Phytoplankton Discrimination Method with Wavelet Descriptor Based Shape Feature Extracted from Microscopic Images (Signal analysis and time-frequency analysis)

Abstract

Image retrieval method based on Euclidian distance between normalized features with their mean and variance in feature space is proposed. Effectiveness of the normalization is evaluated together with a validation ofthe proposed image retrieval method. The proposed method is applied for discrimination and identifying dangerous red tide species based on wavelet utilized classification methods together with texture and color features. Through experiments, it is found that classification performance with the proposed wavelet derived shape information extracted from the microscopic view of the phytoplankton is effective for identifying dangerous red tide species among the other red tide species rather than the other conventional texture, color information. Moreover, it is also found that the proposed normalization of features is effective to improve identification performance. A visualization method for representation of 3D object shape complexity based on the proposed wavelet descriptor is proposed together with its application to image retrievals. Image retrieval method using wavelet descriptor of shape information together with hue and texture information of objects extracted with dyadic wavelet transformation is proposed. Although there are conventional methods for image retrievals with hue and texture information, image retrieval performance (hit ratio) is not so high. Therefore, the proposed method uses shape information derived from objects extracted from original images in addition to the hue and texture information. In order extract object, dyadic wavelet transformation is used to find good focusing image area extraction as objects. Experimental results with several kinds ofphytoplankton show some improvement of hit ratio as well as Euclidian distance among images.