Cornea Recognition Using a Cepstral Approach and SVM
Subscribe/Renew Journal
In this paper, a new technique for feature extraction from corneal images is presented which can be applied for corneal pattern recognition. Most of the previous methods are based on segmentation of the corneal images which are restricted to certain planes. In this paper, a proposed method is applied on corneal images which have two main phases. Firstly, the 2-D images are lexicographically ordered to 1-D signals, and then the Mel Frequency Cepstral Coefficients (MFCCs) and polynomial coefficients are extracted from these 1-D signals or from their transforms. Secondly, the SVM is used to match the extracted features in the testing phase to those of the training phase. Experimental results show that the recognition rate for features extracted from Discrete Sine Transform DST and Discrete Cosine Transform (DCT) achieve better performance compared to other cases. The method in this paper is limited to feature extraction for pattern recognition and the automatic diagnosis case is left for future work.
Keywords
Abstract Views: 215
PDF Views: 2