how to use the LightningModule.log to save the image

[FranticEr]

I want to save the confusion matrix image of validation of each epoch, I try to

def validation_step(self,batch, *args: Any, **kwargs: Any) -> STEP_OUTPUT | None:
x,y=batch
# x=self.Flatten(x)
y_hat=self.Net(x)
loss=self.LossFunction(y_hat,y)

    self.confusionMatrix=confusion_matrix(y_hat,y,task='multiclass',num_classes=10)
    
    acc=accuracy(y_hat,y,task='multiclass',num_classes=10)
    self.log("test_acc", acc, on_step=False, on_epoch=True)
    self.log("testloss", loss, on_step=False, on_epoch=True)
    
    return {'loss':loss}

def on_test_epoch_end(self) -> None:
confusionMatrixImag=sns.heatmap(self.confusionMatrix)
self.log('confusionMatrixImage',confusionMatrixImag,on_epoch=True,on_step=False)
return super().on_test_epoch_end()

but it does not work I can't see the images in the tensorboard webui