An object detection framework based on Retina Net is proposed by using Gaussian mixture model（ GMM） and expectation maximization（ EM）. Firstly

a Soft-IOU layer is added to the back end of Retina Net’s feature extraction to evaluate the intersection over union（ IOU） of the predicted bounding boxes and the ground truth. Then

with the help of the quality score evaluated by Soft-IOU

the GMM for rebars detection is generated.Finally

in order to improve the counting precision

Gaussian mixture clustering method based on EM algorithm is used to solve the problem that the original Retina Net framework is not ideal for dense object detection. The experimental results show that the average precision of the improved method is 3. 3 % higher than that of the Retina Net

and the counting root mean square error is improved by 64. 2. Therefore

the method shows strong adaptability.