Special Object Detection Based On Mask Rcnn
Miaomiao Chen, Fengshan Bai, Z. Gerile · 2021 17th International Conference on Computational Intelligence and Security (CIS) · 2021
Mask RCNN is committed to improving the accuracy of target detection. Mask RCNN target detection algorithm belongs to regional convolutional neural network, it can not only detect the target, but also use deep semantic segmentation target. This method makes the location of special objects more accurate and expression more vivid. When it detect the target, will produce a high quality split mask. The mask adds an additional branch to fast RCNN to predict the target. So its detection speed is faster than fast RCNN. Experiment shows that Mask RCNN have high recognition rate and accurate segmentation rate. it can identify and detect the target object effectively.