Physical models for moving shadow and object detection in video

Nadimi, S.; Bhanu, Bir

doi:10.1109/tpami.2004.51

Cited by 240 publications

(117 citation statements)

References 15 publications

Supporting

Mentioning

116

Contrasting

Order By: Relevance

“…Generally, the cast shadows would be detected as foreground objects since the shadows share the same movement patterns and have a similar magnitude of intensity change as that of the foreground object [1]. Since the shadow regions are sometimes as big as the object regions, the wrong classification as foreground objects can cause various unwanted consequence such as object shape distortion, false object detection and object merging, which has a bad impact on object detection and tracking.…”

Section: Introductionmentioning

confidence: 99%

An Adaptive Cast Shadow Detection with Combined Texture and Color Models

Liu¹,

Zhu²

2014

IJET

View full text Add to dashboard Cite

Abstract-Cast shadows of moving foreground objects can cause miss tracking problem in object detection and tracking, thus shadow detection is an important step used after a moving foreground object is detected. Most of current methods have a significant trade-off between the shadow detection rate and the shadow discrimination rate. In this paper, an effective and adaptive method with combined texture and color models is proposed in order to achieve good shadow detection rate and shadow discrimination rate as well. Firstly, Scale Invariant Local Ternary Pattern (SILTP) is used to select a candidate shadow region. Then HSV color model is employed to detect a new candidate shadow region by using maximum likelihood estimation (MLE) to estimate the thresholds of HSV color model adaptively. Finally the two regions are combined by logical operation and a new shadow region can be obtained. Our experimental results show that the proposed method achieves a better performance in both shadow detection rate and discrimination rate compared to the other current methods. Moreover, the proposed method runs at 100 frames per second and is suitable for the real-time detection and tracking.

show abstract

Section: Introductionmentioning

confidence: 99%

An Adaptive Cast Shadow Detection with Combined Texture and Color Models

Liu¹,

Zhu²

2014

IJET

View full text Add to dashboard Cite

show abstract

“…The second boosting classifier uses our method, that is the photometric invariant contour detector. 1 Using both methods, the learning boosting step selected 100 weak classifiers for each method from a learning set of 50 object images and 200 background images. The background images were extracted from patches of outdoor and indoor images using a randomized process.…”

Section: Methodsmentioning

confidence: 99%

“…In the RGB color model {R, G, B} values correspond directly with V k in (1). The c1c2c3 color model is defined by…”

Section: Color Modelsmentioning

confidence: 99%

“…The presence of noise in the images can lead to the appearance of maxima in the gradient modulus that are not related to any parameter of the image formation process (1). If it is assumed that the sensor noise is normally distributed, and that we know the parameters for each particular sensor, then using (14) and (15) (see appendix) the uncertainties associated to the c1c2c3 and o1o2 color models, as well as the different gradient moduli, can be propagated from the a priori known sensor uncertainties.…”

Section: Contour Detectionmentioning

confidence: 99%

“…Nadimi et al [1] use a multistage approach based on physical models to detect moving shadows in video. Input video frames are passed through a moving object detection stage and then through a series of classifiers which distinguish object pixels from shadow pixels.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Robust Color Contour Object Detection Invariant to Shadows

Scandaliaris

Villamizar

Andrade‐Cetto

et al.

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. In this work a new robust color and contour based object detection method in images with varying shadows is presented. The method relies on a physics-based contour detector that emphasizes material changes and a contourbased boosted classifier. The method has been tested in a sequence of outdoor color images presenting varying shadows using two classifiers, one that learnt contour object features from a simple gradient detector, and another that learnt from the photometric invariant contour detector. It is shown that the detection performance of the classifier trained with the photometric invariant detector is significantly higher than that of the classifier trained with gradient detector.

show abstract

Moving Cast Shadows Detection Methods for Video Surveillance Applications

Amato

Huerta

Mozerov

et al. 2012

Augmented Vision and Reality

View full text Add to dashboard Cite

Moving cast shadows are a major concern in today's performance from broad range of many vision-based surveillance applications because they highly difficult the object classification task. Several shadow detection methods have been reported in the literature during the last years. They are mainly divided into two domains. One usually works with static images, whereas the second one uses image sequences, namely video content. In spite of the fact that both cases can be analogously analyzed, there is a difference in the application field. The first case, shadow detection methods can be exploited in order to obtain additional geometric and semantic cues about shape and position of its casting object ('shape from shadows') as well as the localization of the light source. While in the second one, the main purpose is usually change detection, scene matching or surveillance (usually in a background subtraction context). Shadows can in fact modify in a negative way the shape and color of the target object and therefore affect the performance of scene analysis and interpretation in many applications. This chapter wills mainly reviews shadow detection methods as well as their taxonomies related with the second case, thus aiming at those shadows which are associated with moving objects (moving shadows). IntroductionVideo Surveillance has been in our society for a long time [6,13]. It began in the twentieth century to assist prison officials in the discovery of escape methods. However, it was not until the late-twentieth century that surveillance expanded to include the security of property and people. Video surveillance is more prevalent in Europe than anywhere in the world. For instance, in the past decade, successive UK governments have installed over 2.4 million surveillance cameras (about one for every 14 people). 1 The average Londoners are estimated to have their picture recorded more than three hundred times a day 2 . Traditionally video surveillance was used to display images on monitors inspected by guards or operators. This fact has allowed the observation of an increase number of places using less people and also to perform patrolling duties from the safety of a control room. However, a single operator can only monitor a limited amount of scenes simultaneously and for a limited amount of time, because the process of manual surveillance is very time-consuming and is a really tedious task.The new breakthroughs in technology have led to a new generation of video surveillance. The current generation of video surveillance systems uses digital computing and communication technologies to improve the design of the original architecture, with the ultimate goal to create an automatic video surveillance system. Recent trends in computer vision has delved into the study of cognitive vision systems, which uses visual information to facilitate a series of tasks on sensing, understanding, reaction and communication. In other words, video surveillance systems aim to automatically identify people, objects or events of ...

show abstract

Physical models for moving shadow and object detection in video

Cited by 240 publications

References 15 publications

An Adaptive Cast Shadow Detection with Combined Texture and Color Models

An Adaptive Cast Shadow Detection with Combined Texture and Color Models

Robust Color Contour Object Detection Invariant to Shadows

Moving Cast Shadows Detection Methods for Video Surveillance Applications

Contact Info

Product

Resources

About