Motion Detection Using Colour Templates
نویسندگان
چکیده
The Horridge template model1 is an empirical motion detection model inspired by insect vision. This model has been successfully implemented on several micro-sensor VLSI chips26 using greyscale pixels. The template model is based on movement of detected edges rather than whole objects, which facilitates simple tracking techniques. Simple tracking algorithms developed by Nguyen7 have been successful in tracking coherent movement of objects in a simple environment. Due to the inherent edge detection nature of the template model, two closely spaced objects moving at the same speed relative to the template model sensor will appear to have a common edge and hence be interpreted as one object. Hence when the two objects separate, the tracking algorithm will be upset by the detection of two separate edges, resulting in a loss of tracking. This paper introduces a low-cost insect vision prototype, based on a colour CMOS camera. Although this approach sacrifices auto gain control (AGC) at each pixel, results are valid for controlled lighting conditions. We demonstrate working results, for indoor conditions, by extension of the template model using the colour CMOS sensor to form colour templates. This enables the detection of colour boundaries or edges of closely moving objects by exploiting the difference in colour contrast between the objects. This paper also discusses the effectiveness of this technique in facilitating the independent tracking of multiple objects.
منابع مشابه
Reconstruction of Local Features for Facial Video Compression
In this work, we suggest a feature based facial image reconstruction technique. The encoding technique utilizes affine structure of the face to capture the global facial motion [5]. The structure information is transmitted only once at the bootstrapping stage. After that only the motion information for each frame is used to reconstruct the image at the decoder. An attempt is made to perform loc...
متن کاملColour-Invariant Motion Detection under Fast Illumination Changes1
This paper tackles the problem of robust change detection in image sequences from static cameras. Motion cues are detected using frame differencing with an adaptive background estimation modelled by a mixture of Gaussians. Illumination invariance and elimination or detection of shadows is achieved by using a colour chromaticity representation of the image data.
متن کاملA Novel Background Subtraction for Dynamic Texture Scenes by Exploitation of Fuzzy Colour bar Graph and Morphological Operations
Background subtraction in dynamic scenes is an important and challenging task. This paper proposes an efficient motion detection system based on background subtraction using fuzzy colour histogram and morphological processing. Here two methods are used effectively for object detection followed by people counting and compare these performance based on accurate estimation. In dynamic texture scen...
متن کاملIllumination-Invariant Motion Detection Using Colour Mixture Models
This paper tackles the problem of robust change detection in image sequences from static cameras. Motion cues are detected using frame differencing with an adaptive background estimation modelled by a mixture of Gaussians. Illumination invariance and elimination or detection of shadows is achieved by using a colour chromaticity representation of the image data. The combination of the colourand ...
متن کاملDetection of abl/bcr Fusion Gene in Patients Affected by Chronic Myeloid Leukaemia by Dual-Colour Interphase Fluorescence in situ Hybridisation
Conventional cytogenetic is the standard technique for detection of Philadelphia (Ph) chromosome in chronic myeloid leukemia (CML). Evaluation of abelson murine leukemia/breakpoint cluster region (abl/bcr) fusion using dual-colour fluorescence in situ hybridization (D-FISH) is an alternative approach allowing rapid and reliable detection of the disease. We employed the technique of interphase D...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2003