High accuracy on test and validation data but still can not predict on real data

Question

Hello i am having a classification between two classes A and B and i have trained CNN model. I have high accuracy on all three set of data i.e training (98.7%) validation (99.3%) and test(98%) but still can not predict on real data could any please help to sort out the issue.
p.s i have balance data.

train_augmented = ImageDataGenerator(rescale=1./255,
                                   rotation_range = 40,
                                   horizontal_flip = True,
                                   
                                   zoom_range = 0.2
                                   )
test_augmented = ImageDataGenerator(rescale=1./255)
validation_augmented = ImageDataGenerator(rescale=1./255)

training_set = train_augmented.flow_from_directory(train,
                                                    target_size=(150,150),
                                                    batch_size = 32,
                                                    class_mode = 'binary',
                                                    
                                                    color_mode="rgb",
                                                    shuffle=True
                                                    )
test_set = test_augmented.flow_from_directory(test,
                                                    target_size=(150,150),
                                                    batch_size = 32,
                                                    class_mode = 'binary',
                                                    color_mode="rgb",
                                                    shuffle=False
                                                    )
validation_set = validation_augmented.flow_from_directory(validation,
                                                    target_size=(150,150),
                                                    batch_size = 32,
                                                    class_mode = 'binary',
                                                    color_mode="rgb",
                                                    shuffle=False
                                                    )
model = Sequential()
model.add(Conv2D(32,(3,3),activation='relu',input_shape=(150,150,3)))
model.add(MaxPooling2D(pool_size=(2,2)))
model.add(Dropout(0.5))

model.add(Conv2D(64,(3,3),activation='relu'))
model.add(MaxPooling2D(pool_size=(2,2)))
model.add(Dropout(0.5))

model.add(Flatten())
model.add(Dense(256,activation='relu'))
model.add(Dropout(0.5))
model.add(Dense(1,activation='sigmoid'))
model.summary()
model.compile(Adam(learning_rate=0.001),loss='binary_crossentropy',metrics=['accuracy'])
history = model.fit_generator(training_set,
                    epochs = 30,
                    validation_data = validation_set)

Answer 1

It means that your training/validation/test data is not representative of the real world.

You should get fresh new data from the real world, and use it instead.

Of course I trust that your training/validation/test directories do not contain duplicates.

Answer 2

First of all, your dataset is it balanced ? If not, accuracy is a bad metric to use because it will be strongly influenced by the most represented class. You can use F&-Score which will be better.

Are you sure that your test set is different than your training set ?

Otherwise, like @nicolas-raoul suggest try to be sure that your training / test set are the same.

Good luck with your investigation !