RETRACTED: An efficient object detection using OFSA for thermal imaging (Retracted Article)

The demand for identifying a reliable person is increased because of security issues in our daily life. At present, to identify a person biometric technique such as face recognition is introduced. Since,a person with abnormal behaviour recognition system has reached certain level, their accomplishments in real time applications are restricted by challenges, such as illumination variations. The present visual recognition system is good at controlled illumination conditions and thermal face recognition system is better for detecting disguised persons or when there is no illumination control. Hence, a hybrid system which uses both visual and thermal images for recognising a person is better. The objective of this research is to implement a method which improves the quality of the image by fusing visual and thermal imaging images. Our research methodology has introduced to enhance servo line camera images. Nonlinear image transfer functions were introduced,and the parameters associated with those functions are determined by image statistics for making adaptive algorithms. Next methodswereintroduced for registering the visual images to their consequent thermal images. To get a transformation matrix for the registration, the landmarks in the images are first detected and a subset of those landmarks were selected to obtain the matrix, we propose a hybrid algorithm for detection, tracking and classification using OFSA algorithm to fuse the registered thermal and visual images. In this research, we focus on object detection using OFSA algorithm for more accuracy.