CS-ABC: Cuckoo Search based adaptive boosting classifier for malaria infected erythrocyte detection

Abstract

The contribution of this manuscript is a supervised learning technique that enables to detect the malaria scope in given erythrocyte. Contrast to former efforts in this region, this study utilises multiple observable phase images of the unspotted cells. Automatically, erythrocytes are segmented utilising optical phase thresholds and redeployed to allow quantitative association of the phase images. Redeployed images are examined to extract manifold morphological descriptors on the basis of phase information. When the entire individual descriptors are statistically diverse among uninfected and infected cells, every descriptor will not allow parting of the populations at the satisfactory level aimed at clinical service. The experimental study carried on the proposal and other contemporary models evincing that the proposed CS-ABC (Cuckoo Search based Adaptive Boosting Classifier) is considerably significant with maximal prediction accuracy and minimal misclassification rate that compared to other contemporary models.