Python misc.imsave方法代码示例

# 需要导入模块: from scipy import misc [as 别名]
# 或者: from scipy.misc import imsave [as 别名]
def process_frame(frame_idx, img, model, write_to_dir, conf_threshold, input_size=224):
    """Finds bounding boxes in a video frame, draws these bounding boxes
    and saves the result to HDD.
    """
    # find BBs in frame
    bbs, time_model = find_bbs(img, model, conf_threshold, input_size=input_size)

    # draw BBs
    img_out = np.copy(img)
    for (bb, score) in bbs:
        if score > conf_threshold and bb.width > 2 and bb.height > 2:
            img_out = bb.draw_on_image(img_out, color=[0, 255, 0], thickness=3)

    # save to output directory
    save_to_fp = os.path.join(write_to_dir, "%05d.jpg" % (frame_idx,))
    misc.imsave(save_to_fp, img_out)

    return time_model 

# 需要导入模块: from scipy import misc [as 别名]
# 或者: from scipy.misc import imsave [as 别名]
def crop_det(det_M, img): 
    global track_struct
    crop_det_folder = track_struct['file_path']['crop_det_folder']
    crop_size = track_struct['track_params']['crop_size']
    if not os.path.isdir(crop_det_folder): 
        os.makedirs(crop_det_folder) 
    
    save_patch_list = []
    for n in range(len(det_M)):
        xmin = int(max(0,det_M[n,1])) 
        xmax = int(min(img.shape[1]-1,det_M[n,1]+det_M[n,3])) 
        ymin = int(max(0,det_M[n,2])) 
        ymax = int(min(img.shape[0]-1,det_M[n,2]+det_M[n,4])) 
        img_patch = img[ymin:ymax,xmin:xmax,:] 
        img_patch = misc.imresize(img_patch, size=[crop_size,crop_size]) 
        patch_name = track_lib.file_name(n,4)+'.png'
        save_path = crop_det_folder+'/'+patch_name 
        misc.imsave(save_path, img_patch)
        save_patch_list.append(save_path)
  
    return save_patch_list 

# 需要导入模块: from scipy import misc [as 别名]
# 或者: from scipy.misc import imsave [as 别名]
def main():
    for file in os.listdir(data_image_dir):
        if file.endswith(".png"):
            print("Try to copy %s" % file)
            im = misc.imread(os.path.join(data_image_dir, file), mode='RGB')
            height, width, ch = im.shape
            assert ch == IMAGE_DEPTH
            if height == IMAGE_HEIGHT and width == IMAGE_WIDTH and ch == IMAGE_DEPTH:
                misc.imsave(os.path.join(image_dir, file), im)
            else:
                print("Size: (%d, %d, %d) cannot be used." % (height, width, ch))

    for file in os.listdir(data_label_dir):
        if file.endswith(".png"):
            print("Try to converting %s" % file)
            gt_label = convert_to_label_data(os.path.join(data_label_dir, file))
            if gt_label is not None:
                misc.imsave(os.path.join(label_output_dir, file), gt_label) 

# 需要导入模块: from scipy import misc [as 别名]
# 或者: from scipy.misc import imsave [as 别名]
def setup(self, pre_encode=False):

        target_path = self.root + '/combined_annotations/'
        if not os.path.exists(target_path):
            os.makedirs(target_path)

        if pre_encode:
            print("Pre-encoding segmentation masks...")
            for i in tqdm(self.sbd_train_list):
                lbl_path = self.sbd_path + 'dataset/cls/' + i + '.mat'
                lbl = io.loadmat(lbl_path)['GTcls'][0]['Segmentation'][0].astype(np.int32)
                lbl = m.toimage(lbl, high=self.ignore_index, low=0)
                m.imsave(target_path + i + '.png', lbl)
            for i in tqdm(self.sbd_val_list):
                lbl_path = self.sbd_path + 'dataset/cls/' + i + '.mat'
                lbl = io.loadmat(lbl_path)['GTcls'][0]['Segmentation'][0].astype(np.int32)
                lbl = m.toimage(lbl, high=self.ignore_index, low=0)
                m.imsave(target_path + i + '.png', lbl)
            for i in tqdm(self.files['trainval']):
                lbl_path = self.voc_path + 'SegmentationClass/' + i + '.png'
                lbl = self.encode_segmap(m.imread(lbl_path))
                lbl = m.toimage(lbl, high=self.ignore_index, low=0)
                m.imsave(target_path + i + '.png', lbl) 

# 需要导入模块: from scipy import misc [as 别名]
# 或者: from scipy.misc import imsave [as 别名]
def replace_eyes(image, out_eyes, out_shape, out_path, n):
    x_cen, y_cen, half_w, half_h = out_shape
    copy_image = np.copy(image)
    print(image.shape)
    print(out_eyes.shape) # 41,51
    out_eyes = np.squeeze(out_eyes, axis=0)

    # resize and save as eyes only
    # replace = cv2.resize(out_eyes, (51,41)) #(400, 250)
    save_path_and_name = os.path.join(out_path, '{}.jpg'.format(n))
    misc.imsave(save_path_and_name, out_eyes)

    resize_replace = cv2.resize(out_eyes, (2*half_w, 2*half_h)) * 255 # resize to original
    # resize_replace = np.transpose(resize_replace, axes=(1, 0, 2))
    copy_image[(y_cen - half_h):(y_cen + half_h), (x_cen - half_w):(x_cen + half_w), :] = resize_replace.astype(np.uint8)
    image_save_path_and_name = os.path.join(out_path, 'face_{}.jpg'.format(n))
    # print(image_save_path_and_name)
    misc.imsave(image_save_path_and_name, copy_image)
    return None 

# 需要导入模块: from scipy import misc [as 别名]
# 或者: from scipy.misc import imsave [as 别名]
def data_store(path,action,reward,state):

    if not os.path.exists(path):
        os.makedirs(path)
    else:
        shutil.rmtree(path)
        os.makedirs(path)
    
    df = pd.DataFrame(action, columns=["Steering", "Throttle", "Brake"])
    df["Reward"] = reward
    df.to_csv(path +'car_racing_actions_rewards.csv', index=False)
    
    for i in range(len(state)):
        if rgb_mode == False:
            image = rgb2gray(state[i])
        else:
            image = state[i]

    misc.imsave( path + "img" + str(i) +".png", image) 

# 需要导入模块: from scipy import misc [as 别名]
# 或者: from scipy.misc import imsave [as 别名]
def save_HR_LR(img, size, path, idx):
	HR_img = misc.imresize(img, size, interp='bicubic')
	HR_img = modcrop(HR_img, 4)
	rot180_img = misc.imrotate(HR_img, 180)
	x4_img = misc.imresize(HR_img, 1 / 4, interp='bicubic')
	x4_rot180_img = misc.imresize(rot180_img, 1 / 4, interp='bicubic')

	img_path = path.split('/')[-1].split('.')[0] + '_rot0_' + 'ds' + str(idx) + '.png'
	rot180img_path = path.split('/')[-1].split('.')[0] + '_rot180_' + 'ds' + str(idx) + '.png'
	x4_img_path = path.split('/')[-1].split('.')[0] + '_rot0_' + 'ds' + str(idx) + '.png'
	x4_rot180img_path = path.split('/')[-1].split('.')[0] + '_rot180_' + 'ds' + str(idx) + '.png'

	misc.imsave(save_HR_path + '/' + img_path, HR_img)
	misc.imsave(save_HR_path + '/' + rot180img_path, rot180_img)
	misc.imsave(save_LR_path + '/' + x4_img_path, x4_img)
	misc.imsave(save_LR_path + '/' + x4_rot180img_path, x4_rot180_img) 

# 需要导入模块: from scipy import misc [as 别名]
# 或者: from scipy.misc import imsave [as 别名]
def kernel_summary(sess):
  with sess.as_default():
    for layer in ["conv1", "conv2", "conv3"]:
      with tf.variable_scope(layer, reuse=True):
        weights = tf.get_variable('weights')
      kernels = tf.unpack(tf.transpose(weights, perm=[3,2,0,1]))
      for i,kernel in enumerate(kernels):
      #[12, 6, 6] -> 12 x [8, 8]
        padding = [[1,1], [1,1]]
        padded_kernels = [tf.pad(single_kernel, padding) for single_kernel in tf.unpack(kernel)]

      #12 x [8, 8] -> [6, 12 * 8]
        horizontally_concatenated = tf.concat(1, padded_kernels)

        image = horizontally_concatenated.eval()

        misc.imsave(layer + "_" + str(i) + ".png", image) 

# 需要导入模块: from scipy import misc [as 别名]
# 或者: from scipy.misc import imsave [as 别名]
def get_saliency_for_shallownet(image_url,sal_url):
    arr_files = glob.glob(image_url+"*.jpg")
    for i in range(len(arr_files)):  
        url_image = arr_files[i]
        image = io.imread(url_image)       
        img = misc.imresize(image,(96,96))
        img = np.asarray(img, dtype = 'float32') / 255.
        img = img.transpose(2,0,1).reshape(3, 96, 96)
        xt = np.zeros((1, 3, 96, 96), dtype='float32')
        xt[0]=img
        y = juntingnet.predict(xt)
        tmp = y.reshape(48,48)
        blured= ndimage.gaussian_filter(tmp, sigma=3)
        sal_map = cv2.resize(tmp,(image.shape[1],image.shape[0]))
        sal_map -= np.min(sal_map)
        sal_map /= np.max(sal_map)
        #saliency = misc.imresize(y,(img.shape[0],img.shape[1]))
        aux = url_image.split("/")[-1].split(".")[0]
        misc.imsave(sal_url+'/'+aux+'.png', sal_map) 

# 需要导入模块: from scipy import misc [as 别名]
# 或者: from scipy.misc import imsave [as 别名]
def dotplot2(s1, s2, wordsize=5, overlap=5, verbose=1):
        """ verbose = 0 (no progress), 1 (progress if s1 and s2 are long) or
        2 (progress in any case) """
        doProgress = False
        if verbose > 1 or len(s1)*len(s2) > 1e6:
            doProgress = True

        mat = numpy.ones(((len(s1)-wordsize)/overlap+2, (len(s2)-wordsize)/overlap+2))

        for i in range(0, len(s1)-wordsize, overlap):
            if i % 1000 == 0 and doProgress:
                logging.info("  dotplot progress: {} of {} rows done".format(i, len(s1)-wordsize))
            word1 = s1[i:i+wordsize]

            for j in range(0, len(s2)-wordsize, overlap):
                word2 = s2[j:j+wordsize]

                if word1 == word2 or word1 == word2[::-1]:
                    mat[i/overlap, j/overlap] = 0
        
        imgData = None
        tempDir = tempfile.mkdtemp()
        try:
            path = os.path.join(tempDir, "dotplot.png")
            misc.imsave(path, mat)
            imgData = open(path).read()
        except Exception as e:
            logging.error("Error generating dotplots:'{}'".format(e))
        finally:
            shutil.rmtree(tempDir)
        return imgData 

# 需要导入模块: from scipy import misc [as 别名]
# 或者: from scipy.misc import imsave [as 别名]
def save_images(images, size, image_path):
    return imsave(inverse_transform(images), size, image_path) 

# 需要导入模块: from scipy import misc [as 别名]
# 或者: from scipy.misc import imsave [as 别名]
def imsave(images, size, path):
    return misc.imsave(path, merge(images, size)) 

# 需要导入模块: from scipy import misc [as 别名]
# 或者: from scipy.misc import imsave [as 别名]
def get_representation(img, model, model_dir, out_path):
    saver = tf.train.Saver()

    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())

        ckpt = tf.train.get_checkpoint_state(model_dir)
        # print(ckpt)
        # print(ckpt.model_checkpoint_path)
        if ckpt and ckpt.model_checkpoint_path:
            saver.restore(sess, ckpt.model_checkpoint_path)

        Ax = tf.placeholder(tf.float32, [model.batch_size,model.height,model.width,model.channel],name='Ax')
        enc_Ax = model.splitter('encoder', Ax)
        grad_att_1 =[tf.gradients(enc_Ax[0][:,:,:,i], Ax)[0] for i in range(128)]
        grad_att_2 =[tf.gradients(enc_Ax[1][:,:,:,i], Ax)[0] for i in range(128)]
        grad_att_3 =[tf.gradients(enc_Ax[2][:,:,:,i], Ax)[0] for i in range(256)]
        # from IPython import embed;embed();exit()
        grad_1 = sess.run(grad_att_1, feed_dict={Ax: img})
        grad_2 = sess.run(grad_att_2, feed_dict={Ax: img})
        grad_3 = sess.run(grad_att_3, feed_dict={Ax: img})
        for i in range(128):
            misc.imsave(os.path.join(out_path, '0_{:03d}.jpg'.format(i)), grad_1[i][0])
            misc.imsave(os.path.join(out_path, '1_{:03d}.jpg'.format(i)), grad_2[i][0])
            np.save(os.path.join(out_path, '0_{:03d}.npy'.format(i)), grad_2[i][0])
            np.save(os.path.join(out_path, '1_{:03d}.npy'.format(i)), grad_2[i][0])

        for i in range(256):
            misc.imsave(os.path.join(out_path, '2_{:03d}.jpg'.format(i)), grad_3[i][0])
            np.save(os.path.join(out_path, '2_{:03d}.npy'.format(i)), grad_3[i][0]) 

# 需要导入模块: from scipy import misc [as 别名]
# 或者: from scipy.misc import imsave [as 别名]
def swap_attribute(src_img, att_img, swap_list, model_dir, model, gpu):
    '''
    Input
        src_img: the source image that you want to change its attribute
        att_img: the attribute image that has certain attribute
        swap_list: the swap id list
        model_dir: the directory that contains the checkpoint, ckpt.* files
        model: the DNA_GAN network that defined in train.py
        gpu: for example, '0,1'. Use '' for cpu mode
    Output
        out1: src_img with attributes
        out2: att_img without attributes
    '''
    os.environ["CUDA_VISIBLE_DEVICES"] = gpu
    saver = tf.train.Saver()

    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())

        ckpt = tf.train.get_checkpoint_state(model_dir)
        # print(ckpt)
        # print(ckpt.model_checkpoint_path)
        if ckpt and ckpt.model_checkpoint_path:
            saver.restore(sess, ckpt.model_checkpoint_path)

        Ax = tf.placeholder(tf.float32, [model.batch_size,model.height,model.width,model.channel],name='Ax')
        Be = tf.placeholder(tf.float32, [model.batch_size,model.height,model.width,model.channel],name='Be')

        enc_Ax = model.splitter('encoder', Ax)
        enc_Be = model.splitter('encoder', Be)
        enc_Ae, enc_Bx = model.swap_attribute(enc_Ax, enc_Be, swap_list)
        Ae = model.joiner('decoder', enc_Ae)
        Bx = model.joiner('decoder', enc_Bx)
        out2, out1 = sess.run([Ae, Bx], feed_dict={Ax: att_img, Be:src_img})
        swap = np.concatenate((src_img[0], att_img[0], out1[0], out2[0]), 1)
        misc.imsave('swap.jpg', swap)
        # misc.imsave('out1.jpg', out1[0])
        # misc.imsave('out2.jpg', out2[0]) 

# 需要导入模块: from scipy import misc [as 别名]
# 或者: from scipy.misc import imsave [as 别名]
def interpolation1_(src_img, att_img, inter_num, model_dir, model, gpu):
    '''
    Input
        src_img: the source image that you want to change its attribute
        att_img: the attribute image that has certain attribute
        inter_num: number of interpolation points
        model_dir: the directory that contains the checkpoint, ckpt.* files
        model: the DNA_GAN network that defined in train.py
        gpu: for example, '0,1'. Use '' for cpu mode
    Output
        out: [src_img, inter1, inter2, ..., inter_{inter_num}]
    '''
    os.environ["CUDA_VISIBLE_DEVICES"] = gpu
    saver = tf.train.Saver()

    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())

        ckpt = tf.train.get_checkpoint_state(model_dir)
        # print(ckpt)
        # print(ckpt.model_checkpoint_path)
        if ckpt and ckpt.model_checkpoint_path:
            saver.restore(sess, ckpt.model_checkpoint_path)

        B, src_feat = sess.run([model.B, model.e], feed_dict={model.Be: src_img})
        att_feat = sess.run(model.x, feed_dict={model.Ax: att_img})

        out = src_img[0]
        for i in range(1, inter_num + 1):
            lambda_i = i / float(inter_num)
            out_i = sess.run(model.joiner('G_joiner', B, src_feat + (att_feat - src_feat) * lambda_i) )
            out = np.concatenate((out, out_i[0]), axis=1)
        # print(out.shape)
        misc.imsave('interpolation2.jpg', out)