Combining color-based invariant gradient detector with HoG descriptors for robust image detection in scenes under cast shadows

Villamizar, Michael; Scandaliaris, J.; Sanfeliu, Alberto; Andrade‐Cetto, Juan

doi:10.1109/robot.2009.5152429

Cited by 16 publications

(12 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this sense, Gevers and Smeulders [47] and later Villamizar et al [42], deeply analysed diverse colour models by evaluating their robustness for object recognition under different image parameters. This comparison, summarized in Table I, concluded that the colour model to be chosen depends on the imaging conditions.…”

Section: A Feature Extractionmentioning

confidence: 99%

See 1 more Smart Citation

Object Detection and Recognition for Assistive Robots: Experimentation and Implementation

Martínez-Martín

Pobil

2017

IEEE Robot. Automat. Mag.

View full text Add to dashboard Cite

Abstract-Technological advances are currently being directed to assist the human population in performing ordinary tasks in everyday settings. In this context, a key issue is the interaction with objects of varying size, shape and degree of mobility. Consequently, autonomous assistive robots must be provided with the ability to process visual data in real time so that they can react adequately for quickly adapting to changes in the environment. Reliable object detection and recognition is usually a necessary early step to achieve this goal. In spite of significant research achievements, this issue still remains a challenge when real-life scenarios are considered. In this paper, we present a vision system for assistive robots that is able to detect and recognise objects from a visual input in ordinary environments in real time. The system computes colour, motion and shape cues combining them in a probabilistic manner to accurately achieve object detection and recognition, taking some inspiration from vision science. In addition, with the purpose of processing the input visual data in real-time, a Graphical Processing Unit (GPU) has been employed. The presented approach has been implemented and evaluated on a humanoid robot torso located at realistic scenarios. For further experimental validation, a public image repository for object recognition has been used, allowing a quantitative comparison with respect to other state-of-the-art techniques when real-world scenes are considered. Finally, a temporal analysis of the performance is provided with respect to image resolution and number of target objects in the scene.

show abstract

Section: A Feature Extractionmentioning

confidence: 99%

“…In this way, colour-based invariant gradients have been combined with Histogram of Oriented Gradient (HoG) local features [42] for object detection in outdoor scenes (such as urban scenes) under cast shadows. The approach is, however, limited by the constrained nature of the environments.…”

Section: Introductionmentioning

confidence: 99%

Object Detection and Recognition for Assistive Robots: Experimentation and Implementation

Martínez-Martín

Pobil

2017

IEEE Robot. Automat. Mag.

View full text Add to dashboard Cite

show abstract

“…Bag of Words (BoW) model [13,14,15,16,17] is originated from natural language processing tasks and information retrieval. For image analysis, a visual analogue of a word is used in the BoW model, which is based on the vector quantization process by clustering low-level visual features of local regions or points, such as color, texture, and so forth [18].…”

Section: Bag Of Featuresmentioning

confidence: 99%

“…SIFT and SURF can be used in image categorization. Generally, [10][11][12][13][14][15][16][17][18][19][20][21][22][23][24] better performance and efficiency of training and classification depend on better representation and clustering of features.…”

Section: Introductionmentioning

confidence: 99%

Efficient Modern Description Methods by Using SURF Algorithm for Recognition of Plant Species

Kazerouni

Schlemperz²,

Kuhnert³

2015

AIVP

View full text Add to dashboard Cite

Plants are one of one of most valuable natural resources. There would be no life on earth without plants. They unlike humans and animals manufacture their own food by photosynthesis. In every food chain, the plants occupy the first position and lead the chain as source of food. Environment and climate are largely interlinked with plants. Rainfall, humidity and temperature are influenced by presence of plants. Cutting down plants also imbalance the environment which will indirectly affect human life. Even the economic importance of plants is also quite large to mankind. Plants are great contributors of economy. Many countries rely on agriculture as one of the main source of revenue. Other benefits of plants are significant applications in different fields. Medical and agricultural applications are just some instances of plants application.Plant recognition can be done by using unique characteristic parts of plants. The used part is leaf. Shapes of leaves are useful to do plant recognition and find the species. Bag of words (BoW) and support vector machine (SVM) methods are applied to recognize and identify plants species. Visual contents of images are used and four steps are performed: (i) image preprocessing, (ii) BoW, (iii) train, (v) test. Three combined methods are used on Flavia dataset. The proposed approach is done by Speed-up robust features (SURF) method and two combined method, HARRIS-SURF and features from accelerated segment test-SURF (FAST-SURF). The accuracy of SURF method is higher than other applied methods. It is 92.28395 %. In addition to visional comparison, some quantitative results are measured and compared.

show abstract

“…The best performing methods often use a monolithic representation of people, such as a HOG descriptor [9], and discriminative classifiers. Models of this type have recently been extended to incorporate motion [10], [36] and color [35] features. They have also been applied to upper body detection [18], and have been integrated within larger systems to enable obstacle detection in mobile environments [14].…”

Section: Introduction and Related Workmentioning

confidence: 99%

Vision based victim detection from unmanned aerial vehicles

Andriluka

Schnitzspan

Meyer

et al. 2010

2010 IEEE/RSJ International Conference on Intelligent Robots and Systems

View full text Add to dashboard Cite

Abstract-Finding injured humans is one of the primary goals of any search and rescue operation. The aim of this paper is to address the task of automatically finding people lying on the ground in images taken from the on-board camera of an unmanned aerial vehicle (UAV).In this paper we evaluate various state-of-the-art visual people detection methods in the context of vision based victim detection from an UAV. The top performing approaches in this comparison are those that rely on flexible part-based representations and discriminatively trained part detectors. We discuss their strengths and weaknesses and demonstrate that by combining multiple models we can increase the reliability of the system. We also demonstrate that the detection performance can be substantially improved by integrating the height and pitch information provided by on-board sensors. Jointly these improvements allow us to significantly boost the detection performance over the current de-facto standard, which provides a substantial step towards making autonomous victim detection for UAVs practical.

show abstract

Combining color-based invariant gradient detector with HoG descriptors for robust image detection in scenes under cast shadows

Cited by 16 publications

References 11 publications

Object Detection and Recognition for Assistive Robots: Experimentation and Implementation

Object Detection and Recognition for Assistive Robots: Experimentation and Implementation

Efficient Modern Description Methods by Using SURF Algorithm for Recognition of Plant Species

Vision based victim detection from unmanned aerial vehicles

Contact Info

Product

Resources

About