why plt.imshow(predict) shows a fully black image

Question

I set up a mechine learning model to predict a picture based on a given picture(using tensorflow). but when i use matplotlib, and use pyplot.imshow(predict), the picture showed is all black while the predict data is not all zero.

i am a freshman in python.any information would help a lot. the predict data displayed by pycharm debugger

predict = model.predict(np.array(x)).reshape([512, 512, 3])*255

predict=predict[:,:,0]
# # predict=tf.argmax(predict,axis=2)
# with tf.Session() as sess:
#     predict=predict.eval()


predict = np.array(predict.astype('uint8'))
print(predict.dtype)

pyplot.figure(2)
pyplot.imshow(predict)
pyplot.show()

Answer 1

the model.predict will output a possibility that means when the metric is too big, all data is nearly the same zero.

Answer 2

While plotting the image you can set the minimum and maximum value range for the imshow

pyplot.imshow(predict,vmin=0, vmax=255)