Shadow Detection, Methods and its Evaluation

Mr. Manoj K Sabnis; Dr. (Mrs. ) Amita Sharma

Tracking is one of the main activities of applications like navigation, surveillances, traffic control etc. The success of these applications depends on how successfully tracking services are implemented. Tracking is affected by number of factors like visibility, environmental conditions, motion of objects, nearby similar type of objects and their shadows. The target objects shadows being closely related to the target is tapped as the most important factor to be detected and eliminated. Shadow detection for it elimination started as a pre processing step in tracking but looking at the advantages and its importance, it has now become a independent field of research having its own taxonomy as shadow detection techniques which maps I its own different methods .These methods are further evaluated by different measurable evaluation metrics. This paper aims to target shadow detection methods in its taxonomy framework. It further selects statistical parameters based techniques as the most appropriate and discusses its various methods based on the properties like intensity, colour, geometry and texture. This paper concludes with a analysis of various techniques at a measurable level using evaluation metric.

Select Volume / Issues:

• Volume 3 Issue 6, Nov. 2012

Year:

2012

Type of Publication:

Article

Keywords:

Tracking; Foreground; Background; Subtraction

Journal:

IJECCE

Volume:

3

Number:

6

Pages:

1330-1333

Month:

November