python - How to implement RGB images as tensors in tensorflow? -

i'm new tensorflow , i'm trying create model of stacked sparse denoising auto-encoders. have found way on how load training ( , testing) set through examples here , github cannot use them tensor perform required multiplications etc. (this code loading images)

import tensorflow tf import glob import numpy np pil import image im  im_list = []  #load sets training_set = [] training_set = glob.glob("folder/training_set/*.jpg")  testing_set = [] testing_set = glob.glob("folder/corrupted/*.jpg")   # testing code training set filename_queue = tf.train.string_input_producer(training_set)  reader = tf.wholefilereader() key, value = reader.read(filename_queue)  #data = tf.image.decode_jpeg(value) data = tf.decode_raw(value, tf.uint8)  sess = tf.interactivesession()  sess.run(tf.global_variables_initializer()) #sess.run(tf.global_variables_initializer()) sess.run(tf.local_variables_initializer())  coord = tf.train.coordinator() threads = tf.train.start_queue_runners(coord=coord)  in range (196):     print     m_key = sess.run([key,data])     im_list.append(m_key[1])    coord.request_stop() coord.join(threads) 

by using code manage save images list of uint8 arrays containing data size ~800 ~1000 . images of size 32x32x3 missing.

they other way tried is:

filename_queue = tf.train.string_input_producer(training_set)  image_reader = tf.wholefilereader()  _, image_file = image_reader.read(filename_queue)  imagee = tf.image.decode_jpeg(image_file)  #tf.cast(imagee, tf.float32)  sess = tf.interactivesession()  sess.run(tf.global_variables_initializer())  coord = tf.train.coordinator() threads = tf.train.start_queue_runners(coord=coord)  image = sess.run(imagee)   imaginar = image.astype(np.float32)  #train_step.run(feed_dict={x: imaginar, y_: imaginar_test})  coord.request_stop() coord.join(threads) 

and im trying calculate

y = tf.matmul(x,w) + b            h_x_s = tf.sigmoid(y) h_x = tf.matmul(h_x_s,w_) + b_ y_xi = tf.sigmoid(h_x)  

this way images numpy arrays of 32x32x3 cant find way save them tensor tf.matmul works. errors non fitting shapes of arrays.

# variables x= tf.placeholder(tf.float32,[32, 32, 3]) y_ = tf.placeholder(tf.float32,[32, 32, 3])  w = tf.variable(tf.zeros([32,32,3])) b = tf.variable(tf.zeros([32,32,3]))  w_ = tf.variable(tf.zeros([32,32,3])) b_= tf.variable(tf.zeros([32,32,3])) 

(unsuccessful try)

how should load (and decode) images , sizes should variables , placeholders be? appreciated!

thanks :)

just in case has same problem:

first of use decode_jpeg(data, channels = 3) (channels = 3 means rgb) or other decoder depending on image type.

so can turn 3d image 2d vector. example if image (32,32,3) vector should (1,32*32*3) -> (1, 3072). can using

2d_vec = original_3d_image.reshape(1,-1)

you can turn 3d using

2d_vec.reshape(32,32,3)

do not forget normalize data before use them input. have is

2d_vec = 2d_vec / max_value_of_2d_vec

i have changed lot in code posted before if have questions ask me!