Applied Mechanics and Materials Vols. 511-512

Abstract: In view of complex background of cotton blind stinkbug hazard region and the difficulty in segmentation and classification under natural conditions, an automatic classification method of cotton blind stinkbug hazard level was proposed. In this method, crop regions and disease regions of cotton were extract respectively by H+a*+b* component and Otsu segmentation method based on blind stinkbug hazard cotton leaves. Adhesion cotton leaves separated by Watershed segmentation method and cotton leaf area hazard by blind stinkbug extracted. According to cotton blind the stinkbug hazard rating standard, combination Naive Bayes classifier and color, texture and shape features extracted from images to classify the hazard rating of the blind stinkbug. The results showed that the model classification correct rate was 90.0%, it could classify the hazard rating of the cotton blind stinkbug and provide technical support for the prevention and treatment of the cotton blind stinkbug.

Abstract: This paper proposed a new image enhancement algorithm based on edge sharpening of wavelet coefficients for stereoscopic images. Our scheme uses the multi-scale characteristic of wavelet transform, decomposes the original image into low frequency approximation sub-graph and several high frequency direction. Under the multi-scale, the low frequency approximation sub-graph is processed by edge sharpening method. Then the low frequency sub-graph decomposes in multi-scale again. At last, the low frequency approximation graph after four layers decompose sharpening and the high frequency approximation of the decomposed sub-graph will be refactored to get the new image. Experimental results show that whether PSNR or visual effect, or the subjective assessment of the DMOS value, the proposed method has better enhanced performance than the conventional edge sharpening and wavelet transform. And it has good image edge enhancement, details protection. Meanwhile, the proposed algorithm has the same computational complexity with wavelet transform.

Abstract: Due to the large distance between transmitting array and receiving array in bistatic MIMO, The bistatic MIMO is unable to use the waveform diversity technology base on the monostatic MIMO to improve the resolution of the detection. And bistatic MIMO angle estimation algorithm based on the spatial spectral estimation is unable to get coherent gain from beamforming. In this paper, position calculating of the bistatic sonar is used to get the relationship among DOD DOA and propagation distance. The array manifold of virtual array is constructed in each scan position in 2-d imaging. The improved waveform diversity technology is used in the imaging of bistatic MIMO to improve the resolution of the detection.

Abstract: The eutrophication phenomenon for lake in our country is obvious. Algal blooms have become one of the major environmental issues in the lake. In view of the traditional single point monitoring and analysis limitations, this paper proposes a bloom prediction model based on remote sensing information. Different light wavelengths has different sensitivity of chlorophyll, so extracting the reflectance values to the chlorophyll-sensitive band from satellite remote sensing information, Building water green index information and corresponding to chlorophyll values of lake-sites monitoring area by latitude and longitude geographic information , thereby build a predictive model of algal bloom based on a remote sensing, realize effective prediction of algae bloom in Taihu lake for coastal ,open ocean water and lake reservoir bloom prediction gives an idea.

Abstract: Because of the interference of the noise signals, the actual feature of objects cant be reflected clearly by photo which is measured by optical measurement. Thus, the database of feature of geometric model is established at the platform of Pro/E. Then the information selected will be extracted and exported, and compared to the photo which is gotten from the optical instrument. This can identify the information to prove the correctness when detecting objects photographed, which is used to adjust to the large structure.

Abstract: Multi-scale image segmentation is a kind of effective means to realize the multi-scale expression of geographical entity features. Describing regions of segmented image and their spatial relation is important for multi-scale image segmentation. The paper focuses on irregular pyramid structure based on topology map to research the equivalence of topological relations on different scales in multi-scale image segmentation, which is important for effectively updating the segmentation process.

Abstract: Seam tracking technology is an important area of research automatic arc welding, precise seam tracking is crucial to achieve high quality welds. In order to achieve precise seam tracking, seam deviation ( the weld center arc deviation) detection is a key. Unlike the conventional method by image processing techniques to obtain the seam deviation information directly, but selected image processing area (including the distal end portion of the molten pool welds and the front end of the pool), and analyzed as a pool image centroid characteristic parameters of the weld deviation. Study these parameters to create a new method for visual weld deviation measurement model, establish the linear regression model between pool image centroid deviation and the weld based on regression analysis theory.

Abstract: Using machine vision technology to realize the real-time automatic collection of test parameters including light and color of LED lamps during the experiments. The failure mechanisms were researched and the reference failure mechanisms were given according to the parameters change situation of the failure lamps, in order to implement rapid and automatic detection of test conditions of LED lamps, which can accelerate the research processes of LED lamps and related products, and promote the industrialization process of LED. The fast automatic failure detection system for LED lamps based on machine vision technology proposed in this paper had great significance for LED lamps research organizations such as testing organizations, research institutes, as well as manufacturers, and it was worthy of learning for them.

Abstract: The surveying achievement is the basis for managing the urban real estate, the currency and accuracy of which are the two preconditions to increase efficiency for management of real estate. Based on the validity of surveying data in urban real estate, this paper is to explore the overall method and mechanism for the change of surveying achievements. With the practical experience, operable measures could be proposed for increasing the management ability of urban real estate in China.

Abstract: Since most of the current video surveillance systems are static or dynamic track based on video detection, the paper puts forward the scheme of sound source localization and camera dynamic acquisition in the intelligent video surveillance system. The microphone array is used to locate the sound source, then the camera is steered to monitor sound source position and alarm is executed by image analysis. The experimental results show that the system can achieve effective positioning of the sound source and alarm at abnormal condition under a low-niose environment.