Search

Crop identification is important to precise agriculture, traditional survey is laboring intensive and classic remote sensing image processing methods are highly dependent on expertise, aiming at this problem, a U-Net based deep segmentation network is proposed for crop identification in UAV remote sensing image segmentation. First, we preprocess the UAV remote images of dataset for training. Then, we replace the transposed convolutional layer in U-Net upsampling block with sub pixel convolutional layer, to solve the problem that edge structure information lost and the low segmentation accuracy. The experiment result shows that, compared with original U-Net and other state of art segmentation model, the proposed method can improve the edge smoothness and pixel wise accuracy. The method uses sub pixel convolution layer for upsampling can be applied to other segmentation model, and the way uses segmentation model to identify the crop species can be extended to other remote sensing image processing filed.

In this paper, a surface projection image stitching algorithm based on object coordinate system is proposed to generate stereoscopic image, which can be applied to virtual reality and vehicle assisted driving. According to the actual application, a suitable 3D model is established, and the image is mapped based on the object coordinate system. The 3D model is divided into fusion and non-fusion regions, and the original image is mapped to the non-fusion region. For the stitching region, combined with the model features, a weighted average fusion algorithm based on angle information is designed to correct the color of the stitched image to eliminate uneven illumination and brightness. The experimental results show that the stereoscopic image generated by this method can display the surrounding environment of the car body in a stereoscopic manner, with high image stitching accuracy and good visual effect.

A novel distributed fiber-optic sensor based on Wavelength Division Multiplex (WDM) for determining the position of disturbances is presented. The configuration and operating principle of the system is illustrated. The location principle and the method for the detection system are analyzed. Based on the initial method in the frequency domain, the system realizes the disturbances location using the WDM technology to compare the signals which are the same vibration corresponding to different light sources and light paths. Theory analysis and experiment results show that the proposed algorithm can realize the detection and location of the multipoint disturb signals rapidly and effectively, this method is simply and can be obtained easily, it could increase the length of fiber-optic sensors which has high measurement sensitivity and location precision.

Distributed fiber-optic vibration sensors receive extensive investigation and play a significant role in the sensor panorama. A fiber optic perimeter detection system based on all-fiber interferometric sensor is proposed, through the back-end analysis, processing and intelligent identification, which can distinguish effects of different intrusion activities. In this paper, an intrusion recognition based on the auditory selective attention mechanism is proposed. Firstly, considering the time-frequency of vibration, the spectrogram is calculated. Secondly, imitating the selective attention mechanism, the color, direction and brightness map of the spectrogram is computed. Based on these maps, the feature matrix is formed after normalization. The system could recognize the intrusion activities occurred along the perimeter sensors. Experiment results show that the proposed method for the perimeter is able to differentiate intrusion signals from ambient noises. What’s more, the recognition rate of the system is improved while deduced the false alarm rate, the approach is proved by large practical experiment and project.

This paper researches on the key and difficult issues in stereo measurement deeply, including camera calibration, feature extraction, stereo matching and depth computation, and then put forwards a novel matching method combined the seed region growing and SIFT feature matching. It first uses SIFT characteristics as matching criteria for feature points matching, and then takes the feature points as seed points for region growing to get better depth information. Experiments are conducted to validate the efficiency of the proposed method using standard matching graphs, and then the proposed method is applied to dimensional measurement of mechanical parts. The results show that the measurement error is less than 0.5mm for medium sized mechanical parts, which can meet the demands of precision measurement.

All-fiber interferometer sensor system is a new type of system, which could be used in long-distance, strong-EMI condition for monitoring and inspection. A fiber optic perimeter detection system based on all-fiber interferometric sensor is proposed, through the back-end analysis, processing and intelligent identification, which can distinguish effects of different intrusion activities. In this paper, the universal steps in triggering pattern recognition is introduced, which includes signal characteristics extracting by accurate endpoint detecting, templates establishing by training, and pattern matching. By training the samples acquired in the laboratory, this paper uses the wavelet transformation to decompose the detection signals of the intrusion activities into sub-signals in different frequency bands with multi-resolution analysis. Then extracts the features of the above mentioned intrusions signals by frequency band energy and wavelet information entropy and the system could recognize the intrusion activities occurred along the perimeter sensors. Experiment results show that the proposed method for the perimeter is able to differentiate intrusion signals from ambient noises such as windy and walk effectively. What’s more, the recognition rate of the system is improved while deduced the false alarm rate, the approach is proved by large practical experiment and project.

Distributed optic fiber sensor is a new type of system, which could be used in the long-distance and strong-EMI condition for monitoring and inspection. A method of external modulation with a phase modulator is proposed in this paper to improve the positioning accuracy of the disturbance in a distributed optic-fiber sensor. We construct distributed disturbance detecting system based on Michelson interferometer, and a phase modulator has been attached to the fiber sensor in front of the Faraday rotation mirror (FRM), to elevate the signal produced by interfering of the two lights reflected by the Faraday rotation Mirror to a high frequency, while other signals remain in the low frequency. Through a high pass filter and phase retrieve circus, a signal which is proportional to the external disturbance is acquired. The accuracy of disturbance positioning with this signal can be largely improved. The method is quite simple and easy to achieve. Theoretical analysis and experimental results show that, this method can effectively improve the positioning accuracy.

The multi-sensor image fusion technology can obtain a more comprehensive and more accurate and reliable image, in order to understand the scene or recognize the target more easily. However, most existing algorithms are mainly based on optical remote sensing images, which is highly susceptible by media interference, supplemented by SAR images. The image fusion between SAR images and PAN images also cannot save the textural feature and the color information effectively at the same time. In view of these problems, this paper presents a multi-sensor image fusion algorithm based on region-based selection and IHS transform. The SAR image and PAN image are firstly IHS transformed to achieve the intensity (I), hue (H) and saturation (S) weights. The I weights of SAR image and PAN image are separately decomposed using SIDWT algorithm to extract wavelet coefficients. Then, the I weight of SAR image is divided into regular area and irregular area based on a new adaptive segmentation method. A new fusion rules is presented according to local feature, and then used to fuse corresponding wavelet coefficients of the I weight of SAR image and PAN image. Inverse SIDWT is carried out on the fused wavelet coefficients to get the I weight (I’) of fused image. Finally, the fused image is obtained by inverse IHS transform of I’ weight with the H, S weight of PAN image. Experimental results of real images validated the effectiveness of the proposed algorithm by objective evaluation such as standard deviation, entropy, average gradient, etc.

Aimed to the limitation of present anomaly detection algorism under clutter background for multi-spectral remote sensing data, especially for the situations of dense spread target and exist different attributive of background objects, a bio-inspired anomaly detection algorithm was proposed. Simulate the information processing and fusion mechanism of fly multi-apertures vision system, multi-level background model was proposed to analysis and describe feature of clutter background. Then the threshold value can be chose adaptively according to the level of background model. The proposed algorithm didn’t need the prior knowledge about anomaly, and avoids the choosing of the background widow size. A fusion mechanism was proposed to fuse the different detection results with different level background model. Simulation experiment validated the effectiveness of proposed method.

With quick development of economy, spatial distribution and specialization level of China large scale commodity exchange markets whose turnover are more than 100 million Yuan, have changed greatly. And influencing factors which distribute in the research region have attribute information and spatial information and do not satisfy statistical independence. Commodity exchange market specialization index is brought forward to measure specialization degree, based on the former research and constrained co-local spatial association rule is used to analyze symbiotic pattern between specialization level and influencing factors. Constrained predicate templates and association rule templates can improve mining efficiency greatly. As the result shown, large scale commodity exchange market specialization level on country-region spatial scale went down from 2000 to 2005 and rose at 2006. The interesting association rules extracted based on defined minimum support and confident can provide officers of region governments with rational advices on large scale commodity exchange markets planning and construction.

Spatial distribution pattern is an arrangement of two or more spatial objects according to some spatial relations, such as spatial direction, topological and distance relations. In the real world, spatial objects and spatial distribution pattern all vary continuously along the time-line. Traditional spatial and non-spatial data dissevers this continuous spatio-temporal process. Under analyzing relations among spatial object, its attributes and spatial distribution pattern, we brought metaspatio- temporal process, spatio-temporal process and spatial distribution pattern spatio-temporal process. Rainfall in Eastern China has a typical spatial distribution pattern, being composed of the northern rain area and the southern rain area. Through constructing spatio-temporal process transactions, the association rules can be extracted from spatiotemporal process data set by the Apriori algorithm. The result of the spaio-temporal process association rule mining is consistent with the analysis of the theory. Finally, it is concluded that the spatio-temporal process can describe change of a spatial object in a defined time range, and change trend of one entity can be forecasted through varying trend of others based on the valuable spatio-temporal process association rules.

The objective of this study is to present a novel model for simulating the evolution of gradual change systems. The proposed model, named SDCA-state delay cellar automata, extends the traditional cellar automata (CA) to include state delay, variable time step, and state-delay cell evolving to aid prediction making for the gradual systems developments. The model can perform microcosmic changing happens in one time step through changing state-delay time. In this article, the SDCA model used to simulate the forest fire spreading which is a typical gradual changing system under different external conditions. As experiments' result shown, SDCA model can simulate the forest fire spreading and perform gradual changing with time step varied accurately and simply. Finally, it is concluded that the SDCA model is an effective model in simulating the gradual changing systems' propagation.