An Automatic Algorithm for Object Recognition and Detection Based on ASIFT Keypoints

Reza Oji

DOI: https://doi.org/10.5121/sipij.2012.3503

2012-11-26

Abstract:Object recognition is an important task in image processing and computer vision. This paper presents a perfect method for object recognition with full boundary detection by combining affine scale invariant feature transform (ASIFT) and a region merging algorithm. ASIFT is a fully affine invariant algorithm that means features are invariant to six affine parameters namely translation (2 parameters), zoom, rotation and two camera axis orientations. The features are very reliable and give us strong keypoints that can be used for matching between different images of an object. We trained an object in several images with different aspects for finding best keypoints of it. Then, a robust region merging algorithm is used to recognize and detect the object with full boundary in the other images based on ASIFT keypoints and a similarity measure for merging regions in the image. Experimental results show that the presented method is very efficient and powerful to recognize the object and detect it with high accuracy.

Artificial Intelligence,Computer Vision and Pattern Recognition

What problem does this paper attempt to address?

This paper aims to solve the object recognition problem in image processing and computer vision, especially by combining affine - scale - invariant feature transform (ASIFT) keypoints and region merging algorithms to achieve complete object boundary detection. Specifically, the paper proposes an automated algorithm that can train objects at different viewpoints and scales, find the best keypoints, and use these keypoints to recognize and detect objects in other images, while being able to accurately extract the complete boundaries of objects.

### Overview of the paper's problem - solving methods:
1. **Using the ASIFT algorithm**: ASIFT is a fully affine - invariant algorithm that can handle six affine parameters (translation, scaling, rotation, two camera - axis directions). This makes the keypoints extracted from images taken at different viewpoints and scales more reliable.
2. **Keypoint extraction and matching**: By training objects in multiple images at different viewpoints and scales, the best keypoints are found. These keypoints are used to match the same object in different images.
3. **Region merging algorithm**: Using the keypoints extracted by ASIFT and the similarity measure based on Euclidean distance, region merging is carried out to recognize and detect the complete boundaries of objects. This method does not require users to mark the background and target regions, achieving automation.
4. **Experimental verification**: The effectiveness and accuracy of this method are verified by conducting experiments on datasets of three different objects (shelves, bottles, calendars).

### Main contributions:
- **Automation**: This method is a fully automatic algorithm and does not require user - manual marking.
- **High precision**: The experimental results show that this method has a high accuracy rate in object recognition and boundary detection.
- **Robustness**: By using ASIFT keypoints, this method is highly robust to image changes (such as changes in viewpoint and illumination).

### Conclusion:
The method proposed in the paper performs well in object recognition and boundary detection, especially in dealing with complex scenes and variable image conditions. Future work will further explore the possibility of shape recognition based on the extracted boundaries.