Apparent relief

Automatic extraction and vectorization of contour lines from color topographic maps is an important precursor to obtaining useful information for many vector based GIS applications. In this work, a novel hybridized algorithm is developed for reconstructing the extracted contour lines from color topographic map. The extraction of contour lines from a topographic map leads to broken contour lines due to inherent characteristics of the map, thus posing a challenging problem of identifying gaps and then filling them. This has been addressed by developing algorithms based on connected components, graph theory, Expectation Maximization (EM) and numerical methods. Our algorithm operates by isolating the segments of those contours which have gaps and achieves in reducing the complexity of the matching of such segments by employing the EM algorithm. We also present a new scheme of filling gaps present in thick contours without the application of thinning algorithms.

Bag-of-words based image representation is a successful approach for object recognition. Generally, the subsequent stages of the process: feature detection, feature description, vocabulary construction and image representation are performed independent of the intentioned object classes to be detected. In such a framework, it was found that the combination of different image cues, such as shape and color, often obtains below expected results. This paper presents a novel method for recognizing object categories when using multiple cues by separately processing the shape and color cues and combining them by modulating the shape features by category-specific color attention. Color is used to compute bottom-up and top-down attention maps. Subsequently, these color attention maps are used to modulate the weights of the shape features. In regions with higher attention shape features are given more weight than in regions with low attention. We compare our approach with existing methods that combine color and shape cues on five data sets containing varied importance of both cues, namely, Soccer (color predominance), Flower (color and shape parity), PASCAL VOC 2007 and 2009 (shape predominance) and Caltech-101 (color co-interference). The experiments clearly demonstrate that in all five data sets our proposed framework significantly outperforms existing methods for combining color and shape information.

Chinese wines of different wine styles and varieties have different microstructure, and their micrograph show floccule,stick and granule of variant shape and size. Chinese wines can be classification or graded by the micrographs. Shape and structure of wines' particles in microstructure is the most important feature for recognition and classification of wines. After denoising and segment of micrographs, we describe a set of structure and shape features that can be used to characterize these micrographs. We introduce Euler distance, total number of regions and the combination of shape feature of different regions to describe the structure and region shape of micrograph. Seven kinds total 22 features are selected.Then Chinese wine classification system based on micrograph using combination of shape and structure features and BPNN have been presented. We compare classification results for different choices of features, and find that the best results could be obtained using the combination of shape and structure features.

Person Re-Identification (Re-ID) is a challenging task with wide ranging applications in various fields. This paper presents a novel hand-crafted method for this issue, enhancing the state of the art ones in literature in two aspects. In contrast to most current studies that analyze texture features, we propose a discriminative and compact shape feature by applying Procrustes shape analysis. It not only retains shape distinctiveness of an individual sample, but also alleviates cross-view impacts. Furthermore, we combine the shape feature with some current popular texture features, namely LOMO and mid-level filters, so that the advantages of multiple clues can be jointly used. A score level fusion strategy is finally adopted to optimally integrate their credits. Evaluated on two public benchmarks, i.e. VIPeR and CUHK03, the proposed method achieves very competitive results, indicating its effectiveness in person Re-ID.

3D Reconstruction of Underground Cable Wells with Automatic Extraction of Point Cloud Contour Lines

As assyriologists have to handle tremendous amounts of important documents of ancient history written in cuneiform script, like the epic of Gilgamesh, we are developing an efficent system to support their daily tasks. Previous projects demonstrated the application of holography and early close-range 3D scanners for this task. Based on experiences from our previous projects in archaeology, we are focusing on processing the vast amounts of data of high resolution 3D models from todays close-range 3D scanners like the Breuckmann smartSCAN-3D-HE. The presented method exploits the high-resolution of the 3D models to extract the impressed characters as well as other features like fingerprints. Previous work typically used rendering techniques from computer graphics to visualize the characters, which then had to be processed manually. More recent approaches use methods from differential geometry for detection and extraction of coarse contour lines. These methods are computationally fast, and well-established for industrial application, but cannot cover the variations of human handwriting in form of the – wedge shaped – cuneiform script. To overcome the variations in size of the wedges, we choose a multiscale approach using integrating geometry. A transformation invariant function is achieved by calculating the volumes of multiple concentric spheres intersecting the volume below the 3D model's surface at each point. Due to this multiscale approach, this function is represented by the so-called feature vector. By classifying these feature vectors using auto-correlation, our system – called GigaMesh – can automatically extract characters, requiring only one parameter: the approximated line (wedge) width in mm. Results are shown for cuneiform tablets from the collections of the Assyriologie Heidelberg as well as from the Uruk-Warka Sammlung. Finally an outlook regarding character (en)coding and integration into related projects like the Cuneiform Digital Library Initiative (CDLI) is given.

A distributed approach to contour line extraction in a multi-modal field using wireless sensor networks is proposed in this work. In contrast with previous work on edge or boundary estimation, which is primarily concerned with the region of a certain phenomenon, contour lines offer more detailed information about the underlying phenomenon. To effectively utilize the limited resources of sensors, a cluster-based data fusion technique is adopted so that the phenomenon of interest can be monitored in the base station (BS) without demanding excessive raw data transmission. Simulation results are given to demonstrate the efficiency of the proposed approach. Furthermore, some analysis is conducted to understand the impact of the sensor density and noise on the performance of the proposed approach.

Inverse synthetic aperture radar (ISAR) technique provides high‐resolution images of observed moving objects. However, shadow‐occluded effects and noise, which results in lack of fractions of the object geometry, often corrupt ISAR images. To facilitate the image analysis and to solve the target identification problem in most practical applications additional image‐processing procedures are strongly required. The aim of the present research is based on ISAR geometry and signal modelling to develop procedures improving quality of images and extract the silhouette contour line of the ISAR image, the main geometric parameter in the target recognition process. The problem is solved in two main stages. First, ISAR geometric parameters are described and expressions for linear frequency modulated signal modelling and imaging are derived. Second, image‐processing algorithms for ISAR image improving and contour line extraction are developed. All stages of image‐processing and reconstructing algorithms are illustrated by results of numerical experiments.

This paper deals with the automatic connection of contour lines extracted from a scanned brown geographical map. For the variety of topographical elements contained on a map, the factors causing the interruption of contour line are also multiform, which make the connection task very difficult. On the basis of separating those elements always making the contours break and regarding them as referent layers, a layering-based method is presented. The purpose is to take into account property information (like inclination and configuration) of contour lines when they come across other different symbols, such as gully, cliff, dry land and elevation annotation etc. In this paper, the authors propose that it should be far more effective and direct to adopt different algorithmic operators to different factors than using single one operator to all.

The image recognition and identification procedures are comparatively new in the scope of ISAR (Inverse Synthetic Aperture Radar) applications and based on specific defects in ISAR images, e.g., missing pixels and parts of the image induced by target’s aspect angles require preliminary image processing before identification. The present paper deals with ISAR image enhancement algorithms and neural network architecture for image recognition and target identification. First, stages of the image processing algorithms intended for image improving and contour line extraction are discussed. Second, an algorithm for target recognition is developed based on neural network architecture. Two Learning Vector Quantization (LVQ) neural networks are constructed in Matlab program environment. A training algorithm by teacher is applied. Final identification decision strategy is developed. Results of numerical experiments are presented.

In this work, we employ multi-layer feed forward network algorithm for leaf recognition using its shape, color and vein features. The leaf images are pre-processed and segmented. The shape features like area, convex area, diameter, length, width, perimeter, eccentricity, solidity, major axis length and Minor axis length are extracted by taking the features from the segmented leaf image. The mean, standard deviation, skewness and kurtosis for red, green and blue color features are extracted. Wiener filtering and canny edge detection are used to identify the vein features v1, v2, v3, v4 at different thresholds by calculating gray level value from the gray level histogram. The features are then exposed to Principal Component Analysis (PCA) for feature reduction and the resultant reduced feature set contains five shape features, six color features and three vein features. The features are then taken as input vectors for multi-layer neural network. Total 150 leaves from Columbia leaf image database are taken as samples. 150 leaves represent 10 different leaf species. The system was trained with 104 leaf images and was tested and validated using 23 leaf images each. The success rate of reduced feature set of leaf shapes and color was 84.7% and by combining leaf shape, color and vein features outputted a success rate of 95.3%.

Recently, intelligent transport system (ITS) is actively developed. There are big development areas in it. Within those areas, the research on detection and recognition of the road sign is one of the most important development areas. In this paper, we show the process improvement in order to extract the contour line and recognize the road sign shape from it. We have proposed the new extraction method of the contour line and internal areas from the road sign instantaneously by using the self organization map (SOM)

In this paper, a set of shape features which was based on the special characteristics of Chinese signatures was proposed for off-line Chinese signature verification. Another set of high-pressure features was also developed to extract timing and dynamic information indirectly from the signature image. A database of 100 genuine signatures and 50 simple forgeries was used to investigate the effectiveness of these two sets. Though shape features and high-pressure features had an error rate of 6.5% and 20.0% respectively, the collective use of these two sets reduced the error rate of simple forgery verification to 2.5%. It was concluded that combined use of static and dynamic information is an effective way for off-line Chinese signature verification.

We present a rendering algorithm for the forward projection of volume deformations. Spatial deformations are modeled by allowing the user to define curve skeletons inside the target object. Our rendering method utilises modern consumer graphics hardware and is integrated into an easy to use interface. In addition, we give an analysis of spatial transfer functions computed on the GPU, and give a solution to the problem of cracks appearing in image space when samples are pulled apart.

According to changeability of cloud, cloud-type recognition was primarily based on single-class feature in previous papers which was restricted to a certain degree. A set of features describing the color, texture as well as the shape features were extracted, then the shape and texture features combination methods were discussed. Here Gray-level co-occurrence matrix(GLCM) and Gabor wavelet transform based texture features and Zernike moment based shape features were combined, then support vector machine (SVM) was employed to recognize cloud-type. Experimental results showed that the correct recognition rates of altocumulus, cirrus, clear, cumulus and stratus were improved significantly, with the average recognition rate of 88.6%, and clear sky and stratus’s recognition rate of 100%.KeywordsFeature ExtractionFeature CombinationsCloud-Type RecognitionSupport Vector Machine (SVM)