Algorithm 1 Feature and parameter extraction

1:Detect features E, C, form abstract shape 2:Calculate $P_e,A_e,(x_{ce},y_{ce})$ from abstract shape 3:Image segmentation detects $P_g,A_g,(x_{cg},y_{cg})$ 4:if unable to detect then 5:    Use $P_e,A_e,(x_{ce},y_{ce})$ 6:else 7:    Compare $A_g$ and $A_e$ to evaluate detections