Advanced Materials Research Vols. 760-762

Abstract: In this paper, a new micro-motion ISAR imaging algorithm is proposed, which achieve high cross range resolution by using the changes of azimuth angle caused by micro-motion. Before Wigner-Ville transformation to eliminate the cross-term interference, the echoes of the same range cell are decomposed. Comparing to traditional the Wigner-Ville imaging algorithm and Range-Doppler algorithm, this new algorithm in this paper has better imaging precision and no cross-term interference is existed. At last, the effectivity of the proposed algorithm is demonstrated by the simulated results.

Abstract: SAR image recognition is an important content of of aviation image interpretation work. In this paper, the characteristics of SAR images a practical significance of morphological filtering neural network model and its adaptive BP learning algorithm. As can be seen through the experimental results, the algorithm can not only adapt to the complex and diverse background environment, and has a displacement of the same continuous moving target detection capability, telescopic invariant and rotation invariant features.

Abstract: Wheat diseases image noise was effectively removed using lifting scheme multi-wavelet transform and multi-fractal analysis, and then it used multi-fractal theory to segment diseases image and extract eight multi-fractal spectrum values as wheat shape feature of diseases. Experiments showed that the shape characteristic value of different wheat diseases had great difference, and the shape characteristic value of similar diseases had certain regularity. Therefore, it could extract shape characteristic value to recognize wheat diseases.

Abstract: Histogram equalization (HE) algorithm is wildly used method in image processing of contrast adjustment using images histogram. This method is useful in images with backgrounds and foreground that are both bright or both dark. But the performance of HE is not satisfactory to images with backgrounds and foregrounds that are both bright or both dark. To deal with the above problem, [ gives an improved histogram equalization algorithm named self-adaptive image histogram equalization (SIHE) algorithm. Its main idea is to extend the gray level of the image which firstly be processed by the classical histogram equalization algorithm. This paper gives detailed introduction to SIHE and analyzes the shortage of it, then give an improved version of SIHE named ISIHE, finally do experiments to show the performance of our algorithm.

Abstract: This paper studies on the path recognition system of intelligent vehicles, and makes classification and feature extraction of road. By using the BP neural network it makes the analog simulation, and through the training of network structure, it tries to establish network structure which can reflect the input road sample and the specified road type mappings, and then to realize the automatic identification of road type.

Abstract: Tooth flank pitting and gluing are principal forms of gear defect. The purpose of this research is to extract the image feature of the gear in the different defects by means of image processing technology. Firstly, the image was carried out denoising processing by median filtering and segmentation processing by use of OSTU method. Then, the pixel area was extracted as a feature to distinguish normal gear, tooth surface pitting and gluing, the inertia was extracted as image feature to detect pitting and gluing by Gray level co-occurrence matrix, and the morphological characteristics of the image were extracted. Image feature extraction of different defect form will help to establish an effective image recognition model.

Abstract: The pathological change of lymph node is an important basis of malignant tumor detection and judgment of metastasis of cancer (lung cancer, colorectal cancer, breast cancer, liver cancer, cervical cancer, etc.) An algorithm of lymph node image segmentation based on improved FCM clustering and multi-threshold is proposed to segment the lymph CT image with blurred edge. First, the improved FCM peak clustering is used to sharpen the fuzzy boundary of lymph CT image effectively. Then the multi-threshold algorithm based on image entropy change is introduced to segment enhanced images. The experiment shows that the above algorithm can obtain better segmentation results compared with the traditional FCM clustering method in the case of the fuzzy edge of the lymph node tissue.

Abstract: The classical bilateral filtering algorithm is a non-linear and non-iterative image denoising method in spatial domain which utilizes the spatial information and the intensity information between a point and its neighbors to smooth the noisy images while preserving edges well. To further improve the image denoising performance, a spatially adaptive bilateral filtering image deonoising algorithm with low computational complexity is proposed. The proposed algorithm takes advantage of the local statistics characteristic of the image signal to better preserve the edges or textures while suppressing the noise. Experiment results show that the proposed image denoising algorithm achieves better performance than the classical bilateral filtering image denoising method.

Abstract: Edge detection is a fundamental problem in computer vision. In this paper, we present an effective algorithm to find salient edges from infrared scene images based on Human Visual System. The algorithm integrates three basic edge features: edge contrast, edge density and edge length. In this manner, the proposed algorithm works well to detect salient region boundaries and to suppress false edges from background and texture. The experimental results demonstrate the effectiveness of the proposed algorithm.

Abstract: Data fusion technique can produce fused images with high spatial resolution and abundant spectral information. A new image fusion algorithm based on two-dimension PCA and Curvelet transform will be proposed according to image process models specialities in this paper. First of all, we performed 2DPCA on the MS image to get the 1st principle component (PC1); then we applied Curvelet transform in Pan Image and PC1; lastly decomposition coefficients obtained was processed according to certain rules to get fused coefficients, and afterwards, we performed inverse Curvelet transform on them to acquire fused sub-images. Then we performed inverse 2DPCA transform on the other components and the fused sub-images to get fused images. Experiments will be carried out via application of multispectral and panchromatic images, and it turns out that this new algorithm can improve spatial resolution greatly while maintaining spectral information.