Contour Detection Algorithm Research Articles

STetro-Deep Learning Powered Staircase Cleaning and Maintenance Reconfigurable Robot.

Staircase cleaning is a crucial and time-consuming task for maintenance of multistory apartments and commercial buildings. There are many commercially available autonomous cleaning robots in the market for building maintenance, but few of them are designed for staircase cleaning. A key challenge for automating staircase cleaning robots involves the design of Environmental Perception Systems (EPS), which assist the robot in determining and navigating staircases. This system also recognizes obstacles and debris for safe navigation and efficient cleaning while climbing the staircase. This work proposes an operational framework leveraging the vision based EPS for the modular re-configurable maintenance robot, called sTetro. The proposed system uses an SSD MobileNet real-time object detection model to recognize staircases, obstacles and debris. Furthermore, the model filters out false detection of staircases by fusion of depth information through the use of a MobileNet and SVM. The system uses a contour detection algorithm to localize the first step of the staircase and depth clustering scheme for obstacle and debris localization. The framework has been deployed on the sTetro robot using the Jetson Nano hardware from NVIDIA and tested with multistory staircases. The experimental results show that the entire framework takes an average of 310 ms to run and achieves an accuracy of for staircase recognition tasks and accuracy for obstacle and debris detection tasks during real operation of the robot.

영상기반 비상착륙영역 인식 프레임워크를 위한 심층 합성곱 신경망 기반 영상처리 알고리즘

The UAM (Urban Air Mobility) industry is currently in the limelight To adopt sustainable urban transport systems and contribute to the advancement of aerospace technology, the MOLIT (Ministry of Land, Infrastructure and Transport) of South Korea has developed the K-UAM (Korea Urban Air Mobility) roadmap by collaborating with industries and academia to lead research and development related to the emerged industry. To guarantee the high reliability of industrial-level autonomous flights in urban areas, high priority should be provided to observe flight safety and survivability measures. This paper proposes adopting DNN (Deep Neural Network) based image processing for a vision-based safe landing region recognition framework to satisfy research and development needs. This framework is one of the core elements of the flight contingency plan for UAM and drone transportation in urban environments. Our proposed algorithm can recognize the coarse candidate region to land safely by using a DNN and then determine the refined safety zone by using the contour detection algorithm. To verify the performance of the proposed contribution, a simulation was done and the results were analyzed. The findings proved to be valid with regard to both the quantitative and qualitative terms. Therefore, our proposed framework will not only overcome the image processing complexity of the real world but also improve the enhancement of the flight emergency relief performance.

Two-dimensional high-precision non-contact automatic measurement method based on image corner coordinates

Aiming at some shortcomings of the existing non-contact two-dimensional high-precision measurement methods, this paper proposes a two-dimensional high-precision non-contact automatic measurement method based on the corner coordinates of the image. Firstly, this paper designs a set of simple image acquisition device and explains the advantages of the Canny operator used in the image contour detection algorithm. Subsequently, this paper proposes a dimension calibration algorithm based on image corner coordinates, which can convert the pixel size to the actual size, and achieves the function of the algorithm by hierarchical, multi-step processing of the image. Finally, in order to realize the intelligent positioning and selection of the standard size workpiece position, an automatic measurement and positioning system is designed, which can convert the actual size signal into the pulse time control signal. The experimental results show that the measurement method proposed in this paper has the advantages of fast measurement speed, high robustness, low cost and high degree of automation. When using a black-and-white checkerboard paper with an accuracy of 0.1 mm, the measurement accuracy can reach the micron level.

Dense connection decoding network for crisp contour detection

In the past few years, contour detection algorithm has made obvious progress with the help of convolutional neural networks. The aim of this paper is to present a novel network connecting low- and high-resolution features to make the network achieving richer feature representation. First, VGG net is used as encoding part with outputting the features of different resolutions, and then the feature maps are combined in some specific resolution with up- or down-sample method. The combining process can be stack step-by-step. The proposed network makes the encoding part deeper to extract richer convolutional features. The experiments have shown that the proposed method improves the contour detection performances and outperform some existed convolutional neural networks based methods on BSDS500 and NYUD-V2 datasets.

Autonomous Floor and Staircase Cleaning Framework by Reconfigurable sTetro Robot with Perception Sensors

Cleaning multi-storey buildings need to be considered while developing autonomous service robots. In this paper, we introduce a novel reconfigurable platform called sTetro with the abilitiesto navigate on the floor as well as to detect then climb the staircase autonomously. To this end, an operational framework for this cleaning robot that leverages on customized deep convolution neural network (DCNN) and the RGBD camera to locate staircases in the 3D prebuilt map and then to plan trajectories by maximizing area coverage for both floor and staircase in the multi-storey environments is proposed. While building a 3D map, the staircase location is identified at the 3D point close to the center of the staircase first step using a contour detection algorithm from the boundary of the detected staircase by DCNN. The robot follows the planned trajectory to clear the floor then approaching the staircase location accurately to execute the climbing mode while cleaning the staircase to reach the next floor. The proposed methods archive the high accuracy in identifying the presence of the different staircase types, and the first step locations. Moreover, the multi-storey building evaluations have demonstrated the efficiency of the sTetro in terms of the area coverage both staircase and floor free space.

СИСТЕМА ТЕХНІЧНОГО ЗОРУ З ШТУЧНИМ ІНТЕЛЕКТОМ ДЛЯ ЗАХОПЛЕННЯ ЦИЛІНДРИЧНИХ ОБ'ЄКТІВ РОБОТОМ

The article presents a variant of working with a system of technical vision, which recognizes cylindrical objects. This vision system based on artificial intelligence, which allows you to determine the circles in the image. The coordinates of the value of the circle are necessary for the exact positioning of the robot manipulator. The calculation of the gradient and threshold separation determine the gaps in the intensity of the image of the object. These methods define pixels lying on the border between the object and the background. The further process consists in connection of the segments of a contour separated by small intervals, and in association of separate short segments. Thus, contour detection algorithms accompanies by procedures for constructing object boundaries from the corresponding pixel sequences. The resulting image has sufficient information for artificial intelligence analysis to detect the circle. The software is developed and experimentally tested in operation. The operation of the technical vision system experimentally tested in the work, namely the capture of a cylindrical object. Work vision systems has been experimentally proven to work and is passionate cylindrical object. The coordinates of the value of the circle, which are necessary for the exact location of the robot manipulator were determined by artificial intelligence in 41 milliseconds. The obtained coordinates were transmitted to the microprocessor to adjust the position of the manipulator. The robot accurately captured a cylindrical object.

Plasmodium Detection using Machine Learning

Plasmodium is one of India's biggest public health problems. Early prediction of a malaria epidemic is that the secret to malaria morbidity management, mortality as well as reducing the risk of malaria transmission in the community will benefit politicians, health care providers, medical officers, health ministry and other health organizations to better target medical resources to areas of greatest need. In this project, we acquire data sets from hospital databases, which have the information about the causes of malaria, and the images of cells infected with malaria. We then analyze these data sets and feed them to our machine-learning model. Here we are using contour detection and random forest algorithms for training the model and predicting the output

Contour detection in Cassini ISS images based on Hierarchical Extreme Learning Machine and Dense Conditional Random Field

In Cassini ISS (Imaging Science Subsystem) images, contour detection is often performed on disk-resolved objects to accurately locate their center. Thus, contour detection is a key problem. Traditional edge detection methods, such as Canny and Roberts, often extract the contour with too much interior details and noise. Although the deep convolutional neural network has been applied successfully in many image tasks, such as classification and object detection, it needs more time and computer resources. In this paper, a contour detection algorithm based on H-ELM (Hierarchical Extreme Learning Machine) and DenseCRF (Dense Conditional Random Field) is proposed for Cassini ISS images. The experimental results show that this algorithm’s performance is better than both traditional machine learning methods, such as Support Vector Machine, Extreme Learning Machine and even deep Convolutional Neural Network. The extracted contour is closer to the actual contour. Moreover, it can be trained and tested quickly on the general configuration of PC, and thus can be applied to contour detection for Cassini ISS images.

Contour Detection by Simulating the Curvature Cell in the Visual Cortex and its Application to Object Classification

This paper addresses contour detection by simulating the human visual system and its application to visual object classification. Unlike previously designed bioinspired contour detection algorithms, we consider contour to be the salience of an edge image, and we extract the salience by simulating the endstopped cell and curvature cell in the visual cortex. Generally, we follow a local-to-global feed-forward architecture, in which the size of the receptive field (RF) increases from the primary visual cortex to the higher visual cortex. Edges are first detected by simple cells in small RFs, where textural details are suppressed by non-classical receptive fields (NCRFs) and sparse coding. Second, edges are integrated into local segments by complex cells. Afterwards, they are combined into the salience of edge images by endstopped cells and curvature cells and are ultimately the core of the final contour. In addition, we also apply the bioinspired contour detection algorithm to visual object classification tasks. Experiments on contour extraction show that, compared with state-of-the-art bioinspired algorithms, our algorithm makes a considerable improvement on contour detection. Experiments on visual object classification show that the contours produced by our proposal are powerful representations of the original images, which implies that our proposal is both biologically plausible and technologically useful.

AI Radar Sensor: Creating Radar Depth Sounder Images Based on Generative Adversarial Network.

Significant resources have been spent in collecting and storing large and heterogeneous radar datasets during expensive Arctic and Antarctic fieldwork. The vast majority of data available is unlabeled, and the labeling process is both time-consuming and expensive. One possible alternative to the labeling process is the use of synthetically generated data with artificial intelligence. Instead of labeling real images, we can generate synthetic data based on arbitrary labels. In this way, training data can be quickly augmented with additional images. In this research, we evaluated the performance of synthetically generated radar images based on modified cycle-consistent adversarial networks. We conducted several experiments to test the quality of the generated radar imagery. We also tested the quality of a state-of-the-art contour detection algorithm on synthetic data and different combinations of real and synthetic data. Our experiments show that synthetic radar images generated by generative adversarial network (GAN) can be used in combination with real images for data augmentation and training of deep neural networks. However, the synthetic images generated by GANs cannot be used solely for training a neural network (training on synthetic and testing on real) as they cannot simulate all of the radar characteristics such as noise or Doppler effects. To the best of our knowledge, this is the first work in creating radar sounder imagery based on generative adversarial network.

Delineation of agricultural fields in smallholder farms from satellite images using fully convolutional networks and combinatorial grouping

Accurate spatial information of agricultural fields in smallholder farms is important for providing actionable information to farmers, managers, and policymakers. Very High Resolution (VHR) satellite images can capture such information. However, the automated delineation of fields in smallholder farms is a challenging task because of their small size, irregular shape and the use of mixed-cropping systems, which make their boundaries vaguely defined. Physical edges between smallholder fields are often indistinct in satellite imagery and contours need to be identified by considering the transition of the complex textural pattern between fields. In these circumstances, standard edge-detection algorithms fail to extract accurate boundaries. This article introduces a strategy to detect field boundaries using a fully convolutional network in combination with a globalisation and grouping algorithm. The convolutional network using an encoder-decoder structure is capable of learning complex spatial-contextual features from the image and accurately detects sparse field contours. A hierarchical segmentation is derived from the contours using the oriented watershed transform and by iteratively merging adjacent regions based on the average strength of their common boundary. Finally, field segments are obtained by adopting a combinatorial grouping algorithm exploiting the information of the segmentation hierarchy. An extensive experimental analysis is performed in two study areas in Nigeria and Mali using WorldView-2/3 images and comparing several state-of-the-art contour detection algorithms. The algorithms are compared based on the precision-recall accuracy assessment strategy which is tolerating small localisation errors in the detected contours. The proposed strategy shows promising results by automatically delineating field boundaries with F-scores higher than 0.7 and 0.6 on our two test areas, respectively, outperforming alternative techniques.

People Counting in Elevator Car Based on Computer Vision

To detect the number of passengers in an elevator based on computer vision technology, image preprocessing and human contour detection algorithms were proposed. After the target image was edge-detected and morphologically eroded, a binary image containing only human targets was obtained. The edges of the image were detected using canny operator, it was most suitable for the edge detection of the human body contour in elevator. Person counting was mainly performed by using morphological mark outline connection area counting method. According to the algorithm flow chart, the real elevator monitoring video was tested and verified. The number of people with fewer numbers (less than 10 persons) was highly accurate.

Combining SUN-based visual attention model and saliency contour detection algorithm for apple image segmentation

Accurate segmentation of apple fruit under natural illumination conditions provides benefits for growers to plan relevant applications of nutrients and pesticides. It also plays an important role for monitoring the growth status of the fruit. However, the segmentation of apples throughout various growth stages had only achieved a limited success so far due to the color changes of apple fruit as it matures as well as occlusion and the non-uniform background of apple images acquired in an orchard environment. To achieve the segmentation of apples with different colors and with various illumination conditions for the whole growth stage, a segmentation method independent of color was investigated. Features, including saliency and contour of the image, were combined in this algorithm to remove background and extract apples. Saliency using natural statistics (SUN) visual attention model was used for background removal and it was combined with threshold segmentation algorithm to extract salient binary region of apple images. The centroids of the obtained salient binary region were then extracted as initial seed points. Image sharpening, globalized probability of boundary-oriented watershed transform-ultrametric contour map (gPb-OWT-UCM) and Otsu algorithms were applied to detect saliency contours of images. With the built seed points and extracted saliency contours, a region growing algorithm was performed to accurately segment apples by retaining as many fruit pixels and removing as many background pixels as possible. A total of 556 apple images captured in natural conditions were used to evaluate the effectiveness of the proposed method. An average segmentation error (SE), false positive rate (FPR), false negative rate (FNR) and overlap Index (OI) of 8.4, 0.8, 7.5 and 90.5% respectively, were achieved and the performance of the proposed method outperformed other six methods in comparison. The method developed in this study can provide a more effective way to segment apples with green, red, and partially red colors without changing any features and parameters and therefore it is also applicable for monitoring the growth status of apples.

Detection of Lung Contour with Closed Principal Curve and Machine Learning.

Radiation therapy plays an essential role in the treatment of cancer. In radiation therapy, the ideal radiation doses are delivered to the observed tumor while not affecting neighboring normal tissues. In three-dimensional computed tomography (3D-CT) scans, the contours of tumors and organs-at-risk (OARs) are often manually delineated by radiologists. The task is complicated and time-consuming, and the manually delineated results will be variable from different radiologists. We propose a semi-supervised contour detection algorithm, which firstly uses a few points of region of interest (ROI) as an approximate initialization. Then the data sequences are achieved by the closed polygonal line (CPL) algorithm, where the data sequences consist of the ordered projection indexes and the corresponding initial points. Finally, the smooth lung contour can be obtained, when the data sequences are trained by the backpropagation neural network model (BNNM). We use the private clinical dataset and the public Lung Image Database Consortium and Image Database Resource Initiative (LIDC-IDRI) dataset to measure the accuracy of the presented method, respectively. To the private dataset, experimental results on the initial points which are as low as 15% of the manually delineated points show that the Dice coefficient reaches up to 0.95 and the global error is as low as 1.47 × 10-2. The performance of the proposed algorithm is also better than the cubic spline interpolation (CSI) algorithm. While on the public LIDC-IDRI dataset, our method achieves superior segmentation performance with average Dice of 0.83.

All-sky photogrammetry techniques to georeference a cloud field

Abstract. In this study, we present a novel method of identifying and geolocalizing cloud field elements from a portable all-sky camera stereo network based on the ground and oriented towards zenith. The methodology is mainly based on stereophotogrammetry which is a 3-D reconstruction technique based on triangulation from corresponding stereo pixels in rectified images. In cases where clouds are horizontally separated, identifying individual positions is performed with segmentation techniques based on hue filtering and contour detection algorithms. Macroscopic cloud field characteristics such as cloud layer base heights and velocity fields are also deduced. In addition, the methodology is fitted to the context of measurement campaigns which impose simplicity of implementation, auto-calibration, and portability. Camera internal geometry models are achieved a priori in the laboratory and validated to ensure a certain accuracy in the peripheral parts of the all-sky image. Then, stereophotogrammetry with dense 3-D reconstruction is applied with cameras spaced 150 m apart for two validation cases. The first validation case is carried out with cumulus clouds having a cloud base height at 1500 m a.g.l. The second validation case is carried out with two cloud layers: a cumulus fractus layer with a base height at 1000 m a.g.l. and an altocumulus stratiformis layer with a base height of 2300 m a.g.l. Velocity fields at cloud base are computed by tracking image rectangular patterns through successive shots. The height uncertainty is estimated by comparison with a Vaisala CL31 ceilometer located on the site. The uncertainty on the horizontal coordinates and on the velocity field are theoretically quantified by using the experimental uncertainties of the cloud base height and camera orientation. In the first cumulus case, segmentation of the image is performed to identify individuals clouds in the cloud field and determine the horizontal positions of the cloud centers.

A novel contour-based registration of lateral cephalogram and profile photograph

A contour-based automatic registration method of lateral cephalograms and profile photographs is proposed to facilitate the accurate measurement of anteroposterior (AP) position of maxillary central incisors relative to the forehead. There are mainly three steps in the proposed method. First, a hierarchical contour detection algorithm is employed to obtain contours of forehead and nose in lateral cephalograms and profile photographs, respectively. Curve deviation around Nasion of Soft Tissue (Ns) is further corrected using an iterative polynomial curve fitting algorithm according to the characteristics of cephalograms. Second, the coarse registration is performed based on four landmark pairs using a routine least squares method for each image pair, and then the contours are roughly matched. Finally, the coarse registration is further refined by applying the coherent point drift (CPD) algorithm to all the contour points. Using image pairs of lateral cephalograms and profile photographs from 49 Chinese patients, we show that the proposed method achieves higher performance in terms of accuracy and robustness than the existing methods.

Contour detection in an infrared image using the modified snake algorithm

Infrared and thermal images have been used widely in the different forensics and security applications. Such images show the temperature difference between different objects and scene background. One of the drawbacks of such images is low contrast and noisy images which should be enhanced. This paper presents a new thermal image contour detection algorithm using the modified snake algorithm. The segmentation algorithm based on the image enhancement and the modified model of active contours based on regions, taking into account the calculation of the anisotropic gradient. Some presented experimental results illustrate the performance of the proposed cloud system on real thermal images in comparison with the traditional methods.

Contour detection via stacking random forest learning

Abstract Contour detection is an important and fundamental problem in computer vision which finds numerous applications. Despite significant progress has been made in the past decades, contour detection from natural images remains a challenging task due to the difficulty of clearly distinguishing between edges of objects and surrounding backgrounds. To address this problem, we first capture multi-scale features from pixel-level to segment-level using local and global information. These features are mapped to a space where discriminative information is captured by computing posterior divergence of Gaussian mixture models and sufficient statistics based on deep Boltzmann machine. Then we introduce a stacking random forest learning framework for contour detection. We evaluate the proposed algorithm against leading methods in the literature on the Berkeley segmentation and Weizmann horse data sets. Experimental results demonstrate that the proposed contour detection algorithm performs favorably against state-of-the-art methods in terms of speed and accuracy.

Computer-aided Kidney Segmentation on Abdominal CT Images Using Fuzzy Based Denoising for Gaussian Noise

In this paper, we have proposed an image filtering technique (fuzzy logic based approach) to preserve important signal elements such as edges, smoothing the details of an abdominal CT image such as kidney to make it appear clear and sharpener. At first, abdominal CT images are retrieved from Gaussian noise using fuzzy logic based approach. After that, kidney region have been segmented from the abdomen area using region growing algorithm. In region growing process, the first stage detects the abdomen boundary using contour detection algorithm. The second stage identifies the kidney using seed point location. After successful segmentation, the kidney region is extracted and it is given to some region growing methods such as Region Growing Interest (ROI), pixel filling, erosion, labeling and dilation to find the accurate segmented area of these abdominal CT images. The results of a series of tests on 62 images from 16 patients indicate an interrelationship up to 73% between automatic and manual segmentation.

Implementation in an FPGA circuit of Edge detection algorithm based on the Discrete Wavelet Transforms

The 2D Discrete Wavelet Transform (DWT) is a computationally intensive task that is usually implemented on specific architectures in many imaging systems in real time. In this paper, a high throughput edge or contour detection algorithm is proposed based on the discrete wavelet transform. A technique for applying the filters on the three directions (Horizontal, Vertical and Diagonal) of the image is used to present the maximum of the existing contours. The proposed architectures were designed in VHDL and mapped to a Xilinx Sparten6 FPGA. The results of the synthesis show that the proposed architecture has a low area cost and can operate up to 100 MHz, which can perform 2D wavelet analysis for a sequence of images while maintaining the flexibility of the system to support an adaptive algorithm.