A Genetic Algorithm-Based Moving Object Detection For Real-Time Traffic Surveillance
- 1. A Genetic Algorithm-Based Moving Object Detection For Real-Time Traffic Surveillance
- 2. Surveillance System using genetic algorithm based moving object detection.
- 3. Moving object detection scheme for surveillance. Genetic Dynamic Saliency Map with Background Subtraction algorithm. Less computation and higher detection accuracy rate.
- 4. Moving object detection based on images obtained from fixed CCTV cameras In real-time embedded systems where the computational and memory resources are scarce, Temporal differencing uses the pixel wise differences between two or three consecutive images in an image sequence to extract moving regions and are highly adaptive to the dynamic scene changes. It accomplishes this task by extracting and modelling each pixel value independently through a mixture of Gaussians of a particular allocation. The pixels which are determined as belonging to the background category within the current frame are then described in the distribution
- 5. Algorithms based on temporal difference fail to extract all the relevant pixels of a foreground object especially when the object has uniform texture or moves slowly. The computational complexity of different algorithms such as object detection is also a challenging task ,when 1) unexpected number of multiple moving objects 2) size variation and poorly textured objects 3) rapid change in illumination conditions 4) shadows and multiple occlusions
- 6. GDSM and Background subtraction for detecting moving objects in surveillance. Background Subtraction is carried out using the threshold obtained from the object candidates using the GDSM. CSD is used to reduce The blur noise around objects, generated by the repetitive image resizing in the Gaussian Pyramid the weights of CSD are optimized using GA, resulting in tighter object regions
- 7. Object detection algorithms intended for embedded systems should be fast, simple and Effective with superior performance. Object detection algorithms intended for embedded systems should be fast, simple and effective with superior performance. Gaussians focuses on robust background modeling and updating to adapt the background model to the varying illumination conditions during different times of the day, geometry reconfiguration of the background structure
- 9. Software Requirement: Language - Java (JDK 1.7), JSP. OS - Windows 7 Ultimate 32-bit. Database - MYSQL 5.0,SQLYOG Apache Tomcat 7 Eclipse or NetBeans IDE 7.1.2 Hardware Requirement : 1 GB RAM 80 GB Hard Disk Above 2GHz Processor Internet connection Data Card