Advanced Materials Research Vols. 1006-1007

DOI:

https://doi.org/10.4028/www.scientific.net/AMR.1006-1007

Export:

MARCXML

ToC:

Table of Contents

Paper Title Page

Characteristics Detection Method for Material Surface Based on Polarization Vision Information

Authors: Rong Hua Li, Bao Zhang Qu, Feng Xiang Zhai, Ying Zhou

Abstract: Comparing to ordinary light intensity measurements, polarization measurements have many advantages in studying the surface，atmosphere and exploring space. In order to obtain high-quality polarization information, the paper designs a three-channel synchronous acquisition device. It researches the feasibility of obtaining the characteristics information of material surface by polarization vision. It respectively measures the leaves, automobiles, cement wall these three samples’ reflected light’s polarization characteristics using the method of getting Stokes parameters. The experimental results show that the polarization characteristics have a great relationship with the characteristics of target, the measurement wavelength and the viewing angle. The method mentioned in this paper is vital to achieving material surface characteristics. At the same time , it provides a new way for geology and mineral resources exploration.

747

Review on Application of Near-Infrared Spectroscopy Identification Technology in Sorting Impurities from Waste Paper

Authors: Hai Ming Xin, Dong Ri Shan, Cheng Min Gao

Abstract: With rapid increase of waste paper recycling, more and more researchers focus on the advanced technologies of sorting impurities form waste paper. The principals of main identification and sorting technologies are introduced in this paper, whose advantages and disadvantages are discussed. Because of importance of sorting technologies based on near-infrared spectroscopy, we emphasis its principle, application and research status etc.. At last, research and development tendency of identification and sorting technology are prospected.

752

Researches on Segmentation of Overlapped Convex Targets

Authors: Yun Zhao, Ming Zhu Zhu, Ying Shen

Abstract: Situation of multiple overlapped targets usually encountered when image processing technology was applied in counting and analysis. Effect of the segmentation directly affects the accuracy of counting and measurement. Three basic methods include concave points matching, distance transform and Hough Transform were introduced and compared in this paper.

756

Scale Distorted Target Recognition Based on Machine Vision

Authors: Yu Chen, Fu Rong Huo, Li Qin Zheng

Abstract:

Machine vision means to carry out measurement or judgment with machine instead of human eyes. In the field of target recognition, optical correlation technology is a main way to realize machine vision. Targets can be recognized and located with high precision taking advantage of optoelectronic hybrid joint transform correlator (OHJTC). However, when scale or angular distortion of the detected target exists relative to the reference template, the intensity of correlation peaks will decrease to a great extent, which restricted the recognition results greatly. In this paper, the development and principle of maximum average correlation height (MACH) algorithm is introduced. Through amounts of experiments, the controlling parameters of the synthesized filter are optimized, which makes MACH filter suppress background noise and widen recognition range of targets. To show the feasibility of this algorithm, simulative and optical experiments of the improved MACH filter are carried out. As an example, the recognition results of a fighter target in sky are given, which shows the scale distortion tolerance can reach up to ±23%. The actual effect of the improved MACH filter algorithm has been confirmed very well.

760

Vector Quantization by Minimizing Kullback-Leibler Divergence between the Class Label Distributions over Quantization Input and Output

Authors: Hao Xiang Wang, Shan Yue, Yang Li

Abstract: This paper proposes a new method for vector quantization by minimizing the Divergence of Kullback-Leibler between the class label distributions over the quantization inputs, which are original vectors, and the output, which is the quantization subsets of the vector set. In this way, the vector quantization output can keep as much information of the class label as possible. An objective function is constructed and we developed an iterative algorithm to minimize it as well. The novel method is evaluated on bag-of-features based image classification problems.

764

An Image Quality Assessment Approach Based on Saliency Map in Space Domain

Authors: Bin Wang

Abstract: This paper proposes a no-reference image assessment approach (IQA) based on saliency map in the space domain of the image. The saliency map of the image is extracted by Itti model at first. Next, the saliency-map weighted normalized image is used to get the histogram of the image, then the histogram is modeled by generalized gaussian distribution and the parameter of the generalized gaussian distribution is estimated by parameter estimating approach. Parameters of the generalized gaussian distribution are used as the feature vector for the training and testing image. The feature vectors of the testing image are fed to support vector regresion machine to evaluate the image quality score. Experimental results show that our approach outperforms the recent method of no-reference IQA.

768

Research on Surface Defect Detection Technique of Rolling Element Based on Computer Vision

Authors: Chuan Ren, Xiao Yu Xiu, Guo Hui Zhou

Abstract: This paper proposed a new method of surface defect detection of rolling element based on computer vision, which adopted CCD digital camera as image sensor, and used digital image processing techniques to defect the surface defects of rolling element. The main steps include collect image, use an improved median filter to reduce the noise, increase or decrease the exposure to achieve the image enhancement, create a binary image with threshold method and detect the edge of the image, and use subtraction method for surface defects identification. The experiment indicates that the above methods the advantages of simple, the capability of noise resistance, high speed processing and better real-time.

773

A Low Cost Scheme of Scene Depth Map for Stereoscopic Video Streaming Using Single Kinect Sensing

Authors: Din Yuen Chan, Chi Wen Hsieh, Tai Lang Jong

Abstract: This study mainly described a low-cost stereoscopic sensing scheme using a single Kinect device of Microsoft. Basically, the integration of infrared (IR) depth map with visible light image of Kinect could support the virtual 3-dimensional view. However, the precision and unstable displays are problems practically that means the unstable variation of 3D displays and distorted projections of Kinect IR projector make the depth maps non-reliable so that the virtual view usually non-acceptable with defects in many cases especially in low lamination and corners of the sensing objects. The study proposed an algorithm to rebuild a good depth map includes: free-holes filling, extrapolation and prediction of filling-loss holes, rectification and trimming of depth-edge deformation, modification of the depths for no-IR-detectable object and the sparse IR-detectable depths, and finally applying the bilateral filtering for the edge smoothing. The simulation result demonstrated that the proposed algorithm can provide an acceptable virtual 3-dimensional images and the calculation loads are suitable for real-time stereoscopic displays.

779

The Study and Experiment of ERT Imaging System

Authors: Zi Peng Chen, Peng Li, Xi Fang, Zhuo Qiu Li

Abstract: The Electrical Resistance Tomography (ERT) is a new technology developed in recent years. Because of the characteristics as simple structure, low cost and non-invasive, the ERT has already application in geophysics, physical geography, mineral detection et al.in this paper, the ERT is used to detect the damage of the remote carbon fiber felt with wireless sensing technology. In experiment, the wireless technology is used to avoid the noise and the introduction of the long wires, which realized the remote detection; the Improved Newton -- Raphson algorithm is used to achieve the detective imaging, which had good results.

783

Model Optimization and Testing on Real-Time Video Moving Target Detection

Authors: Shu Ling Zhang, Zhi Hong Zhang

Abstract: Zhang presented a statistical model of real-time video moving target detection based on Bayesian statistical theory. This article discusses the algorithm parameter selection and detection efficiency of the model by using the experimental simulation method. This article generates a reference background based on unsupervised learning methods, and uses a color space that has a better environmental adaptability to represent the background, and uses dynamic threshold method to classify the results of background subtraction and frame difference. By comparing experimental of different methods, it shows that this algorithm has a greater advantage in terms of accuracy and timeliness.

787