A Study of Feature Based Techniques in Visual Object Recognition

ELECTRONICS & COMMUNICATION ENGINEERING

Robotics is one of the research area in computer age. So, to make the robots as capable as humans, to allow them to interact with real environment so many algorithms are developed and will be developed. Some of those algorithms are developed in the area of computer vision to allow the robots for accurate recognition. In all those algorithms feature extraction technique is most important part of the algorithm. As the features are robust to different affine transformations like translation, scale, rotation, flipped, etc. the algorithm will be more robust to those transformations. So, feature extraction techniques are one of the important part of the image retrieval systems. Some of those feature extraction techniques, with their invariance properties are discussed here for the image retrieval system. Object recognition in computer vision is the task of determining similar objects in an image or in video sequence. The shape based object recognition is made difficult by many adversaries like variations in pose, occlusion, scaling, poor illumination, rotations, etc. A variety of techniques developed for shape based object recognition to make the recognition easier and accurate. Matching is one of the central issues of object recognition and important part of the object recognition system. A common goal is to measure the image data and compare it and verify it for accurate recognition process. Matching process in object recognition involves some type of search and when search takes place the set of extracted features will be compared with the stored features for recognition process. To accomplish this task relevant and comparable information should be extracted from the input data.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.