20 February 2022 in Food Hydrocolloids
1 July 2019 in Signal Processing
16 September 2020 in IEEE Transactions on Instrumentation and Measurement
Reconstructing normal section profiles is an effective approach at monitoring the shape of a 3D revolving-symmetry structure. The existing reconstruction methods typically require a sensor to work in a constrained pose and thus suffer from poor flexibility and limited view-angle. This paper proposes a pose-unconstrained normal section profile reconstruction framework for 3D revolving-symmetry structures, which estimates the normal section profile using multiple 3D general section profiles acquired by a multi-line structured light vision sensor. First, we establish a model to estimate the revolving axis and calculate the normal section profile using corresponding points. Then, the model is embedded into an iterative algorithm to optimize the corresponding points and calculate the accurate normal section profile. Simulations show that our algorithm is applicable to 3D revolving structures of various shapes and is robust against noise and local surface defects. Real experiments were conducted on reconstructing the normal section profile of a 3D wheel. The results demonstrate that our algorithm reaches the mean precision of 0.065 mm and repeatability with the STD of 0.007mm. It is also robust to pose and position of the sensor.
30 December 2021 in IEEE Transactions on Instrumentation and Measurement
Recognizing small objects with fine-grained categories in a cluttered scenario is challenging in some industrial applications. This article proposes a 3-D fine-grained recognition framework for segmenting and classifying small-scale objects with local structural differences from cluttered point clouds. The framework consists of two stages: improved semantic segmentation of 3-D scenarios and fine-grained 3-D instance recognition. In the first stage, normal-angle cues are integrated into discriminative feature learning to enhance local structure representation for small objects. In addition, sampling strategies are used to improve the efficiency of cluttered scenario segmentation. In the second stage, we design a fine-grained classification subnetwork (FGC-Net) based on an attention mechanism that predicts subcategory labels of segmented object instances in metacategories. The experiments were conducted on an aeroengine point cloud collected in the real-world industrial scene. The results demonstrate that the proposed algorithm recognizes small components of multiple fine-grained subcategories with the best scene segmentation mIoU of 88.01% and the overall fine-grained recognition rate of 92.1%. The detailed intermediate result and ablation analysis are also presented. The method has been applied in the actual industrial manufacturing and inspecting field.
13 May 2024 in IEEE Transactions on Industrial Informatics
Incompleteness and irregularity are inherent challenges in 3-D point clouds, which hinder point-cloud-based 3-D object or scene understanding, especially for complicated industrial scenarios. Previous point cloud completion methods typically suffer from the nonuniform density of the restored point cloud. In this article, we proposed a global and uniform point cloud completion algorithm via a density-sensitive transformer for reconstructing complete and fine-grained 3-D small-scale industrial components. We first designed a position and feature encoding module to aggregate discriminative point cloud features. Then, we constructed a density-sensitive transformer structure with a coarse point cloud generator that allowed for recovering the global structure and uniform local details of the 3-D object. The experimental results on a real-world industrial component dataset and a public multiobject dataset demonstrated that our method achieved state-of-the-art performance with the former CD- $\mathcal{l}$ 1 of 8.17 × 10 -3 , avg-NUC of 0.61 and the latter CD- $\mathcal{l}$ 1 of 5.33 × 10 -3 , avg-NUC of 0.74. Our method better balanced the global point cloud integrity and density uniformity in a coarse-to-fine pipeline, which benefited high-quality 3-D object reconstruction. The method has been applied to practical aero-engine component reconstruction in the real-world scenario.
1 August 2023 in IEEE Transactions on Industrial Informatics
Segmenting small-scale and close 3-D objects from a complicated 3-D point cloud scene is a crucial yet challenging task for industrial scene understanding. Previous 3-D instance segmentation work (such as PointGroup and SoftGroup) suffers from limited accuracy when dealing with objects in close proximity. This article proposes a center-aware 3-D instance segmentation framework that performs two cooperative stages, including bottom-up center-guided instance proposal generation and top-down center-aware instance mask refinement, to jointly address the challenges of segmenting close objects. In the first stage, the object clusters are first segmented from the cluttered background. We propose to determine the center of each object by predicting its k -nearest neighbors with their respective offsets, which enables an instance proposal generated for each object. In the second stage, center-based relative position encoding is performed on all points in each instance proposal. We concatenate the explicitly encoded geometric features (low-level) with high-level semantic features and construct the center-aware feature representation for instance mask prediction. We conducted experiments on a real-world industrial point cloud dataset that involves the important aero-engine component—close clamps in the cluttered scenario. The results show that our proposed method achieves state-of-the-art overall 3-D instance segmentation precision with $\mathbf{AP}_{\mathbf{avg}}$ of 0.905 and $\mathbf{AP}_{\mathbf{90}}$ of 0.736.