Automatic object detection and matching based on proposed signature

Abstract

The traditional approaches to object to detection and classifications only look at local pieces of an image; it is within a sliding window or the regions around an interest point detector. This paper introduces a new idea for automatic object detection based on its own proposed signature. This signature is completely distinct, simple to use, saved in small memory, and it is determined in different lights on objects. Exceedingly, the matching is very fast and more accurate under not clearing constraint. After determination of signatures, the matching process among the inputs object's signature and other saved standard shapes signatures done. This comparison is making among input signatures with other memorized signatures of saved basic standard shapes, like circle, square ellipse rectangle triangle and so on. This proposed method examines 95 different objects and achieved 92.5% of expected right matching.

Proposed Method for Detecting Objects

“…They are consisting of segmentation, construction of objects' signatures in image and matching them to classify the object based on its signature [7]. The segmentation process represents the main stone in this algorithm, which is given initial hypotheses of object positions, scales and supporting based on matching.…”

“…The greatest common divisor of the above approaches is that they can fail when the regional image information is insufficient (target is very small or unclear), and this is considered as a weakness of them [7]. In this way, the image matching based on features is depending on analyzing the extracted features and find the corresponding relationship between them [24].…”