Weak Texture Research Articles

Mono‐MVS: textureless‐aware multi‐view stereo assisted by monocular prediction

AbstractThe learning‐based multi‐view stereo (MVS) methods have made remarkable progress in recent years. However, these methods exhibit limited robustness when faced with occlusion, weak or repetitive texture regions in the image. These factors often lead to holes in the final point cloud model due to excessive pixel‐matching errors. To address these challenges, we propose a novel MVS network assisted by monocular prediction for 3D reconstruction. Our approach combines the strengths of both monocular and multi‐view branches, leveraging the internal semantic information extracted from a single image through monocular prediction, along with the strict geometric relationships between multiple images. Moreover, we adopt a coarse‐to‐fine strategy to gradually reduce the number of assumed depth planes and minimise the interval between them as the resolution of the input images increases during the network iteration. This strategy can achieve a balance between the computational resource consumption and the effectiveness of the model. Experiments on the DTU, Tanks and Temples, and BlendedMVS datasets demonstrate that our method achieves outstanding results, particularly in textureless regions.

Effect of Solution Annealing Time on the Microstructure and Mechanical Properties of Selective-Laser-Melted 2205 Duplex Stainless Steel

The 2205 duplex stainless steel (DSS) produced by selective laser melting (SLM) exhibits high strength (1078.8 MPa) but poor plasticity (15.2%) owing to the high cooling rate during SLM, which inhibits the formation of austenite and creates a nearly entirely ferritic microstructure. The dual-phase nature can be restored through solution annealing, which enables well-matched strength and plasticity, but which has not been extensively studied. We investigate the effects of 5 min, 30 min, and 120 min of solution annealing at 1000 °C on the dual-phase ratio, grain size, texture strength, inclusions, grain boundary characteristics, and mechanical properties of SLM-manufactured 2205 DSS. After 30 min of solution annealing, the elongation increased to 32.2% owing to the restoration of the dual-phase structure, the reduction in dislocation density, the weakening of texture, and the decrease in grain size. Increasing solution annealing time also corresponded to a decrease in the ultimate tensile strength (from 831.7 to 787.5 MPa) and yield strength (from 610.3 to 507.8 MPa) due to grain coarsening and the gradual transformation of ferrite to austenite. Furthermore, the mechanism of the transformation from ferrite to austenite was proposed, and it was observed that the transformation of MnSiO3 to MnCrO4 provided nucleation sites for austenite.

Understanding the weakening of texture in multiaxial forged ultrafine-grained strong and ductile Mg-2Zn-2Gd alloy

ABSTRACT Texture is an important aspect that impacts the formability of magnesium (Mg) and its alloys. Recently, the unique characteristics of multiaxial forging of Mg-2Zn-2Gd alloy were discussed, where it was demonstrated that multiaxial forged processed alloy is characterized by high strength and high ductility combination and the texture is significantly weakened or almost texture-free. The objective of the study described here is to obtain some fundamental insights that explain the underlying reasons for the weakening of texture in the multiaxial forged Mg-2Zn-2Gd alloy.

Significant improvement of bending properties for WE43 magnesium alloys under the coupling effects of weak basal texture and high deformation temperature

Significant improvement of bending properties (approximately 64.7%) for WE43 magnesium alloy was realized by weakening the basal texture and raising the deformation temperature. The mechanisms for such improvement were investigated in detail by combining the experiment and finite element model. The results concluded that the coupling effects of weak basal texture and high deformation temperature was the key reason producing the above improvement of bending properties. Specially, the TD sample with weak basal texture was more conducive to activating basal slip than the RD sample, resulting in its higher strain hardenability and consequently more significant bending properties. Based on weak basal texture, raising the deformation temperature could further improve the bending properties. This improvement mechanism was attributed to the abundant activations of non-basal slip instead of basal slip, which could increase the independent slip systems and efficiently delay the development of the strong basal texture. Moreover, the occurrence of partial DRV and DRX relieved the stress concentrations by consuming the dislocations, further improving the bending properties. The activations of these deformation mechanisms matching the corresponding bending stress states caused the crystal rotation and the texture variety in the bending samples. Basal slip activation aligned the (0002) basal planes parallel to tensile stress in the outer region or perpendicular to compressive stress in the inner region. Additionally, prismatic slip activation induced crystal rotation around the c-axis, developing <10-10>//LD texture in the outer region but <10-10>⊥LD texture in the inner region.

Evolution of Residual Stress, Crystal Orientation, and Texture on Preheating Weld Treatment of Low Carbon Steel ASTM A572 Grade 42

Preheating is one type of heat treatment on the material prior to starting the welding process by increasing the temperature of the material. It is used to reduce the cooling rate during the welding processes to minimize the risk of residual stress and cracking during welding. Besides the benefits of welding, it also has a negative impact, especially on the integrity of the material, because the heating process at high temperatures will cause residual stress, which will affect the mechanical properties, chemical composition, and microstructure of the material, especially on Heat Affected Zone (HAZ). This study aims to analyze the effect of preheating at a temperature of 200 ºC on the welding for ASTM A572 Grade 42 steel with a thickness of 40mm using the SMAW (Shielded Metal-Arch welding) method. ASTM A572 Grade 42 was used in this study with carbon content of less than 0.25%. Based on the results of measurements and analysis using the XRD method, it was found that preheating resulted in reducing the residual stress on both weld metal and HAZ areas. Preheating treatment also moved the peak diffraction to the right side, which means the heat treatment affected compressive residual stress rather than tensile residual stress. For crystal orientation, area [110] has the highest peak diffraction and highest intensity. This area also was found with smaller size crystal size and higher dislocation and microstrain. While bigger crystal size with lower dislocation and microstrain were found in the area [200]. For texture, the highest density was found in the area [200], while the weaker texture was found in the areas [110] and [211]. The texture was influenced by plastic deformation due to atomic structure and its dislocation.

Cooperative achievement in grain refinement and texture weakening of Mg-Gd-Y-Sm-Zr alloy fabricated by asymmetric extrusion-shear process

Mg-Gd-Y-Sm-Zr alloy with fine grains and weak texture was fabricated by extrusion-shear (ES) process. The microstructure evolution mechanism of the alloy during asymmetric plastic deformation process was analyzed. At the forward extrusion stage, the initial coarse grains are refined significantly under the synergistic action of discontinuous dynamic recrystallization (DDRX), continuous dynamic recrystallization (CDRX), and deformation bands (DBs) induced dynamic recrystallization (DRX). The shear deformation during equal channel angular pressing (ECAP) can further refine the residual coarse grains after the previous extrusion, as well as weaken the basal texture of the alloy. The effect of extrusion ratio (λ = 9, 16, 25, 36) on the alloy after ES forming was studied. When λ is 25, the component has superior mechanical properties with yield tensile strength (YTS) of 216 MPa, ultimate tensile strength (UTS) of 303 MPa and elongation of 25.47%. The improvement of YTS after ES forming is mainly attributed to the grain boundary strengthening, while the excellent elongation is benefited from the weak bimodal texture. Abnormal grain growth will occur when λ is 36, leading to a decrease in the strength and plasticity of the alloy.

Effect of rotational-die ECAP parameters on microstructure and mechanical properties of Mg97Y2Zn alloys

In this work, the Mg97Y2Zn (at%) alloys have been processed by the rotary-die (RD) ECAP technique under different temperature or pass conditions to tailor the microstructure and mechanical properties. The results reveal that the DRX ratio (VDRX) increases with the increase of ECAP pass or temperature, weakening the texture intensity. The refinement of DRXed grains saturates after 9 passes of ECAP at 350 °C, while the size of DRXed grains (dDRX) monotonically decreases with the reduction of temperature from 400 °C to 300 °C in the case of the same pass number (9P). The combination of fine dDRX, large VDRX, and directionally arranged LPSO phases along the extrusion direction endows the 350 °C–9P alloy with the optimum strength and ductility synergy, exhibiting the tensile yield strength (TYS) of 267 MPa, ultimate tensile strength (UTS) of 318 MPa, and elongation of 10.3%. However, the larger VDRX, weaker texture, less low-angle grain boundaries combine to improve the elongation (14.7%) of the 400 °C–9P alloy at the expense of the TYS (231 MPa) and UTS (303 MPa). This paper provides useful information to optimize the RD-ECAP parameters for the fabrication of high-performance Mg97Y2Zn alloys.

Dual-stream pyramid and attention network for stereo matching

In order to accurately estimate the disparity of ill-posed regions, such as weak texture and occlusion regions, we propose DSPANet, a stereo matching network that incorporates a dual-stream pyramid module and a channel and spatial attention module. The dual-stream pyramid module captures numerous complementary features from different layers by utilizing multi-resolution inputs and feature extraction blocks. This approach enables the learning of local detailed features at various scales. These features at various scales are then combined to calculate the stereo matching cost. By incorporating channel and spatial attention module into the feature extraction process to obtain richer and more concise contextual information, the matching cost can be constructed more accurately, which provides powerful conditions for subsequent cost aggregation. In the cost aggregation stage, we utilize the stacked hourglass module for both encoding and decoding. Additionally, we incorporate 3D global attention upsampling during the decoding stage, which enables high-level features to provide guidance information to low-level features in a simple way. We evaluate our proposed method on the Scene Flow dataset, as well as the KITTI2012 and KITTI2015 datasets. The experimental results demonstrate that our DSPANet achieves superior performance and effectively enhances the matching results in ill-posed regions. Our code has been implemented using PyTorch and will be released after paper publication at https://github.com/Shi-LiQing/DSPANet.

Effect of nano-sized Mo particles on microstructure and mechanical properties of Mo/AZ31 composites

It is a challenge to enhance both strength and ductility for the Mg matrix composites. In the present work, nanostructured Mo-reinforced AZ31 composites are synthesized via powder metallurgy method. The addition of nano-sized Mo particles reduces the grain size and decreases the texture strength of the composites. The Al5Mo and Al8Mo3 phases are formed through Al atoms diffusion and interaction with Mo atoms. The synergistically reinforcing effect of Al5Mo, Al8Mo3 and nano-sized Mo particles improves the mechanical properties of the composites. The nano-Mo/AZ31 composite with 1.0 wt% Mo addition exhibits superior mechanical properties, with yield strength, ultimate tensile strength, and elongation of 221 MPa, 328 MPa and 12.4 %, respectively. These results mean the improvements of 20.1 %, 17.1 % and 21.6 % compared to the AZ31 Mg alloy. The strength enhancement is attributed to the grain refinement and Orowan strengthening, while the improvement in ductility is attributed to grain size reduction, texture weakening, and the activation of a broader range of slip systems.

Effect of initial grain size on the recrystallization behavior and recrystallization texture of a Mg–3Gd alloy

The effect of initial grain size on the recrystallization and recrystallization texture of a rolled Mg–3Gd (wt.%) alloy is studied in detail. The results show that the deformation microstructure of an initially coarse-grained (CG) sample has a larger twinned area and a higher density of twin boundaries than a fine-grained (FG) sample. After annealing, the CG sample recrystallizes preferentially in the twinned area, whereas the FG sample adopts the higher density grain boundaries as the nucleation sites. Furthermore, weak recrystallization texture components appear from the grain nucleation stage, regardless of the initial grain size, and are preserved after complete recrystallization due to uniform grain growth. The majority of recrystallization texture is deviated 20°–45° away from normal direction (ND), accounting for more than 50 %. Especially, the recrystallization texture of the FG sample is a “Rare Earth texture”, in contrast to the widely reported texture modification unrelated to grain boundary nucleation. Only a scattered basal texture is observed in the CG sample, which also differs from the reported “Rare Earth texture” originating from shear band nucleation in dilute Mg–Gd alloys. Finally, based on the Johnson–Mehl–Avrami–Kolmogorov (JMAK) model, the recrystallization kinetics are calculated, and it is found that the initial grain size mainly affects the nucleation rate, and has limited effect on the grain growth rate.

Unveiling the microstructure evolution and deformation mechanisms of micro-alloying Mg-1.5Zn-0.5Zr-0.5Sr alloy via extrusion-shearing

Obtaining high strength without sacrificing ductility in Mg alloys is of great importance. However, the hexagonal close-packed crystal structure critically limits the deformation of Mg alloys. In this work, novel Mg-1.5Zn-0.5Zr-0.5Sr (wt.%) alloys were fabricated through homogenization, direct extrusion (DE), and extrusion-shearing (ES) processes. The effects of three different processes on microstructures, texture, and mechanical properties were thoroughly investigated. The results show that the average grain size of the alloy was greatly refined after hot extrusion. The alloy after the DE process exhibited a typical bimodal microstructure with dynamic recrystallized (DRXed) grains showing a strong <101‾0 > fiber texture parallel to extrusion direction (ED), while the small-sized unDRXed grains showed relatively weak basal texture. In addition, the strength was significantly improved, with the value of yield strength (YS) reaching ∼229 MPa and ultimate tensile strength (UTS) reaching ∼262 MPa. This is mainly attributed to grain boundary strengthening and dislocation strengthening. Compared to the DE-processed alloy, the basal texture intensity was significantly decreased and non-basal slip was activated after the ES process, which indicates that the shearing deformation during ES could improve the ductility, with the fracture elongation (FE) increased from ∼13.9 % to ∼21.4 %. This work aims to unveil the deformation mechanisms in Mg-1.5Zn-0.5Zr-0.5Sr alloy and provides an efficient way for preparing micro-alloying Mg alloys with low cost and significant improvement of comprehensive properties.

Study on strengthening and toughening mechanism of Ti-Ni-Al ternary metallic-intermetallic laminated composites: Effect of interlaminar phase and Ti2Ni morphology

Further improving the strengthening and toughening properties of traditional binary metallic-intermetallic laminated (MIL) composites has become an urgent problem to be solved. In this work, a thin 0.02 mm Ni layer was added between the Ti and Al layers to prepare novel Ti-Ni-Al ternary MIL (TNA-TMIL) with multi-layer heterostructure (HS) characteristics, which significantly improved strength and toughness. TNA-TMIL composites include intermetallic layers (Ni-Al and Ni-Ti), unreacted metal layers (Ti-6Al-4V and Ni) and special morphology Ti2Ni in the Ti-6Al-4V layers. With the rise of hot-pressing temperature from 710 ℃ to 900 ℃, the eutectoid reaction and the increase of Ni element in Ti-6Al-4V alloy caused the morphology of Ti2Ni in the Ti-6Al-4V layer transition from island-like to network-like/needle-like to transition regions and finally formed needle-like Ti2Ni morphology. The electron backscatter diffraction (EBSD) analysis of the needle-like Ti2Ni shows that it has a weak cubic texture, and its fine grains are composed of dynamic recrystallization (DRX) and deformed grains. Nano-indentation and compression experiments indicated that the mechanical properties of TNA-TMIL composites are related to the different HS combinations and the morphology changes of Ti2Ni in the Ti-6Al-4V layer. This discovery may provide experience for designing and developing ternary MIL composites.

XLPE cable joint defects measurement method based on point cloud remapping

For the problems of low detection accuracy caused by weak texture and strong reflection on the surface of Cross-linked polyethylene (XLPE) high-voltage cable joint, this paper proposed a joint point cloud remapping and image segmentation method for the cable joint defects measurement, which improved the defects measurement accuracy from the correlation of the disordered point cloud and the salience of defects. Firstly, the discrete noise in the joint point cloud is removed using a modified radius filtering algorithm. Secondly, for the problems of disorderliness and weak association of the cable joint point cloud, we utilized Rodrigues' rotation formula (RRF) to correct the position of the XLPE cable joint point cloud. Then, for the demand of improving feature expression of non-significant defects, a point cloud remapping method is proposed, which remapping insignificant defects point cloud into saliency image for more effective detection. Lastly, an image segmentation method based on 8-neighborhood growth is applied for explicit defects detection, and the detected defects are quantified by the proposed measurement approach. Extensive experiments are conducted on the collected cable joints' point cloud and simulated cable joints' point cloud. With the help of the proposed method, we can achieve the best effects in DE, FR quantitative comparison, and the measurement inaccuracies are less than 4.5%. Experimental results demonstrate that the proposed method enjoys state-of-the-art performance in cable joint defects measurement.

Learning accurate template matching with differentiable coarse-to-fine correspondence refinement

Template matching is a fundamental task in computer vision and has been studied for decades. It plays an essential role in manufacturing industry for estimating the poses of different parts, facilitating downstream tasks such as robotic grasping. Existing methods fail when the template and source images have different modalities, cluttered backgrounds, or weak textures. They also rarely consider geometric transformations via homographies, which commonly exist even for planar industrial parts. To tackle the challenges, we propose an accurate template matching method based on differentiable coarse-to-fine correspondence refinement. We use an edge-aware module to overcome the domain gap between the mask template and the grayscale image, allowing robust matching. An initial warp is estimated using coarse correspondences based on novel structure-aware information provided by transformers. This initial alignment is passed to a refinement network using references and aligned images to obtain sub-pixel level correspondences which are used to give the final geometric transformation. Extensive evaluation shows that our method to be significantly better than state-of-the-art methods and baselines, providing good generalization ability and visually plausible results even on unseen real data.

SA-MVSNet: Self-attention-based multi-view stereo network for 3D reconstruction of images with weak texture

Multi-view stereo (MVS) reconstruction is a key task of image-based 3D reconstruction, and deep learning-based methods can achieve better results than traditional algorithms. However, most of the current deep learning-based MVS methods use convolutional neural networks (CNNs) to extract image features, which cannot achieve the aggregation of long-distance context information and capture robust global information. In addition, in the process of fusing depth maps into point clouds, the confidence filters will filter out the depth values with low confidence in weak texture areas. These problems will lead to the low completeness of 3D reconstruction of weak texture and texture-less areas. To address the above problems, this paper proposes SA-MVSNet based on the PatchmatchNet with a self-attentive mechanism. First, we design a coarse-to-fine network framework to advance depth map estimation. In the feature extraction network, a module with a pyramid structure based on Swin Transformer Block is used to replace the original Feature Pyramid Network (FPN), and the self-correlation between weak texture areas is enhanced by applying a global self-attention mechanism. Then, we also propose a self-attention-based adaptive propagation module (SA-AP), which applies a self-attention calculation within depth value propagation window to obtain the relative weight values of current pixel and others, and then adaptively samples the depth values of neighbors on the same surface for propagation. Experiments show that SA-MVSNet has significantly improved the completeness of 3D reconstruction for the images with weak texture on DTU (provided by Danish Technical University), BlendedMVS, and Tanks and Temple datasets. Our code is available at https://github.com/miaowang525/SA-MVSNet.

Grain alignment in hexaferrite permanent magnets by compaction at room and elevated temperatures.

Shape-controlled precursors enable grain alignment without using an applied magnetic field in the permanent magnet material strontium hexaferrite. The effect is investigated by conducting four series of experiments using different compaction methods: two cold and two hot compactions. The hypothesis is that magnetic short-circuiting will diminish the grain alignment (texture) in cold compacted samples. The two cold compactions were performed below the Curie temperature, one being a simple cold-pressing (CP) and the other being a cold-pressing followed by sintering (CPS). The CP samples showed a relatively weak texture and low remanent magnetisation, while the CPS method yielded a slightly sharpened texture and improved remanent magnetisation compared with CP. In the two hot compactions, the pressing was done above the Curie temperature, using spark plasma sintering (SPS) and induction pressing (IP). SPS and IP samples both showed remarkably sharper textures than the cold-pressed pellets, with roughly doubled texture strength, and consequently better remanent magnetisation. The bulk density of the hot compacted samples increased almost twofold compared with that of the cold compactions. Hot compaction outperforms cold compaction with regard to density, sharpness of texture and remanent magnetisation. This is explained by inter-particle magnetic short-circuiting during cold pressing, which hinders the alignment process and reduces the bulk density.

A new stereo matching algorithm based on improved four-moded census transform and adaptive cross pyramid model

&lt;p&gt;Stereo matching is still very challenging in terms of depth discontinuity, occlusions, weak texture regions, and noise resistance. To address the problems of poor noise immunity of local stereo matching and low matching accuracy in weak texture regions, a stereo matching algorithm (iFCTACP) based on improved four-moded census transform (iFCT) and a novel adaptive cross pyramid (ACP) structure were proposed. The algorithm combines the improved four-moded census transform matching cost with traditional measurement methods, which allows better anti-interference performance. The cost aggregation is performed on the adaptive cross pyramid structure, a unique structure that improves the traditional single mode of the cross. This structure not only enables regions with similar color and depth to be connected but also achieves cost smoothing across regions, significantly reducing the possibility of mismatch due to inadequate corresponding matching information and providing stronger robustness to weak texture regions. Experimental results show that the iFCTACP algorithm can effectively suppress noise interference, especially in illumination and exposure. Furthermore, it can markedly improve the error matching rate in weak texture regions with better generalization. Compared with some typical algorithms, the iFCTACP algorithm exhibits better performance whose average mismatching rate is only 3.33$ \% $.&lt;/p&gt;

Transformer network for stereo matching of weak texture objects

Transformer network for stereo matching of weak texture objects

Binocular vision SLAM with fused point and line features in weak texture environment

Binocular vision SLAM with fused point and line features in weak texture environment

Comparison of electron-beam-welded joints manufactured by rolled and selective laser melted reduced activation ferritic/martensitic steel

As the core component of China Fusion Engineering Test Reactor (CFETR), the forming and manufacturing of the first wall faces severe challenges due to the constraints of materials, service environment and geometric structure. The microstructure and mechanical properties of welded joints obtained by rolled and laser selective melted (SLMed) reduced activation ferritic/martensitic (RAFM) steel were investigated in depth. The microstructure analysis results showed that the weld metal (WM) was composed of coarse lath martensite and δ-ferrite, while the microstructure were mainly coarse grains in heat affected zone (HAZ) which was easy to break. The average grains size of SLMed RAFM steel (S-RAFMs) was smaller than that of rolled RAFM steel (R-RAFMs), since the rapid melting and solidification of laser in the additive manufacturing process prevented grains to grow. The microstructure of R-RAFMs joint was relatively more uniform and there was no obvious texture and preferred orientation, while the WM of S-RAFMs joint had a weak cubic texture. The results of mechanical properties showed that the S-RAFMs joints exhibited higher strength (UTS = 697.46 MPa) and the R-RAFMs joints exhibited better plasticity (EL = 13.49 %) under the same tensile conditions. The tensile and impact fracture morphologies showed that the fracture mechanism of two joints was ductile-brittle mixed fracture, and the dimples of R-RAFMs joints were finer and denser. It was clearly observed that the hardness of WM (∼390–430 HV0.3) was higher than that of the base metal (BM) (∼190–230 HV0.3) from the hardness distributions along the width directions. The average hardness of BM in R-RAFMs joint (227 HV0.3) was lower than S-RAFMs joint (197.3 HV0.3), which was mainly related to the fine grain strengthening effect during SLM forming. The above results may provide theoretical guidance for the high-performance manufacturing of the first wall and promote the application of SLM and EBW technology in the field of nuclear fusion.