Hybrid method based on topography for robust detection of iris center and eye corners

Abstract

A multistage procedure to detect eye features is presented. Multiresolution and topographic classification are used to detect the iris center. The eye corner is calculated combining valley detection and eyelid curve extraction. The algorithm is tested in the BioID database and in a proprietary database containing more than 1200 images. The results show that the suggested algorithm is robust and accurate. Regarding the iris center our method obtains the best average behavior for the BioID database compared to other available algorithms. Additional contributions are that our algorithm functions in real time and does not require complex post processing stages. ACM Reference Format:Villanueva, A., Ponz, V., Sesma-Sanchez, L., Ariz, M., Porta, S., and Cabeza, R. 2013. Hybrid method based on topography for robust detection of iris center and eye corners. ACM Trans. Multimedia Comput. Commun.

Eye center localization and gaze gesture recognition for human–computer interaction

“…Three publicly available databases were tested in our experiments: the BioID database [27], the GI4E database [16] and the extended Yale Face Database b [28]. The BioID database is the most widely employed database for eye centre localisation studies since it contains an array of variations including illumination, face scale, moderate head pose and the presence of glasses; The GI4E database is known for containing images of 103 subjects with 12 different gaze directions; The extended Yale Face Database b is captured under extremely challenging lighting conditions and also contains various head poses.…”

“…Further evaluations on the GI4E database are compared to 6 other methods (similarly to [16] and [34]), as shown in Fig. 7.…”

“…This eye model is derived from the relative rotational invariance of a pupil/iris region. An appearance feature based method [16] analyses greyscale pixel values in relation to their neighbouring pixels. Their method utilises images scaled to different resolutions for increased robustness.…”

Eye centre localisation: an unsupervised modular approach

“…The proposed method gains a total score of 6, outperforming all the other methods in comparison. Similarly to the experiments by Villanueva et al (2013) and Baek et al (2013), we also evaluated the proposed method on the GI4E dataset and compared it to 6 other methods. As shown by Figure 7, the proposed method outperforms all the other methods in comparison and proves to be robust against eye/head movement by achieving 97.9% accuracy for ݁ ௫ 0.05.…”

Real-time eye pupil localization using Hough regression forest

“…To build our forest, we use annotated datasets [23] and [25]. We perform face detection through all the images and extract rough regions around the eyes using anthropomorphic relations.…”