Fig. 1
From: Tile-based microscopic image processing for malaria screening using a deep learning approach
General overview of the proposed framework. First, the high-resolution image is sliced into overlapping tiles and fed as an input to the detection network for training. Then, the trained model is used during inference to predict parasite location at individual tiles. Finally, the initial prediction results are fused to generate refined detections to be superimposed on the input high-resolution image