Authors:Keren Kapach; Ehud Barnea; Rotem Mairon; Yael Edan; Ohad Ben-Shahar Pages: 4 - 34 Abstract: Despite extensive research conducted in machine vision for harvesting robots, practical success in this field of agrobotics is still limited. This article presents a comprehensive review of classical and state-of-the-art machine vision solutions employed in such systems, with special emphasis on the visual cues and machine vision algorithms used. We discuss the advantages and limitations of each approach and we examine these capacities in light of the challenges ahead. We conclude with suggested directions from the general computer vision literature which could assist our research community meet these challenges and bring us closer to the goal of practical selective fruit harvesting robots. Keywords: agricultural computer vision; agrovision; agrobotics; fruit harvesting robots; robot vision; machine vision; visual cues. Citation: International Journal of Computational Vision and Robotics, Vol. 3, No. 1/2 (2012) pp. 4 - 34 PubDate: 2012-04-07T23:20:50-05:00 DOI: 10.1504/IJCVR.2012.046419 Issue No:Vol. 3, No. 1/2 (2012)
Authors:Bailing Zhang; Chihang Zhao; Jie He Pages: 35 - 51 Abstract: The identification of the make and model of vehicles from images captured by surveillance camera, also referred to as vehicle type recognition, is a challenging task in intelligent transportation system and automatic surveillance. In this paper, we first comparatively studied two feature extraction methods for image description, i.e., the MPEG-7 edge orientation histogram (EOH) and the pyramid histogram of oriented gradients (PHOGs). EOH captures the spatial distribution of edges by detecting five predefined types of edge directions. PHOG represents the local shape by a histogram of edge orientations computed for each image sub-region, quantised into a number of bins. Compared with previously proposed feature extraction approaches for vehicle recognition, EOH has the advantage of small feature size, economic calculation cost and relative good performance and PHOG has the ascendency in its description of more discriminating information. A composite feature description from PHOG and EOH can further increase the accuracy of classification by taking their complementary information. We then investigate the applicability of the random subspace (RS) ensemble method for vehicle classification based on the combined features. A base classifier is trained with a randomly sampled subset of the original feature set and the ensemble assigns a class label by majority voting. Experimental results using more than 600 images from 21 types show the effectiveness of the proposed approach. The composite feature is better than any single feature in the classification accuracy and the ensemble model produces better performance compared to any of the individual neural network base classifier. With moderate ensemble size 30, the random subspace ensembles offers a classification rate close to 94%, showing the promising potential in real applications. Keywords: vehicle type recognition; vehicle make recognition; edge histogram descriptors; pyramid histogram; oriented gradients; PHOGs; multiple layer perceptron; MLP; random subspace ensemble; vehicle identification; vehicle images; surveillance cameras; inte Citation: International Journal of Computational Vision and Robotics, Vol. 3, No. 1/2 (2012) pp. 35 - 51 PubDate: 2012-04-07T23:20:50-05:00 DOI: 10.1504/IJCVR.2012.046414 Issue No:Vol. 3, No. 1/2 (2012)
Authors:Anand Singh Jalal; Vrijendra Singh Pages: 52 - 74 Abstract: Tracking multiple objects in a scenario that exhibits complex interaction is very challenging. In this work, we propose a multi-resolution framework for multi-object tracking in complex wavelet domain to resolve the challenges resulting from occlusions and splits. In the proposed approach the appearance model is computed at high resolution to achieve more discriminative power to object model, whereas all other tasks are performed at low resolution to get benefited from noise resilience nature of wavelet domain. In this paper, we have also discussed a simple and effective unimodal background subtraction approach to extract moving objects by exploiting the low sub-band characteristics of the object image in complex wavelet domain at low resolution. A scheme exploiting the spatial and appearance information is used to detect and correct the occlusion and split state. Experimental results illustrate the effectiveness and robustness of the proposed framework in ambiguous situations in several indoor and outdoor video sequences. Keywords: multi-object tracking; complex wavelets; multi-resolution; computer vision; object tracking; multiple objects; moving objects; occlusion; split state; ambiguous situations; video sequences. Citation: International Journal of Computational Vision and Robotics, Vol. 3, No. 1/2 (2012) pp. 52 - 74 PubDate: 2012-04-07T23:20:50-05:00 DOI: 10.1504/IJCVR.2012.046415 Issue No:Vol. 3, No. 1/2 (2012)
Authors:Gang Hu; Qigang Gao Pages: 75 - 95 Abstract: Gesture recognition has been an attractive research area for decades. Recently, the video game industry has become the major driving force for the development of advanced gesture control technologies. Conventional video games are controlled via physical devices. In contrast, the emerging trend is using camera-based human computer interface (HCI) to capture human gestures and control game playing directly. This paper presents a novel approach for facilitating the development of gesture control-based video games. A time-of-flight (TOF) camera is adopted to provide both depth and greyscale image sequences. 3D perceptual gesture features are extracted and grouped into a generic gesture representation for target gesture recognition. The game control parameters are derived from the recognised gestures on the fly. This framework includes five key modules: 1) perceptual feature extraction; 2) object tracking by perceptual grouping; 3) representation and modelling; 4) gesture recognition; 5) game control parameter generation. A proof-of-concept dart game is implemented for demonstration and evaluation. Keywords: gesture control; video games; video game interfaces; time-of-flight; TOF 3D camera; perceptual feature extraction; object tracking; tracking; perceptual grouping; gesture representation; modelling; gesture recognition; human-computer inte Citation: International Journal of Computational Vision and Robotics, Vol. 3, No. 1/2 (2012) pp. 75 - 95 PubDate: 2012-04-07T23:20:50-05:00 DOI: 10.1504/IJCVR.2012.046416 Issue No:Vol. 3, No. 1/2 (2012)
Authors:Suvarna S. Nandyal; Basavaraj S. Anami; A. Govardhan; P.S. Hiremath Pages: 96 - 114 Abstract: This paper presents a methodology for identification and classification of images of the medicinal plants based on level set segmentation. The medicinal plants are identified using structural features, namely, height, shape, size of leafy part, flowers, fruits, and branching patterns. In this work, the level sets are used for segmentation of images of medicinal plants. The two segments, namely, leafy part (canopy) and stem, are obtained. The geometrical ratios of length to width of leafy and stem parts of images are used as features. The classification of images of medicinal plants into herbs, shrubs and trees using minimum distance, neural network and SVM classifiers is performed. The experiments are carried on 400 images of medicinal plants of different classes, such as Calotropis gigantea, Aloe vera, Catharantus roseus, Carica Papaya, Azadirachita indica and Cocos nucifera. The classification accuracies obtained by different classifiers are compared. It is observed that the combination of level set segmentation and SVM classifier yielded better classification results. The knowledge of these medicinal plants is useful for practitioners of Ayurveda system of medicine, botanists and common man for home remedies. Keywords: medicinal plants; herbs; shrubs; trees; geometric features; level set segmentation; minimum distance classifier; plant classification; plant images; SVM classification; support vector machines; image segmentation. Citation: International Journal of Computational Vision and Robotics, Vol. 3, No. 1/2 (2012) pp. 96 - 114 PubDate: 2012-04-07T23:20:50-05:00 DOI: 10.1504/IJCVR.2012.046417 Issue No:Vol. 3, No. 1/2 (2012)
Authors:Shenghong Zhong; Wenxiong Kang; Xunchi Wu Pages: 115 - 128 Abstract: This paper introduces the design and realisation of a multi-functional intelligent access control system, in which the hand vein recognition technique plays an essential part. Along with the hand vein image acquisition, processing and matching, the authentication of incoming people can be achieved. Moreover, modules that join newly, such as body induction, voicemail and visualisation, can help lower the system's energy consumption, enhance its versatility and security as well as reliability. This system can provide an intact solution towards user's security. The experiments and running trials also showed that the recognition rate can reach 93.9%, while the false accept rate is under 0.001%. Keywords: smart housing; DSP; vein recognition; intelligent access control; computational vision; hand recognition; hand veins; image acquisition; image processing; image matching; biometrics; security. Citation: International Journal of Computational Vision and Robotics, Vol. 3, No. 1/2 (2012) pp. 115 - 128 PubDate: 2012-04-07T23:20:50-05:00 DOI: 10.1504/IJCVR.2012.046418 Issue No:Vol. 3, No. 1/2 (2012)
Authors:Sashikala Mishra; Kailash Shaw; Debahuti Mishra; Srikanta Patnaik Pages: 129 - 137 Abstract: Classification is a technique where we discover the hidden class level of the unknown data. As different classification methods produces different accuracy according to the class level; classifier fusion is the solution to achieve more accuracy in every level of the input data. Selection of a suitable classifier in classifier fusion is a tedious task. In the proposed model, the output of the three classifiers is fed to the dynamic classifier fusion technique. This model will use each classifier for every individual data. We have used principal component analysis (PCA) to deal with issues of high dimensionality in biomedical classification. Three types of classification techniques on microarray data like multi layer perceptron (MLP), FLANN and PSO-FLANN have been implemented and compared; it has been observed that MLP is showing better result. We have also proposed a model for classifier fusion, where the model will choose the relevant classifiers according to the different region of datasets. Keywords: principal component analysis; PCA; classifier fusion; FLANN; data classification; PSO-FLANN; MLP; modelling; biomedical data. Citation: International Journal of Computational Vision and Robotics, Vol. 3, No. 1/2 (2012) pp. 129 - 137 PubDate: 2012-04-07T23:20:50-05:00 DOI: 10.1504/IJCVR.2012.046420 Issue No:Vol. 3, No. 1/2 (2012)
Proudly sponsored by
LM Information Delivery
One of Europe's leading subscription and information management providers offering cost-efficient solutions for academic and research libraries.
SUNCAT is the largest freely available source of information about
serials holdings in the UK. Researchers are able to locate serials held
in 85 UK research libraries.