Real-time unstructured road detection based on CNN and Gibbs energy function

2021 ◽  
pp. 1-19
Author(s):  
Mingzhou Liu ◽  
Xin Xu ◽  
Jing Hu ◽  
Qiannan Jiang
Keyword(s):  
Gibbs Energy ◽  
Energy Function ◽  
Gaussian Mixture Models ◽  
Texture Features ◽  
Gaussian Mixture ◽  
Gibbs Energy Function ◽  
Road Detection ◽  
The Road ◽  
Detection Algorithms ◽  
Random Sample Consensus

Road detection algorithms with high robustness as well as timeliness are the basis for developing intelligent assisted driving systems. To improve the robustness as well as the timeliness of unstructured road detection, a new algorithm is proposed in this paper. First, for the first frame in the video, the homography matrix H is estimated based on the improved random sample consensus (RANSAC) algorithm for different regions in the image, and the features of H are automatically extracted using convolutional neural network (CNN), which in turn enables road detection. Secondly, in order to improve the rate of subsequent similar frame detection, the color as well as texture features of the road are extracted from the detection results of the first frame, and the corresponding Gaussian mixture models (GMMs) are constructed based on Orchard-Bouman, and then the Gibbs energy function is used to achieve road detection in subsequent frames. Finally, the above algorithm is verified in a real unstructured road scene, and the experimental results show that the algorithm is 98.4% accurate and can process 58 frames per second with 1024×960 pixels.

scholarly journals Deep-Learning-Based Road Crack Detection Frameworks for Dashcam-captured Images under Different Illumination Conditions

2021 ◽  
Author(s):  
Da-Ren Chen ◽  
Wei-Min Chiu
Keyword(s):  
Object Detection ◽  
Large Scale ◽  
Crack Detection ◽  
State Of The Art ◽  
Gaussian Mixture Models ◽  
Gaussian Mixture ◽  
Machine Learning Techniques ◽  
Detection Accuracy ◽  
The Road ◽  
Art Object

Abstract Machine learning techniques have been used to increase detection accuracy of cracks in road surfaces. Most studies failed to consider variable illumination conditions on the target of interest (ToI), and only focus on detecting the presence or absence of road cracks. This paper proposes a new road crack detection method, IlumiCrack, which integrates Gaussian mixture models (GMM) and object detection CNN models. This work provides the following contributions: 1) For the first time, a large-scale road crack image dataset with a range of illumination conditions (e.g., day and night) is prepared using a dashcam. 2) Based on GMM, experimental evaluations on 2 to 4 levels of brightness are conducted for optimal classification. 3) the IlumiCrack framework is used to integrate state-of-the-art object detecting methods with CNN to classify the road crack images into eight types with high accuracy. Experimental results show that IlumiCrack outperforms the state-of-the-art R-CNN object detection frameworks.

scholarly journals Fusing Appearance and Prior Cues for Road Detection

2019 ◽  
Vol 9 (5) ◽  
pp. 996
Author(s):  
Fenglei Ren ◽  
Xin He ◽  
Zhonghui Wei ◽  
Lei Zhang ◽  
Jiawei He ◽  
...  
Keyword(s):  
Pedestrian Detection ◽  
Autonomous Driving ◽  
Relevant Information ◽  
Gaussian Mixture ◽  
Morphological Processing ◽  
Road Detection ◽  
Collision Warning ◽  
The Road ◽  
Illuminant Invariant ◽  
Institute Of Technology

Road detection is a crucial research topic in computer vision, especially in the framework of autonomous driving and driver assistance. Moreover, it is an invaluable step for other tasks such as collision warning, vehicle detection, and pedestrian detection. Nevertheless, road detection remains challenging due to the presence of continuously changing backgrounds, varying illumination (shadows and highlights), variability of road appearance (size, shape, and color), and differently shaped objects (lane markings, vehicles, and pedestrians). In this paper, we propose an algorithm fusing appearance and prior cues for road detection. Firstly, input images are preprocessed by simple linear iterative clustering (SLIC), morphological processing, and illuminant invariant transformation to get superpixels and remove lane markings, shadows, and highlights. Then, we design a novel seed superpixels selection method and model appearance cues using the Gaussian mixture model with the selected seed superpixels. Next, we propose to construct a road geometric prior model offline, which can provide statistical descriptions and relevant information to infer the location of the road surface. Finally, a Bayesian framework is used to fuse appearance and prior cues. Experiments are carried out on the Karlsruhe Institute of Technology and Toyota Technological Institute (KITTI) road benchmark where the proposed algorithm shows compelling performance and achieves state-of-the-art results among the model-based methods.

The Gibbs Energy Description of Nd:YAG System

2012 ◽  
Vol 512-515 ◽  
pp. 479-483 ◽  
Author(s):  
Xiao Jin Cui ◽  
Wei Pan
Keyword(s):  
Complex System ◽  
Gibbs Energy ◽  
Direct Effect ◽  
Energy Function ◽  
Yttrium Aluminum Garnet ◽  
Thermochemical Data ◽  
Gibbs Energy Function ◽  
Sublattice Model ◽  
Yttrium Aluminum ◽  
Compounds Stability

Sublattice model is used to describe the complex system, Neodymium-doped yttrium aluminum garnet (Nd:YAG), and then Gibbs energy function of Nd:YAG system is seriously evaluated, mainly focusing on the instability of NAG, which has a direct effect on the Gibbs energy of NAG as well as Nd:YAG. Incorporating the selected literature thermochemical data of Nd:YAG system with the reevaluated parameters in Gibbs energy function according to a method utilized for defining the instability of NAG, the Gibbs energy function is well described. Trying to be more convincible, the method utilized for defining compounds stability and reevaluating the parameters has been test in Al2O3-Nd2O3, Al2O3-Y2O3, Al2O3-Gd2O3, Al2O3-Sm2O3 systems, achieving a satisfying agreement.

B2 Phases and their Defect Structures: Part I. Ab Initio Enthalpy of Formation and Enthalpy of Mixing in the Al-Ni-Pt-Ru System

2004 ◽  
Vol 842 ◽  
Author(s):  
Sara Prins ◽  
Raymundo Arroyave ◽  
Chao Jiang ◽  
Zi-Kui Liu
Keyword(s):  
Gibbs Energy ◽  
Energy Function ◽  
Correlation Functions ◽  
Enthalpy Of Mixing ◽  
Enthalpies Of Formation ◽  
Gibbs Energy Function ◽  
First Principle ◽  
Defect Structures ◽  
Sublattice Model ◽  
Local Pair

ABSTRACTThe enthalpies of formation of the bcc phases in the Al-Ni-Pt-Ru system, particularly in the Al-Ru binary and Pt-Al-Ru ternary subsystems, were calculated by first principle methods. The enthalpies of formation for stoichiometric bcc-B2 phases have been calculated using both the GGA and LDA approximations, while the enthalpies of formation for B2 phases with large amounts of constitutional defects (both vacancies and anti-site atoms) were calculated using the Special Quasirandom Structures (SQS) approach. The enthalpies of mixing for the disordered bcc-A2 phases have also been calculated with SQS by mimicking the random bcc alloy with the local pair and multisite correlation functions. The calculated B2 lattice parameters for the different defect structures were compared with experimental results. These results are used as input values for the CALPHAD modified sublattice model to describe the A2/B2 phases with one Gibbs energy function.

scholarly journals Drivable Road Corridor Detection using Flood Fill Road Detection Algorithm

2019 ◽  
Vol 9 (2) ◽  
pp. 5011-5014
Keyword(s):  
Image Processing ◽  
Edge Detection ◽  
Detection Algorithm ◽  
Image Processing Technique ◽  
Road Detection ◽  
The Road ◽  
Detection Algorithms ◽  
Color Differences ◽  
On The Road ◽  
Flood Fill

Current image processing techniques for drivable road detection make use of lane markings. However, most roads lack lane markings which make such techniques obsolete. For such conditions, an image processing technique is required which identifies the boundaries of the road based on the color differences between the road and the surroundings. This paper proposes a flood fill road detection approach in which we first analyze a sample of the road and compute its RGB pixel distribution. The pixel range is used to detect the other road pixels in the image. Edge detection algorithms are then applied on the detected road to give road edge. It classifies the road on the basis of the visible differences between the road and its neighborhood. It allows for subtle color differences on the road surface, and unlike a color mask, due to the inherent growing nature of a flood fill algorithm, it does not detect neighborhood elements beyond the boundary having features similar to the road. This technique also manages to detect any obstructions on the road as opposed to other edge detection algorithms. We also propose methods to enable quick computation of otherwise expensive flood-fill algorithm. The method was tested on both marked and unmarked lanes and produced satisfying results for both images and videos.

Sign in / Sign up

Export Citation Format

Share Document