Description

The object detection is the process to identify the objects in the train image using the query image. In proposed detect the object of the image without training process. The without training method is rectify the memory allocation problem in object detection method. But it has attain some major problems to identify the object very well. In training based method all type of the image trained well using the feature extraction algorithms. But in our method there is no training so the object is detect by the orientation of the figure. Here we use the SIFT and PCA for extract the descriptors points and reduce that points are reduced to the query image and calculate average Euclidean distance of the points and match the object easily. The proposed method rectify the memory allocation problem for store the train feature and same time give the object image as the query and it find the object in train image by the SIFT and PCA methods.