Shape Detection Research Articles

Small hand-designed convolutional neural networks outperform transfer learning in automated cell shape detection in confluent tissues.

Mechanical cues such as stresses and strains are now recognized as essential regulators in many biological processes like cell division, gene expression or morphogenesis. Studying the interplay between these mechanical cues and biological responses requires experimental tools to measure these cues. In the context of large scale tissues, this can be achieved by segmenting individual cells to extract their shapes and deformations which in turn inform on their mechanical environment. Historically, this has been done by segmentation methods which are well known to be time consuming and error prone. In this context however, one doesn't necessarily require a cell-level description and a coarse-grained approach can be more efficient while using tools different from segmentation. The advent of machine learning and deep neural networks has revolutionized the field of image analysis in recent years, including in biomedical research. With the democratization of these techniques, more and more researchers are trying to apply them to their own biological systems. In this paper, we tackle a problem of cell shape measurement thanks to a large annotated dataset. We develop simple Convolutional Neural Networks (CNNs) which we thoroughly optimize in terms of architecture and complexity to question construction rules usually applied. We find that increasing the complexity of the networks rapidly no longer yields improvements in performance and that the number of kernels in each convolutional layer is the most important parameter to achieve good results. In addition, we compare our step-by-step approach with transfer learning and find that our simple, optimized CNNs give better predictions, are faster in training and analysis and don't require more technical knowledge to be implemented. Overall, we offer a roadmap to develop optimized models and argue that we should limit the complexity of such models. We conclude by illustrating this strategy on a similar problem and dataset.

Microwave-Sensor Array for Decoupling Detection of Distance, Shape, Dielectric, and Morphology

Microwave-Sensor Array for Decoupling Detection of Distance, Shape, Dielectric, and Morphology

Deep Time Series Forecasting With Shape and Temporal Criteria.

This paper addresses the problem of multi-step time series forecasting for non-stationary signals that can present sudden changes. Current state-of-the-art deep learning forecasting methods, often trained with variants of the MSE, lack the ability to provide sharp predictions in deterministic and probabilistic contexts. To handle these challenges, we propose to incorporate shape and temporal criteria in the training objective of deep models. We define shape and temporal similarities and dissimilarities, based on a smooth relaxation of Dynamic Time Warping (DTW) and Temporal Distortion Index (TDI), that enable to build differentiable loss functions and positive semi-definite (PSD) kernels. With these tools, we introduce DILATE (DIstortion Loss including shApe and TimE), a new objective for deterministic forecasting, that explicitly incorporates two terms supporting precise shape and temporal change detection. For probabilistic forecasting, we introduce STRIPE++ (Shape and Time diverRsIty in Probabilistic for Ecasting), a framework for providing a set of sharp and diverse forecasts, where the structured shape and time diversity is enforced with a determinantal point process (DPP) diversity loss. Extensive experiments and ablations studies on synthetic and real-world datasets confirm the benefits of leveraging shape and time features in time series forecasting.

Integrated Approach for Clothing Detection and Comparison using Structural Shape Detection and Texture Analysis

This paper presents a novel approach for detecting and finding similar clothing articles in images by combining structural shape detection and texture comparison. We utilized the Mask R-CNN model for structural shape detection and explored color hashing and Local Binary Pattern (LBP) histogram for texture comparison. Experiments were conducted using the DeepFashion2 dataset, demonstrating varying performance across different clothing articles and texture comparison algorithms. While the proposed solution achieved good detection rates for certain clothing articles, it faced challenges in detecting others due to the unequal distribution of articles in the dataset and limitations in texture comparison.

A Deep Detection Model based on Multi-task Learning for Appearance Defect of Solid Propellants

Solid propellants (SPs), as a high-energy material, are commonly used in military and industrial power systems, such as solid rocket and missiles. The SPs, however, confronts severe difficulties of inevitable defects while being made, thus bringing about the significance of inspection. However, the literatures before typically tackled this problem separately, which subsequently combines different models for the variety of defect patterns. Despite of the effectiveness, this act of matters usually brings excessive complexity and additionally computational burden. In this article, we managed to solve this problem in an integrated framework, which unite both the size detection task and shape detection task at the same time, but with different training strategies. To be specific, our framework is mostly consisted of two stage. Firstly, the SPs region is output using a semantic segmentation network, and size measurements are completed with traditional image processing to determine the size defects of the SPs. Then, the depth features of the segmentation network are combined with the semantic segmentation map to make a spatial attention mechanism, which is input to the deep classifier to complete the shape defect detection. The focus of model is gradually shifted from the segmentation task to the classification task as the number of training sessions increases by introducing dynamic balancing factors. The experimental results show that the multi-task learning approach can greatly improve the generalization and robustness of the model, and the accuracy and speed are improved for appearance defect detection of SPs.

Development and Testing of a High-Resolution Three-Dimensional Seismic Detection System for Gas Hydrate

As a novel type of mineral resource, gas hydrate has received a considerable amount of attention worldwide. This seismic detection method can detect abnormal phenomena such as the BSR, blank zones, velocity anomalies and polarity inversion of gas hydrate and become an important method of gas hydrate detection. The occurrence area of gas hydrate in the South China Sea is usually buried deep beneath the seabed. The current method cannot meet the needs of the shape and structure detection of gas hydrate deposits. With the support of the National Key R&D Program of China, some key technologies have led to developmental breakthroughs, such as ultra-high-energy plasma sources, small-group-interval high-resolution seismic streamers, and distributed three-dimensional seismic acquisition. The seismic profile obtained north of the South China Sea shows that the stratum penetration depth reaches nearly 1000 m at a depth of 1500 m, and the vertical resolution is better than 1.5 m. This system can serve the needs of high-resolution exploration of gas hydrate resources.

Single-shot 3D imaging with point cloud projection based on metadevice

Three-dimensional (3D) imaging is a crucial information acquisition technology for light detection, autonomous vehicles, gesture recognition, machine vision, and other applications. Metasurface, as a subwavelength scale two-dimensional array, offers flexible control of optical wavefront owing to abundant design freedom. Metasurfaces are promising for use as optical devices because they have large field of view and powerful functionality. In this study, we propose a flat optical device based on a single-layer metasurface to project a coded point cloud in the Fourier space and explore a sophisticated matching algorithm to achieve 3D reconstruction, offering a complete technical roadmap for single-shot detection. We experimentally demonstrate that the depth accuracy of our system is smaller than 0.24 mm at a measurement distance of 300 mm, indicating the feasibility of the submillimetre measurement platform. Our method can pave the way for practical applications such as surface shape detection, gesture recognition, and personal authentication.

Detection of the Irregular Shape of the Southern Limb of Menoetius from Observations of the 2017–2018 Patroclus–Menoetius Mutual Events

This work presents the analysis of seven mutual events of the Patroclus–Menoetius system (PMS) observed during the last season of mutual events, in 2017–2018. We compare the obtained light curves with those predicted using Grundy et al. and discuss the differences in the timing of the events and the drop of magnitude. Based on models of these observations, we present a collection of orbital parameters that provide the best fit for the observed events and compare the new solutions for the orbit of the system with solutions provided in the literature. Furthermore, we also discuss an interesting finding in the light curve of the only superior event in our collection. This light curve (the one with the best signal-to-noise ratio in our data set) shows the imprint of a possible crater in the south pole of Menoetius as deep as a fourth of its radius.

Research on Cigarette Shape Detection based on Dual Channel Acquisition

The cigarette shape detection system of cigarette machine is to online detect the unqualified cigarettes with warping, paper sticking, flattening and mouth burst in the high-speed flow cigarettes on the production line, and eliminate them in real time to ensure the product quality of finished cigarettes. At present, the photoelectric technology is adopted for the cigarette shape detection sensor, and the output signal range is analog quantity within 0-15Vdc. According to the field installation error, sensor difference and attenuation in use, the front and rear signal ranges are inconsistent. Corresponding modules are required to adjust the reference voltage and magnification of the front and rear channels and output standard 0-10vdc analog quantity. The current module is installed under the max detection wheel and adjusted through the general potentiometer of the internal circuit board. It is independent of the quality detection system and is far away from the MLP man-machine interface of the quality data display. Therefore, the adjustment process is very inconvenient, the device improves the calibration intelligent level of this part, forms a closed loop with the quality detection system, improves the adjustment accuracy, prolongs the service life, improves the staff work efficiency and facilitates the staff operation.

Real-time object detection and selection with the LabVIEW program

Nowadays, the demand for production is increasing due to the increase in the human population. For this reason, different developing technologies are used to meet the required production. In developing technologies, image processing techniques are used to save manpower and time and to minimize possible errors. In this study, image processing techniques were used to detect and select the colors and shapes of the objects coming over the conveyor belt system. Real-time images were preferred in the study. In the implemented system, the selection process was carried out by using the LabVIEW program to define the colors and shapes of the objects. LabVIEW NI- IMAQdx was used to find the colors of the objects. To facilitate the definition of shapes, the images taken from the Vision Assistant module in the LabVIEW program were converted to HSL format, and shape definitions were made using different algorithms. After these processes were done, the servo motors in the conveyor belt system were enabled to communicate with each other with the help of the Arduino program and the selection of the objects was carried out. According to the results obtained, the average accuracy rate for three-dimensional objects was 95.349 %. This rate is considered to be a very high rate for object detection of correct color and shape.

Vision-Action Semantic Associative Learning Based on Spiking Neural Networks for Cognitive Robot

Establishing cognitive environments help cognitive robots understand human actions, languages and observed objects. In this paper, a cognitive robotic model based on a novel spiking bidirectional associative memory (BAM) method is presented to establish a cognitive environment. The spiking BAM network uses a supervised spike-based learning rule to learn the relationship between the semantic information of vision and action, where a new coding scheme is proposed to encode and decode vision and action semantic information. Vision semantic information is obtained by the deeply salient shape detection (DSSD) method. Action semantic information is obtained by a spiking neural network (SNN) with the proposed coding method. Experimental results demonstrate that the spiking BAM achieves a good convergence ability and relatively high recall accuracy. Specially, for the cognitive development of the robot, this system can be used to improve the robot’s ability to infer human intentions in its natural interactions with humans.

ROBUST TECHNIQUES FOR BUILDING FOOTPRINT EXTRACTION IN AERIAL LASER SCANNING 3D POINT CLOUDS

Abstract. The building footprint is crucial for a volumetric 3D representation of a building that is applied in urban planning, 3D city modeling, cadastral and topographic map generation. Aerial laser scanning (ALS) has been recognized as the most suitable means of large-scale 3D point cloud data (PCD) acquisition. PCD can produce geometric detail of a scanned surface. However, it is almost impossible to get point clouds without noise and outliers. Besides, data incompleteness and occlusions are two common phenomena for PCD. Most of the existing methods for building footprint extraction employ classification, segmentation, voting techniques (e.g., Hough-Transform or RANSAC), or Principal Component Analysis (PCA) based methods. It is known that classical PCA is highly sensitive to outliers, even RANSAC which is known as a robust technique for shape detection is not free from outlier effects. This paper presents a novel algorithm that employs MCMD (maximum consistency within minimum distance), MSAC (a robust variant of RANSAC) and a robust regression to extract reliable building footprints in the presence of outliers, missing points and irregular data distributions. The algorithm is successfully demonstrated through two sets of ALS PCD.

Investigation of Rectangular Shape Object Detection and Classification Using Python

The existing fishcake machine can perform tasks to increase the productivity of fishcakes in the market. However, high productivity is not the only main purpose of a successful food manufacturing industries. The quality of the food itself need to accommodate the standard quality that has been lined by authorities to ensure that the quality of the food product must be at the best standard along with its high productivity. The objective of this work is to develop a monitoring method for detection of irregular shape and foreign particles of the fishcake during the fishcake production. This work will implement and uses both hardware and software elements to provide the best solution for the purpose of fishcake quality sustainability. Real-time image processing is the method used in this study by developing python program through Raspberry Pi 4. The algorithm of the programming is the main concern in this research. The library used for this study is OpenCV which is a common API used for computer vision.

GEOMETRIC FEATURES INTERPRETATION OF PHOTOGRAMMETRIC POINT CLOUD FROM UNMANNED AERIAL VEHICLE

Abstract. Recent days point clouds have become one of the most common 3D sources of information which is provides accurate geometry features of the object. 3D point clouds can be derived from either photogrammetry, Lidar or SAR in some cases depending upon the application. These point clouds consisting of 3D geospatial location of an object in form of XYZ coordinates which can be used in various ways to deduct information related to that object either based on visualisation or geometrical interpretation. Quality assessment standards for these point clouds are still very much in nascent stage with optimum accuracy in relative terms only. In this paper, multiple scale of point cloud has been used to understand the level of information these clouds consist on these multiple scales. Based on the 3D spatial information of these point cloud in local neighbourhood, some of invariant geometric properties can be computed for each 3D point with respective covariance matrix. These can be used to describe the local 3D structure using eigenvalues for these matrices. Using these Geometric features an approach is developed to understand the point cloud quality assessment. The proposed methodology exploits these special geometric properties to evaluate the 3D scene structure. Further, the point cloud is classified using shape detection algorithm which evaluates the geometric features to detect the mathematical shapes in the point cloud. This paper also enlightens on different geometric features that can be extracted from a point cloud and the importance of it.

Deep Learning for Live Cell Shape Detection and Automated AFM Navigation.

Atomic force microscopy (AFM) provides a platform for high-resolution topographical imaging and the mechanical characterization of a wide range of samples, including live cells, proteins, and other biomolecules. AFM is also instrumental for measuring interaction forces and binding kinetics for protein–protein or receptor–ligand interactions on live cells at a single-molecule level. However, performing force measurements and high-resolution imaging with AFM and data analytics are time-consuming and require special skill sets and continuous human supervision. Recently, researchers have explored the applications of artificial intelligence (AI) and deep learning (DL) in the bioimaging field. However, the applications of AI to AFM operations for live-cell characterization are little-known. In this work, we implemented a DL framework to perform automatic sample selection based on the cell shape for AFM probe navigation during AFM biomechanical mapping. We also established a closed-loop scanner trajectory control for measuring multiple cell samples at high speed for automated navigation. With this, we achieved a speed-up in AFM navigation and reduced the time involved in searching for the particular cell shape in a large sample. Our innovation directly applies to many bio-AFM applications with AI-guided intelligent automation through image data analysis together with smart navigation.

5D-fusion imaging for surface shape, polarization, and hyperspectral measurement.

We present a five-dimensional (5D) imager that is capable of simultaneous detection of the surface shape, spectral characteristics, and polarization states of macroscopic objects, and straightforwardly fuse collected data into a 5D data set. A polarized module that uses a polarized camera obtains polarized images, while a 3D hyperspectral module reconstructs the target as a 3D point cloud using a fringe projection technique. A liquid-crystal tunable filter is placed in front of the camera of this module to acquire spectral data that can be assigned to corresponding point clouds directly. The two modules are coupled by a dual-path configuration that allows the polarization information to be merged into a comprehensive point cloud with spectral information, generating a new 5D model. The 5D imager shows excellent performance, with a spectral resolution of 10 nm, depth accuracy of 30.7 µm, and imaging time of 8 s. Sample experiments on a toy car with micro scratch defects and a yellowing plant are presented to demonstrate the capabilities of the 5D imager and its potential for use in a broad range of applications, such as industrial manufacturing inspection, plant health monitoring, and biological analysis.

Three-dimensional ice shape detection based on flash pulse infrared thermal wave testing

Ice detection is an important issue in the field of icing prevention and de-icing. In this study, an experimental platform was built for ice detection using flash pulse infrared thermal wave detection, followed by a quantitative recognition approach for identifying the three-dimensional shape of ice. The new method was combined the edge recognition with thickness calculation of inverse heat transfer problem. And the edge of the ice was based on the Principal Component Analysis (PCA) and the Canny algorithm. Thus, by processing the ice edges and giving an initial thickness, the finite element model of the ice was established to numerically simulate the temperature distribution for ice thickness inversion based on the forward heat transfer problem. Meanwhile, the thickness of the ice was inversed by the Levenberg-Marquardt (LM) method based on the inverse heat transfer problem. The resulting edges and thickness of the ice were found to be in good agreement with the experimental results, demonstrating the feasibility of the proposed methods. This paves the way to explore an effective accurate and quantitative identification method for three-dimensional ice shape detection in infrared thermal wave detection.

Measurement report: An exploratory study of fluorescence and cloud condensation nuclei activity of urban aerosols in San Juan, Puerto Rico

Abstract. Many atmospheric aerosols are cloud condensation nuclei (CCN), capable of activating as cloud droplets when the relative humidity exceeds 100 %. Some primary biological aerosol particles (PBAPs), such as plant spores, pollen, or bacteria, have been identified as such CCN. Urban environments are a source of these bioaerosols, some of which are naturally produced by the local flora or are transported from surrounding regions and others of which are a result of human activities. In the latter case, open sewage, uncovered garbage, mold or other products of such activities can be a source of PBAPs. There have been relatively few studies, especially in the tropics, where PBAPs and CCN have been simultaneously studied to establish a causal link between the two. The metropolis of San Juan, Puerto Rico, is one such urban area with a population of 2 448 000 people (as of 2020). To better understand the fluorescent characteristics and cloud-forming efficiency of aerosols in this region, measurements with a wideband integrated bioaerosol spectrometer (WIBS), a condensation nuclei (CN) counter and a CCN spectrometer were made at the University of Puerto Rico – Río Piedras Campus. Results show that the CCN / CN activation ratio and the fraction of fluorescing aerosol particles (FAPs) have repetitive daily trends when the FAP fraction is positively correlated with relative humidity and negatively correlated with wind speed, consistent with previous studies of fungi spores collected on substrates. The results from this pilot study highlight the capabilities of ultraviolet-induced fluorescence (UV-IF) measurements for characterizing the properties of FAPs as they relate to the daily evolution of PBAPs. The use of multiple excitation and emission wavelengths, along with shape detection, allows the differentiation of different PBAP types. These measurements, evaluated with respect to previous, substrate-based analysis of the local fungal and pollen spores, have established a preliminary database of measurements that future, longer-term studies will build upon.

Modern Sophisticated Instrumental Techniques Used in the Characterization of Bhasma

A nano-metallic drug called Bhasma is created from metal and minerals. It has long been regarded in Ayurveda as a trustworthy and quick-acting dose type. The following procedures—Shodhana, Bhavana, Mardana, and Marana—are used to prepare Bhasma. These processes assist in breaking down particles into their smallest components, get rid of undesirable traits, and give the substance new properties that make them more palatable to the body's tissues and organs. In the classic, various Bhasma Parikshas, such as Varitaratwa, Unnama, Rekha purnatwa, Nirutha, and Apunarbhava, are mentioned. However, these tests are insufficient to determine the safety and efficacy of Bhasma. For the examination of Bhasma, various contemporary factors are used to ensure its acceptance on a global scale like physical evaluation, chemical analysis, and instrumental analysis. Bhasma efficiency is reliant on its homogenous chemical composition, nanocrystalline structure, and biological activity. These highly sophisticated analytical tools, including XRD, SEM, TEM, NTA, ICP-MS, XRF, AAS, and FTIR, are required for the detection of crystal size, particle size, shape, surface area, dispersion state, and morphological features, in addition to chemical and physical properties like reactivity, optical behaviour, magnetism, zeta potential, and others. To understand the toxicity, pharmacokinetics, and pharmacodynamics of Bhasma, all of these methods are employed. As a result, the current paper's main focus is on characterizing Bhasma using modern sophisticated instrumental techniques.

Experimental verification of different approaches for the determination of gas bubble equivalent diameter from optical imaging

The key group of research methods used for the analysis of formation and flow of gas bubbles in liquids are those involving digital imaging and shadowgraphy. Despite their frequent use, there is no comprehensive analysis of the compliance of different procedures to determine bubble size based on experimental data. In this work, fifteen different approaches for shape recognition were used to determine the equivalent diameter of a bubble and then the results obtained were compared with bubble equivalent diameters determined experimentally. The experiments concerned bubbles of equivalent diameters in the range of 2.4–5.4 mm which correspond to the size of the bubbles most commonly encountered in industrial practice. Particularly, three algorithms for bubble shape detection were evaluated and then three sets of parameters to express the size of the bubble and two types of bubble equivalent diameter were derived. It was found that the most accurate results are obtained by using image binarization and Canny edge detection for the determination of the bubble shape, combined with the Feret diameter employed to express its size and with the volumetric equivalent diameter instead of the superficial one. As the binarization method is not versatile, we recommend using the Canny edge detector.