Vgg16 model in keras

can we add Batch Normalization after each convolution layer in keras pre trained model also.

Yes, it is in best practice to use blocks of Conv-BatchNorm-ReLU but please do ensure that there’s no norm already present.