accordance with variables (dependency estimation, segment length, threshold, network measure). Combined integration measures provided the highest classification accuracies (CAs) (75.00%80.65%) when PC is applied to longer segments in accordance with particular threshold as the mean. The segregation measures also provided useful CAs (74.13%80.00%), while the combination of both measures did not. The results reveal that discrete emotional states are characterized by balanced network measures even if both segregation and integration measures vary depending on arousal scores of audio-visual stimuli due to neurotransmitter release during video watching.

ResNet and its application to medical image processing: Research progress and challenges

Off-cut wood is currently considered as waste thus is dumped by saw mills as they find no means of utilising them. Sawn timber material of furniture factories and short length of sawn timbers are also considered to be wastes in the timber industry. Finger jointing technique which interlock the end joints formed by machining a number of similar tapered symmetrical fingers are recognised in this regards as effective and sustainable means of utilisation of timber wastes. The present study was undertaken to assess the strength grade of finger jointed timber based on BS 5268-2:2002. Seven timber species which are commonly used in Sri Lanka were employed for the assessment with and without finger joints. Values of Modulus of Rupture (MOR), Modulus of Elasticity (MOE), Compression parallel to grain and Compression perpendicular to grain strength of the specimens were measured. Finger jointing was performed with constant geometry (finger length 19 mm, tip width 1 mm and finger pitch 4 mm) using polyvinyl acetate (PVA, P-SWR) adhesive at normal exposure conditions. The strength properties were evaluated by Universal Testing Machine (UTM) according to BS 373:1957. Strength classes relevant to the grade stresses were not significant for finger jointed and clear specimens of Satin, Mahogany, Jack and Grandis. Both clear and finger jointed timber specimens obtained D40 for Satin and Teak, D30 for Jack, Mahogany and Grandis. Teak shows properties similar to both D35 and D40 when used as finger jointed timber. Kumbuk was shown to change from D40 to D30 while using as finger jointed timber. Finger jointed Pine showed stress grade of C22, C24 and C27.

Forestry in Vietnam: The potential role for native timber species