• Corpus ID: 16308105

Illumination invariant interest point detection for vision based recognition tasks

  title={Illumination invariant interest point detection for vision based recognition tasks},
  author={Flore Faille},
Vision based recognition systems learn the appearance of given objects or scenes using images. These objects or scenes can then be recognised and localised in other images. Such recognition systems usually reduce the amount of processed image data by detecting interest points: small characteristic image patches. To improve the robustness of vision based recognition systems under illumination changes, new interest point detectors are developed in this work. They achieve stable interest point… 

The Influence Of Illumination Variety On Repeatability Quality Of Feature Detectors Using Discrete Cosine Transform Algorithm

This research has improved the quality of the repetition of detector features by removing the coefficient of high discrete cosine transform (DCT) with a coefficient of 25%, 50% 75%.

Image Processing for Improvement of Facial Keypoints Detector

This paper discusses the improvement of facial detection algorithms using the DCT algorithm and image processing, and implemented Discrete Cosine Transform, by eliminating the high and low coefficient because there is noise and illumination.

Evaluation of Feature Detectors on Repeatability Quality of Facial Keypoints In Triangulation Method

In this research, the detectors are Harris-Stephens, SURF, FAST, Minimum Eigenvalue, and BRISK have been tested and analyzed through black box test, and the repeatability score for a given pair of images is computed.



Stable Interest Point Detection under Illumination Changes Using Colour Invariants

A new detection method is presented here, which is based on the very popular Harris detector and on the m space, which yields a detection which is invariant to shadows, shading and illumination colour for matte surfaces.

A fast method to improve the stability of interest point detection under illumination changes

  • Flore Faille
  • Computer Science
    2004 International Conference on Image Processing, 2004. ICIP '04.
  • 2004
Based on a simple image formation model and on ideas of homomorphic filtering, the Harris corner detector is modified to reduce the illumination influence, and the computation effort for detection is only minimally increased.

Colour Image Retrieval and Object Recognition Using the Multimodal Neighbourhood Signature

The multimodal neighbourhood signature method is shown to operate successfully under changing illumination, viewpoint and object pose, as well as non-rigid object deformation, partial occlusion and the presence of background clutter dominating the scene.

Illumination-Invariant Motion Detection Using Colour Mixture Models

This paper tackles the problem of robust change detection in image sequences from static cameras using frame differencing with an adaptive background estimation modelled by a mixture of Gaussians.

The Illumination-Invariant Matching of Deterministic Local Structure in Color Images

  • D. SlaterG. Healey
  • Computer Science, Mathematics
    IEEE Trans. Pattern Anal. Mach. Intell.
  • 1997
An algorithm is derived for the recognition of local surface structure which is invariant to these scene transformations of the feature matrices and which is demonstrated with a series of experiments on images of real objects.

Interest Points Detection in Color Images

A new interest point detector for color images is presented based on a non linear filtering of the image which preserves edges, and on the use of a color point detector based on the Harris detector.

Computing illumination-invariant descriptors of spatially filtered color image regions

It is shown, using a set of classification experiments, that the filtered distribution invariants can significantly improve the capability of a recognition system in environments where illumination cannot be controlled.

Color Invariant Edge Detection

A color Differential geometry approach is used to detect material edges, invariant with respect to illumination color and imaging conditions, and the performance of the color invariants is demonstrated by some real-world examples, showing the invariants to be successful in discounting shadow edges and illumination color.

Reflectance based object recognition

An algorithm is developed that estimates a reflectance ratio for each region in an image with respect to its background that is efficient as it computes ratios for all image regions in just two raster scans.

Adapting Interest Point Detection to Illumination Conditions

Three methods are proposed to enhance a state-of-the-art algorithm: the Harris corner de- tector, based on different principles: a local image normalization as preprocessing, a local threshold adaption, and a lo- cal automated threshold selection based on clustering.