How to improve the probability of registration and precision of localization is a hard problem, which is desiderated to solve. The two basic approaches (normalized cross-correlation and phase correlation) for image re...
详细信息
How to improve the probability of registration and precision of localization is a hard problem, which is desiderated to solve. The two basic approaches (normalized cross-correlation and phase correlation) for image registration are analysed, two improved approaches based on spatial-temporal relationship are presented. This method adds the correlation matrix according to the displacements in x- cirection and y- directions, and the registration pose is searched in the added matrix. The method overcomes the shortcoming that the probability of registration decreasing with area increasing owing to geometric distortion, improves the probability and the robustness of registration.
A hierarchical retrieval scheme of the accessory image database is proposed based on textile industrial accessory contour feature and region feature. At first smallest enclosed rectangle[1] feature (degree of accessor...
详细信息
A hierarchical retrieval scheme of the accessory image database is proposed based on textile industrial accessory contour feature and region feature. At first smallest enclosed rectangle[1] feature (degree of accessory coordination) is used to filter the image database to decouple the image search scope. After the accessory contour information and region information are extracted, the fusion multi-feature of the centroid distance Fourier descriptor and distance distribution histogram is adopted to finish image retrieval accurately. All the features above are invariable under translation, scaling and rotation. Results from the test on the image database including 1,000 accessory images demonstrate that the method is effective and practical with high accuracy and fast speed.
To solve the heterogeneous image scene matching problem, a non-linear pre-processing method for the original images before intensity-based correlation is proposed. The result shows that the proper matching probability...
详细信息
To solve the heterogeneous image scene matching problem, a non-linear pre-processing method for the original images before intensity-based correlation is proposed. The result shows that the proper matching probability is raised greatly. Especially for the low S/N image pairs, the effect is more remarkable.
patternrecognition methods were used to treat the experimentally measured data of Pitzer’s coefficients of 107 electrolytes, to find the relationships between the ionic structural parameters of these electrolytes an...
详细信息
patternrecognition methods were used to treat the experimentally measured data of Pitzer’s coefficients of 107 electrolytes, to find the relationships between the ionic structural parameters of these electrolytes and Pitzer’s coefficients. It is found that these relationships can be approximately expressed as linear equations of four dimensionless numbers, (R_+/R_-), (R_++R_-)/Z_+Z_-, (Z_+/Z_-) and (R_t/R_l), where R_+ and R_- are the cationic and anionic radii respectively; Z_+ and Z_- are the cationic and anionic charge numbers respectively, and (R_t/R_l) denotes the nonsphericity of some non-spherical ions. Besides, it is found that the difference of the nuclear magnetic resonance measured rotational relaxation time of water molecules around cations and anions, |Δτ|, has good correlation with Pitzer’s coefficients. The relationships can be interpreted by the theory of corresponding states of ionic solutions. Based on the relationships, an example of application to some hydrometallurgical process was discussed.
To modeling and classify underwater sound, hidden Markov tree (HMT) model in wavelet domain is adopted. Taking advantage of the models, the simulation time sequence of ocean noise can be produced. An improved classifi...
详细信息
An improved approach for JSEG is presented for unsupervised color image segmentation. Instead of color quantization algorithm, an automatic classification method based on adaptive mean shift (AMS) based clustering is ...
详细信息
Since feature selection can remove the irrelevant features and improve the performance of learning systems, it is an crucial step in machine learning. The feature selection methods using support vector machines have o...
详细信息
Microscopic halftone imagerecognition and analysis can provide quantitative evidence for printing quality control and fault diagnosis of printing devices, while halftone image segmentation is one of the significant s...
详细信息
Microscopic halftone imagerecognition and analysis can provide quantitative evidence for printing quality control and fault diagnosis of printing devices, while halftone image segmentation is one of the significant steps during the procedure. Automatic segmentation on microscopic dots by the aid of the Fuzzy C-Means (FCM) method that takes account of the fuzziness of halftone image and utilizes its color information adequately is realized. Then some examples show the technique effective and simple with better performance of noise immunity than some usual methods. In addition, the segmentation results obtained by the FCM in different color spaces are compared, which indicates that the method using the FCM in the f 1f 2f 3 color space is superior to the rest.
The mean shift algorithm is an efficient method for tracking object in the color image sequence. However, in the infrared object-tracking scenario, there is a singular feature space, i.e. the grey space, for represent...
详细信息
ICA (independent component analysis) feature extraction is an efficient sparse coding method for noise reduction. In many ICA-based de-noising processing, however, they need noise-free source data to train the basis v...
详细信息
暂无评论