Oriented Feature Extraction Research Articles

Remote Sensing for Land Administration 2.0

Contemporary land administration (LA) systems incorporate the concepts of cadastre and land registration. Conceptually, LA is part of a global land management paradigm incorporating LA functions such as land value, land tenure, land development, and land use. The implementation of land-related policies integrated with well-maintained spatial information reflects the aim set by the United Nations to deliver tenure security for all (Sustainable Development Goal target 1.4, amongst many others). Innovative methods for data acquisition, processing, and maintaining spatial information are needed in response to the global challenges of urbanization and complex urban infrastructure. Current technological developments in remote sensing and geo-spatial information science provide enormous opportunities in this respect. Over the past decade, the increasing usage of unmanned aerial vehicles (UAVs), satellite and airborne-based acquisitions, as well as active remote sensing sensors such as LiDAR, resulted in high spatial, spectral, radiometric, and temporal resolution data. Moreover, significant progress has also been achieved in automatic image orientation, surface reconstruction, scene analysis, change detection, classification, and automatic feature extraction with the help of artificial intelligence, spatial statistics, and machine learning. These technology developments, applied to LA, are now being actively demonstrated, piloted, and scaled. This Special Issue hosts papers focusing on the usage and integration of emerging remote sensing techniques and their potential contribution to the LA domain.

A new key performance indicator oriented industrial process monitoring and operating performance assessment method based on improved Hessian locally linear embedding

ABSTRACT The industrial process monitoring and operating performance assessment techniques are of great significance to ensure the safety and efficiency of the production and to improve the comprehensive economic benefits for the modern enterprises. In this paper, a new key performance indicator (KPI) oriented nonlinear process monitoring and operating performance assessment method is proposed based on the improved Hessian locally linear embedding (HLLE), in view of the problems of strong nonlinearity, high dimension and information redundancy in actual industrial process data. Firstly, in order to characterise the similarities of samples in both temporal and spatial dimensions, a new measurement, based on Finite Markov theory, is defined to replace the Euclidean distance in traditional HLLE. Secondly, by mining the relationships between process variables and the key performance indicator, the KPI oriented feature extraction method is developed. On this basis, the monitoring statistics is constructed and the corresponding control limit is determined for the real-time fault detection. After that, a new operating performance assessment approach based on sliding window Kullback–Leibler divergence is put forward to facilitate maintenance or adjustments. Finally, the proposed method is applied to the hot strip mill process, and the results show the effectiveness.

An oriented feature extraction and recognition approach for concave-convex mixed interacting features in cast-then-machined parts

Feature recognition is an important technology of computer-aided design/computer-aided engineering/computer-aided process planning/computer-aided manufacturing integration in cast-then-machined part manufacturing. Graph-based approach is one of the most popular feature recognition methods; however, it cannot still solve concave-convex mixed interacting feature recognition problem, which is a common problem in feature recognition of cast-then-machined parts. In this study, an oriented feature extraction and recognition approach is proposed for concave-convex mixed interacting features. The method first extracts predefined features directionally according to the rules generated from attributed adjacency graphs–based feature library and peels off them from part model layer by layer. Sub-features in an interacting feature are associated via hints and organized as a feature tree. The time cost is reduced to less than [Formula: see text] by eliminating subgraph isomorphism and matching operations. Oriented feature extraction and recognition approach recognizes non-freeform-surface features directionally regardless of the part structure. Hence, its application scope can be extended to multiple kinds of non-freeform-surface parts by customizing. Based on our findings, implementations on prismatic, plate, fork, axlebox, linkage, and cast-then-machined parts prove that the proposed approach is applicable on non-freeform-surface parts and effectively recognize concave-convex mixed interacting feature in various mechanical parts.

An Event-Based Neurobiological Recognition System with Orientation Detector for Objects in Multiple Orientations.

A new multiple orientation event-based neurobiological recognition system is proposed by integrating recognition and tracking function in this paper, which is used for asynchronous address-event representation (AER) image sensors. The characteristic of this system has been enriched to recognize the objects in multiple orientations with only training samples moving in a single orientation. The system extracts multi-scale and multi-orientation line features inspired by models of the primate visual cortex. An orientation detector based on modified Gaussian blob tracking algorithm is introduced for object tracking and orientation detection. The orientation detector and feature extraction block work in simultaneous mode, without any increase in categorization time. An addresses lookup table (addresses LUT) is also presented to adjust the feature maps by addresses mapping and reordering, and they are categorized in the trained spiking neural network. This recognition system is evaluated with the MNIST dataset which have played important roles in the development of computer vision, and the accuracy is increased owing to the use of both ON and OFF events. AER data acquired by a dynamic vision senses (DVS) are also tested on the system, such as moving digits, pokers, and vehicles. The experimental results show that the proposed system can realize event-based multi-orientation recognition. The work presented in this paper makes a number of contributions to the event-based vision processing system for multi-orientation object recognition. It develops a new tracking-recognition architecture to feedforward categorization system and an address reorder approach to classify multi-orientation objects using event-based data. It provides a new way to recognize multiple orientation objects with only samples in single orientation.

Double-orientation code and nonlinear matching scheme for palmprint recognition

Many palmprint authentication approaches have been proposed in recent years. Among them, the orientation based coding approach, in which the dominant orientation features of palmprints are extracted and encoded into bitwise codes, is one of the most promising approaches. The distance between codes created from two palmprint images is calculated in the matching stage. Reliable orientation feature extraction and efficient matching are the two most crucial problems in orientation based coding approaches. However, conventional coding based approaches usually extract only one dominant orientation feature by adopting filters with discrete orientations, which is sensitive to the noise and rotation. This paper proposed a novel double-orientation code (DOC) scheme to represent the orientation feature of palmprint and designed an effective nonlinear angular matching score to evaluate the similarity between the DOC. Extensive experiments performed on three types of palmprint databases demonstrate that the proposed approach has excellent performance in comparison with previously proposed state-of-the-art approaches.

Remote Sensing Based Spatial Statistics to Document Tropical Rainforest Transition Pathways

In this paper, grid cell based spatial statistics were used to quantify the drivers of land-cover and land-use change (LCLUC) and habitat degradation in a tropical rainforest in Madagascar. First, a spectral database of various land-cover and land-use information was compiled using multi-year field campaign data and photointerpretation of satellite images. Next, residential areas were extracted from IKONOS-2 and GeoEye-1 images using object oriented feature extraction (OBIA). Then, Landsat Thematic Mapper (TM) and Enhanced Thematic Mapper Plus (ETM+) data were used to generate land-cover and land-use maps from 1990 to 2011, and LCLUC maps were developed with decadal intervals and converted to 100 m vector grid cells. Finally, the causal associations between LCLUC were quantified using ordinary least square regression analysis and Moran’s I, and a forest disturbance index derived from the time series Landsat data were used to further confirm LCLUC drivers. The results showed that (1) local spatial statistical approaches were most effective at quantifying the drivers of LCLUC, and (2) the combined threats of habitat degradation in and around the reserve and increasing encroachment of invasive plant species lead to the expansion of shrubland and mixed forest within the former primary forest, which was echoed by the forest disturbance index derived from the Landsat data.

Iris Recognition Based on Local Gabor Orientation Feature Extraction

Iris Recognition Based on Local Gabor Orientation Feature Extraction

Object oriented classification of high resolution data for inventory of horticultural crops

Abstract. High resolution satellite images are associated with large variance and thus, per pixel classifiers often result in poor accuracy especially in delineation of horticultural crops. In this context, object oriented techniques are powerful and promising methods for classification. In the present study, a semi-automatic object oriented feature extraction model has been used for delineation of horticultural fruit and plantation crops using Erdas Objective Imagine. Multi-resolution data from Resourcesat LISS-IV and Cartosat-1 have been used as source data in the feature extraction model. Spectral and textural information along with NDVI were used as inputs for generation of Spectral Feature Probability (SFP) layers using sample training pixels. The SFP layers were then converted into raster objects using threshold and clump function resulting in pixel probability layer. A set of raster and vector operators was employed in the subsequent steps for generating thematic layer in the vector format. This semi-automatic feature extraction model was employed for classification of major fruit and plantations crops viz., mango, banana, citrus, coffee and coconut grown under different agro-climatic conditions. In general, the classification accuracy of about 75–80 per cent was achieved for these crops using object based classification alone and the same was further improved using minimal visual editing of misclassified areas. A comparison of on-screen visual interpretation with object oriented approach showed good agreement. It was observed that old and mature plantations were classified more accurately while young and recently planted ones (3 years or less) showed poor classification accuracy due to mixed spectral signature, wider spacing and poor stands of plantations. The results indicated the potential use of object oriented approach for classification of high resolution data for delineation of horticultural fruit and plantation crops. The present methodology is applicable at local levels and future development is focused on up-scaling the methodology for generation of fruit and plantation crop maps at regional and national level which is important for creation of database for overall horticultural crop development.

Locality oriented feature extraction for small training datasets using non-negative matrix factorization

This paper proposes a simple and effective method to construct descriptive features for partially occluded face image recognition. This method is aimed for any small dataset which contains only one or two training images per subject, namely Locality oriented feature extraction for small training datasets (LOFESS). In this method, gallery images are first partitioned into sub-regions excluding obstructed parts to generate a collection of initial basis vectors. Then these vectors are trained with Non-negative matrix factorization algorithm to find part-based bases. These bases finally build up a local occlusion-free feature space. The main contribution in this paper is the incorporation of locality information into LOFESS bases to preserve spatial facial structure. The presented method is applied to recognize disguised faces wearing sunglasses or scarf in a control environment without any alignment required. Experimental results on the Aleix-Robert database show the effectiveness of the LOFESS method.

Detecting subcanopy invasive plant species in tropical rainforest by integrating optical and microwave (InSAR/PolInSAR) remote sensing data, and a decision tree algorithm

In this paper, we propose a decision tree algorithm to characterize spatial extent and spectral features of invasive plant species (i.e., guava, Madagascar cardamom, and Molucca raspberry) in tropical rainforests by integrating datasets from passive and active remote sensing sensors. The decision tree algorithm is based on a number of input variables including matching score and infeasibility images from Mixture Tuned Matched Filtering (MTMF), land-cover maps, tree height information derived from high resolution stereo imagery, polarimetric feature images, Radar Forest Degradation Index (RFDI), polarimetric and InSAR coherence and phase difference images. Spatial distributions of the study organisms are mapped using pixel-based Winner-Takes-All (WTA) algorithm, object oriented feature extraction, spectral unmixing, and compared with the newly developed decision tree approach. Our results show that the InSAR phase difference and PolInSAR HH–VV coherence images of L-band PALSAR data are the most important variables following the MTMF outputs in mapping subcanopy invasive plant species in tropical rainforest. We also show that the three types of invasive plants alone occupy about 17.6% of the Betampona Nature Reserve (BNR) while mixed forest, shrubland and grassland areas are summed to 11.9% of the reserve. This work presents the first systematic attempt to evaluate forest degradation, habitat quality and invasive plant statistics in the BNR, and provides significant insights as to management strategies for the control of invasive plants and conversation in the reserve.

Word Indexing of Ancient Documents Using Fuzzy Classification

This paper proposes a fuzzy classification system to perform word indexing in ancient printed documents. The indexing system receives a given word selected by an user. The word is preprocessed using an aspect ratio filter, assuring that only interesting word candidates are considered. The image is classified by oriented feature extraction using Gabor filter banks. The oriented features are used to generate membership functions that characterize the selected word. This target word image is then compared to the potential matches, using a similarity matrix. The indexing system is flexible and lightweight when compared to other optimal recognizers, which allows its use in "real-time" applications. A significant test revealed that the indexer achieved very good results in terms of precision and recall in texts from XVIIth century.

Sensor Model Based Preprocessing of 3-D Laser Range Image Data and Motion Oriented Feature Extraction for Mobile Robot Applications

The information processing system of the full-scale autonomous mobile robot MACROBE consists of a multilayered motion planning and control system (MPCS) cooperating with an advanced multisensory system. As a primary sensor within the multisensory system, a 3-D laser range camera was developed. This paper summarizes the basic specifications of the range imaging device. It introduces a sensor model useful for the understanding and analysis of typical sensor measurement errors. Sensor model based error correction methods as used for raw image data processing are discussed. Furthermore several schemes for the extraction of motion-oriented environmental features from 3-D range images are described. These features are used as input for MACROBE’s MPCS. In addition, logical sensors as a means of efficient organization of higher level sensor data processing are introduced. An outlook on the development of advanced logical sensors concludes the paper.