图像数据增强的内容（可根据需要自定义选择）：
1.直方图均衡化
2.clahe自适应对比度直方图均衡化
3.白平衡
4.亮度增强
5.亮度，饱和度，对比度增强
6.去除图像上的高光部分
7.自适应亮度增强
8.随机遮挡
9.图像高斯模糊
10.压缩图像
# -*- coding: utf-8 -*-"""******不改变原始xml的一些数据增强方法 type 1-10*******把增强后的图像和xml一起放入新的文件夹rootpath:picture_xml原始路径savepath：picture_xml保存路径*******改变原始 xml的一些数据增强方法 type 11-15******修改图片的同时修改对应的xmlfile_path:传入类别的信息txt，最好和生成labelmap的顺序一致rootpath:picture_xml原始路径savepath：picture_xml保存路径11:自定义裁剪，图像大小 w,h，例如 w=400,h=60012：自定义平移，平移比例 w,h [0-1] 例如w=0.1,h=0,213：自定义缩放，调整图像大小 w,h,例如 w=400,h=60014：图像翻转15:图像任意旋转，传入旋转角度列表anglelist=[90,-90]"""import cv2import randomimport mathimport os,shutilimport numpy as npfrom PIL import Image, ImageStatfrom skimage import exposureimport matplotlib.pyplot as pltimport tensorlayer as tlfrom lxml.etree import Element, SubElement, tostringimport xml.etree.ElementTree as ETdef hisColor_Img(path):"""对图像直方图均衡化:param path: 图片路径:return: 直方图均衡化后的图像"""img = cv2.imread(path)ycrcb = cv2.cvtColor(img, cv2.COLOR_BGR2YCR_CB)channels = cv2.split(ycrcb)cv2.equalizeHist(channels[0], channels[0]) #equalizeHist(in,out)cv2.merge(channels, ycrcb)img_eq=cv2.cvtColor(ycrcb, cv2.COLOR_YCR_CB2BGR)return img_eqdef clahe_Img(path,ksize):""":param path: 图像路径:param ksize: 用于直方图均衡化的网格大小，默认为8:return: clahe之后的图像"""image = cv2.imread(path, cv2.IMREAD_COLOR)b, g, r = cv2.split(image)clahe = cv2.createCLAHE(clipLimit=2.0, tileGridSize=(ksize,ksize))b = clahe.apply(b)g = clahe.apply(g)r = clahe.apply(r)image = cv2.merge([b, g, r])return imagedef whiteBalance_Img(path):"""对图像白平衡处理"""img = cv2.imread(path)b, g, r = cv2.split(img)Y = 0.299 * r + 0.587 * g + 0.114 * bCr = 0.5 * r - 0.419 * g - 0.081 * bCb = -0.169 * r - 0.331 * g + 0.5 * bMr = np.mean(Cr)Mb = np.mean(Cb)Dr = np.var(Cr)Db = np.var(Cb)temp_arry = (np.abs(Cb - (Mb + Db * np.sign(Mb))) < 1.5 * Db) & (np.abs(Cr - (1.5 * Mr + Dr * np.sign(Mr))) < 1.5 * Dr)RL = Y * temp_arry# 选取候选白点数的最亮10%确定为最终白点，并选择其前10%中的最小亮度值L_list = list(np.reshape(RL, (RL.shape[0] * RL.shape[1],)).astype(np.int))hist_list = np.zeros(256)min_val = 0sum = 0for val in L_list:hist_list[val] += 1for l_val in range(255, 0, -1):sum += hist_list[l_val]if sum >= len(L_list) * 0.1:min_val = l_valbreak# 取最亮的前10%为最终的白点white_index = RL < min_valRL[white_index] = 0# 计算选取为白点的每个通道的增益b[white_index] = 0g[white_index] = 0r[white_index] = 0Y_max = np.max(RL)b_gain = Y_max / (np.sum(b) / np.sum(b > 0))g_gain = Y_max / (np.sum(g) / np.sum(g > 0))r_gain = Y_max / (np.sum(r) / np.sum(r > 0))b, g, r = cv2.split(img)b = b * b_gaing = g * g_gainr = r * r_gain# 溢出处理b[b > 255] = 255g[g > 255] = 255r[r > 255] = 255res_img = cv2.merge((b, g, r))return res_imgdef bright_Img(path,ga,flag):"""亮度增强 Tensorlayer:param ga: ga为gamma值，>1亮度变暗，<1亮度变亮:param flag:True: 亮度值为(1-ga,1+ga)False:亮度值为ga,默认为1:return: 亮度增强后的图像"""image = tl.vis.read_image(path)tenl_img = tl.prepro.brightness(image, gamma=ga, is_random=flag)return tenl_imgdef illumination_Img(path,ga,co,sa,flag):"""亮度,饱和度，对比度增强 Tensorlayer:param ga: ga为gamma值，>1亮度变暗，<1亮度变亮:param co: 对比度值，1为原始值:param sa: 饱和度值，1为原始值:param flag:True: 亮度值为(1-ga,1+ga)，对比度(1-co,1+co)，饱和度(1-sa,1+sa)False:亮度值为ga,对比度co,饱和度sa:return:增强后的结果"""image = tl.vis.read_image(path)tenl_img= tl.prepro.illumination(image, gamma=ga, contrast=co , saturation=sa, is_random=flag)return tenl_imgdef create_mask(imgpath):image = cv2.imread(imgpath, cv2.IMREAD_GRAYSCALE)_, mask = cv2.threshold(image, 200, 255, cv2.THRESH_BINARY)return maskdef xiufu_Img(imgpath,maskpath):"""去除图像上的高光部分"""src_ = cv2.imread(imgpath)mask = cv2.imread(maskpath, cv2.IMREAD_GRAYSCALE)#缩放因子(fx,fy)res_ = cv2.resize(src_,None,fx=0.6, fy=0.6, interpolation = cv2.INTER_CUBIC)mask = cv2.resize(mask,None,fx=0.6, fy=0.6, interpolation = cv2.INTER_CUBIC)dst = cv2.inpaint(res_, mask, 10, cv2.INPAINT_TELEA)return dstdef image_brightness(rgb_image):'''检测图像亮度(基于RMS)'''stat = ImageStat.Stat(rgb_image)r, g, b = stat.rmsreturn math.sqrt(0.241*(r**2) + 0.691*(g**2) + 0.068*(b**2))def calc_gamma(brightness):return brightness/127.0def image_gamma_transform(pil_im, gamma):image_arr = np.array(pil_im)image_arr2 = exposure.adjust_gamma(image_arr, gamma)if len(image_arr.shape) == 3: # 格式为(height(rows), weight(colums), 3)r = Image.fromarray(np.uint8(image_arr[:, :, 0]))g = Image.fromarray(np.uint8(image_arr[:, :, 1]))b = Image.fromarray(np.uint8(image_arr[:, :, 2]))image = Image.merge("RGB", (r, g, b))return imageelif len(image_arr.shape) == 2: # 格式为(height(rows), weight(colums))return Image.fromarray(np.uint8(image_arr))def autobright_Img(rootpath,savepath):"""自适应亮度增强"""list=os.listdir(rootpath)for i in range(0, len(list)):path = os.path.join(rootpath, list[i])if os.path.isfile(path):if list[i].endswith("jpg") or list[i].endswith("JPG") or list[i].endswith("png") or list[i].endswith("PNG"):print("adjust_bright running....")print(list[i])im = Image.open(path)brightness = image_brightness(im)newimage = np.array(image_gamma_transform(im, calc_gamma(brightness)))newname="adjust_bright"+list[i][:-4]saveflie=os.path.join(savepath, newname+".jpg")plt.imsave(saveflie, newimage)shutil.copyfile(os.path.join(rootpath, list[i][:-4] + ".xml"),os.path.join(savepath, newname + ".xml"))def probability_random_event(rate, event):"""随机变量的概率函数"""## 参数rate为list<int># 返回概率事件的下标索引start = 0index = 0randnum = random.randint(1, sum(rate))for index, scope in enumerate(rate):start += scopeif randnum <= start:breakreturn event[index]def erase_Img(root_path,save_path):"""随机遮挡"""for file in os.listdir(root_path):file_name, extension = os.path.splitext(file)if extension == '.xml':print("erase running....")print(file_name+".jpg")xml_path = os.path.join(root_path, file)tree = ET.parse(xml_path)root = tree.getroot()image_path = os.path.join(root_path,file_name+'.jpg')image = cv2.imread(image_path)for obj in root.findall('object'):#文件夹图像遮挡的比例，参数可修改is_erase = probability_random_event([7,3],[True,False])if is_erase:#boundingbox遮挡的方向，参数可修改erase_orientation = probability_random_event([6,2,1,1],['down', 'up', 'left','right'])#遮挡方块的大小，参数可修改erase_scope = random.uniform(0.1,0.3)xml_box = obj.find('bndbox')_xmin = int(xml_box.find('xmin').text)_xmax = int(xml_box.find('xmax').text)_ymin = int(xml_box.find('ymin').text)_ymax = int(xml_box.find('ymax').text)box_width = _xmax - _xminbox_height = _ymax - _yminnew_xmin, new_xmax, new_ymin, new_ymax = _xmin, _xmax, _ymin, _ymaxif erase_orientation == 'down':new_ymax = int(_ymax - box_height*erase_scope)image[new_ymax:_ymax,new_xmin:new_xmax,:] = 255if erase_orientation == 'up':new_ymin = int(_ymin + box_height*erase_scope)image[_ymin:new_ymin,new_xmin:new_xmax,:] = 255if erase_orientation == 'left':new_xmin = int(_xmin + box_width*erase_scope)image[new_ymin:new_ymax,_xmin:new_xmin,:] = 255if erase_orientation == 'right':new_xmax = int(_xmax - box_width*erase_scope)image[new_ymin:new_ymax,new_xmax:_xmax,:] = 255cv2.imwrite(os.path.join(save_path,"earse_"+file_name+'.jpg'),image)xml_box.find('xmin').text = str(new_xmin)xml_box.find('ymin').text = str(new_ymin)xml_box.find('xmax').text = str(new_xmax)xml_box.find('ymax').text = str(new_ymax)tree.write(os.path.join(save_path,"earse_"+file_name+'.xml'))def blur_Img(rootpath,savepath,ksize,new_rate):"""随机模糊图像"""img_list=[]for imgfiles in os.listdir(rootpath):if (imgfiles.endswith("jpg") or imgfiles.endswith("JPG")):img_list.append(imgfiles)filenumber = len(img_list)rate =new_rate # 自定义抽取文件夹中图片的比例，参数可修改picknumber = int(filenumber * rate)sample = random.sample(img_list, picknumber)for name in sample:print("blur running....")print(name)namepath=os.path.join(rootpath,name)ori_img = cv2.imread(namepath)size = random.choice(ksize) # 设置高斯核的大小，参数可修改，size>9，小于9图像没有变化kernel_size = (size, size)image = cv2.GaussianBlur(ori_img, ksize=kernel_size, sigmaX=0, sigmaY=0)cv2.imwrite(os.path.join(savepath,"blur_"+name), image)shutil.copyfile(os.path.join(rootpath, name.split(".")[0] + ".xml"),os.path.join(savepath, "blur_"+name.split(".")[0] + ".xml"))def compress_Img(infile_path, outfile_path,pic_size):"""压缩图像不改变图片尺寸压缩到指定大小:param infile: 压缩源文件:param outfile: 压缩文件保存地址"""count=0for infile in os.listdir(infile_path):if infile.endswith(".jpg") or infile.endswith(".GPG") or \infile.endswith(".jpeg") or infile.endswith(".GPEG"):print("compress_ running....")print(infile)filename, extend_name = os.path.splitext(infile)img_path=os.path.join(infile_path,infile)imgsaved_path=os.path.join(outfile_path,"compress_"+infile)img = cv2.imread(img_path, 1)# 获取文件大小:KBimg_size = os.path.getsize(img_path)/ 1024if img_size>pic_size:cv2.imwrite(imgsaved_path, img, [cv2.IMWRITE_JPEG_QUALITY, 30])shutil.copyfile(os.path.join(infile_path, filename + ".xml"),os.path.join(outfile_path, "compress_"+filename + ".xml"))else:shutil.copyfile(img_path,imgsaved_path)shutil.copyfile(os.path.join(infile_path, filename + ".xml"),os.path.join(outfile_path, "compress_"+filename + ".xml"))#######改变原始 xml的一些数据增强方法 type 11-15############def rotate_image(src, angle, scale=1):""":param src: 图像的地址:param angle: 旋转角度:param scale: 旋转范围:return: 旋转后的图像"""w = src.shape[1]h = src.shape[0]rangle = np.deg2rad(angle)nw = (abs(np.sin(rangle) * h) + abs(np.cos(rangle) * w)) * scalenh = (abs(np.cos(rangle) * h) + abs(np.sin(rangle) * w)) * scalerot_mat = cv2.getRotationMatrix2D((nw * 0.5, nh * 0.5), angle, scale)rot_move = np.dot(rot_mat, np.array([(nw - w) * 0.5, (nh - h) * 0.5, 0]))rot_mat[0, 2] += rot_move[0]rot_mat[1, 2] += rot_move[1]dst = cv2.warpAffine(src, rot_mat, (int(math.ceil(nw)), int(math.ceil(nh))), flags=cv2.INTER_LANCZOS4)return dstdef rotate_xml(src, xmin, ymin, xmax, ymax, angle, scale=1):"""旋转xml"""w = src.shape[1]h = src.shape[0]rangle = np.deg2rad(angle)# 获取旋转后图像的长和宽nw = (abs(np.sin(rangle)*h) + abs(np.cos(rangle)*w))*scalenh = (abs(np.cos(rangle)*h) + abs(np.sin(rangle)*w))*scalerot_mat = cv2.getRotationMatrix2D((nw*0.5, nh*0.5), angle, scale)rot_move = np.dot(rot_mat, np.array([(nw-w)*0.5, (nh-h)*0.5,0]))rot_mat[0, 2] += rot_move[0]rot_mat[1, 2] += rot_move[1]point1 = np.dot(rot_mat, np.array([xmin, ymin, 1]))point2 = np.dot(rot_mat, np.array([xmax, ymin, 1]))point3 = np.dot(rot_mat, np.array([xmax, ymax, 1]))point4 = np.dot(rot_mat, np.array([xmin, ymax, 1]))concat = np.vstack((point1, point2, point3, point4))concat = concat.astype(np.int32)rx, ry, rw, rh = cv2.boundingRect(concat)return rx, ry, rx+rw, ry+rhdef get_classdict(file_path):"""得到所有类别的字典,总类别:param file_path: 总类别的文本文件:return: 所有类别的列表，字典"""classes = []classes_dict = {}for line in open(file_path):classes.append(line.strip("\n"))for i in range(len(classes)):classes_dict[classes[i]]=i+1return classes,classes_dictdef trans_img(img,img_xml,classes_dict):"""把数据转换为 ann_list的格式,ann_list=[类别，位置信息【list】]:param img: img:param img_xml: img_xml:param classes_dict: 总类别字典:return: class_list:xml中的类别,ann_list=[位置信息【list】"""ann_list=[]class_list=[]tree = ET.parse(img_xml)root = tree.getroot()#找到图像的w,hsize = root.find('size')xml_width = int(size.find('width').text)xml_height = int(size.find('height').text)for obj in root.iter('object'):position=[]#类别信息xml_name = str(obj.find('name').text)if xml_name in classes_dict.keys():xml_class=classes_dict[xml_name]xml_box = obj.find('bndbox')_xmin = int(xml_box.find('xmin').text)_xmax = int(xml_box.find('xmax').text)_ymin = int(xml_box.find('ymin').text)_ymax = int(xml_box.find('ymax').text)c_x=((_xmin+_xmax)/2.0)/xml_widthc_y = ((_ymin + _ymax) / 2.0) / xml_heighto_w=(_xmax-_xmin)/xml_widtho_h = (_ymax - _ymin) / xml_heightposition.append(c_x)position.append(c_y)position.append(o_w)position.append(o_h)class_list.append(xml_class)ann_list.append(position)return class_list,ann_list#######type11-14:自定义裁剪，w,h>0，自定义平移，w,h:0-1，自定义缩放w,h >0，图像翻转 ###########def handle_img(rootpath,savepath,classes,classes_dict,img,img_xml,w,h,type):"""用tensorlayer 进行数据增强，包括裁剪，平移，缩放，翻转，得到增强后的图像和修改对应的xml,并保存在新的文件夹中:param rootpath: 图片+xml原始路径:param savepath: 图片+xml保存路径:param classes: 总类别:param classes_dict: 总类别词典:param img: 图像:param img_xml: img_xml:param w: 1：裁剪的尺寸w 2：平移比例 w 3：缩放的尺寸w:param h: 1：裁剪的尺寸h 2：平移比例 h 3：缩放的尺寸h:param type: 不同的数据增强方法。type=1:自定义裁剪，w,h>0 2：自定义平移，w,h:0-1 3：自定义缩放w,h >0 4：图像翻转:return: 直接保存数据增强后的图像+xml"""anns = []#得到总类别#sku包含的类别，归一化后得到坐标信息 ,is_random可以修改，默认裁剪中间img1=os.path.join(rootpath,img)img1_xml=os.path.join(rootpath,img_xml)cla, ann = trans_img(img1, img1_xml, classes_dict)image = tl.vis.read_image(img1)##随机裁剪图片######if type==11:im_crop, clas, coords = tl.prepro.obj_box_crop(image, cla,ann, wrg=w, hrg=h, is_rescale=True, is_center=True,is_random=False)tl.vis.save_image(im_crop,os.path.join(savepath,"crop_"+img))image = im_crop.copy()######位移#########if type ==12:im_shfit, clas, coords = tl.prepro.obj_box_shift(image, cla,ann, wrg=w, hrg=h,is_rescale=True, is_center=True, is_random=True)tl.vis.save_image(im_shfit,os.path.join(savepath,"shift_"+img))image = im_shfit.copy()#######调整图片大小 #####if type==13:clas = claim_resize, coords = tl.prepro.obj_box_imresize(image,coords=ann, size=[w, h], is_rescale=True)tl.vis.save_image(im_resize, os.path.join(savepath, "resize_" + img))image =im_resize.copy()#########图像翻转###########if type==14:clas = claim_flip, coords=tl.prepro.obj_box_horizontal_flip(image, coords=ann, is_rescale=True, is_center=True, is_random=False)tl.vis.save_image(im_flip, os.path.join(savepath, "flip_" + img))image = im_flip.copy()imh, imw = image.shape[0:2]# clas类别信息 anns新的坐标信息，im_crop裁剪的图片for i in range(len(coords)):pos = []x, y, x2, y2 = tl.prepro.obj_box_coord_centroid_to_upleft_butright(coords[i])x, y, x2, y2 = tl.prepro.obj_box_coord_scale_to_pixelunit([x, y, x2, y2], (imh, imw))pos.append(x),pos.append(y),pos.append(x2),pos.append(y2)anns.append(pos)###新建xml###########node_root = Element('annotation')node_folder = SubElement(node_root, 'folder')node_folder.text = '1'node_filename = SubElement(node_root, 'filename')if type==11:node_filename.text = "crop_" + imgif type==12:node_filename.text = "shift_" + imgif type==13:node_filename.text = "resize_" + imgif type==14:node_filename.text = "filp_" + imgnode_size = SubElement(node_root, 'size')node_width = SubElement(node_size, 'width')node_width.text = str(imw)node_height = SubElement(node_size, 'height')node_height.text = str(imh)node_depth = SubElement(node_size, 'depth')node_depth.text = '3'for i in range(len(clas)):node_object = SubElement(node_root, 'object')node_name = SubElement(node_object, 'name')node_name.text = str(classes[clas[i]-1])node_difficult = SubElement(node_object, 'difficult')node_difficult.text = '0'node_bndbox = SubElement(node_object, 'bndbox')node_xmin = SubElement(node_bndbox, 'xmin')node_xmin.text = str(anns[i][0])node_ymin = SubElement(node_bndbox, 'ymin')node_ymin.text = str(anns[i][1])node_xmax = SubElement(node_bndbox, 'xmax')node_xmax.text = str(anns[i][2])node_ymax = SubElement(node_bndbox, 'ymax')node_ymax.text = str(anns[i][3])xml = tostring(node_root, pretty_print=True) # 格式化显示，该换行的换行if type == 11:img_newxml=os.path.join(savepath,"crop_"+img_xml)if type == 12:img_newxml = os.path.join(savepath, "shift_" + img_xml)if type == 13:img_newxml = os.path.join(savepath, "resize_" + img_xml)if type == 14:img_newxml = os.path.join(savepath, "flip_" + img_xml)file_object = open(img_newxml, 'wb')file_object.write(xml)file_object.close()########type=5:旋转图像###################def route_pic_xml(rootpath,routepath,anglelist,rate):"""旋转图像和修改对应的xml,并保存在新的文件夹中:param rootpath: 原始路径:param routepath: 旋转后的路径:param anglelist: 旋转角度列表:param rate: 旋转图像的比例"""for angle in anglelist:pathDir = []for i in os.listdir(rootpath):if i.endswith("jpg") or i.endswith("JPG"):pathDir.append(i)new_rate =rate # 自定义抽取图片的比例picknumber = int(len(pathDir) * new_rate)print(picknumber)sample = random.sample(pathDir, picknumber)for name in sample:a, b = os.path.splitext(name) # 分离出文件名aimg = cv2.imread(os.path.join(rootpath, a + '.jpg'))rotated_img = rotate_image(img, angle)cv2.imwrite(os.path.join(routepath, str(angle) + "d"'_' + a + '.jpg'), rotated_img)print(a + ".jpg")tree = ET.parse(os.path.join(rootpath, a + '.xml'))root = tree.getroot()for box in root.iter('bndbox'):xmin = float(box.find('xmin').text)ymin = float(box.find('ymin').text)xmax = float(box.find('xmax').text)ymax = float(box.find('ymax').text)x, y, x1, y1 = rotate_xml(img, xmin, ymin, xmax, ymax, angle)box.find('xmin').text = str(x)box.find('ymin').text = str(y)box.find('xmax').text = str(x1)box.find('ymax').text = str(y1)tree.write(routepath +"/"+ str(angle) + "d" + '_' + a + '.xml')print(a + ".xml")if __name__=='__main__':#####根据type值选择不同的数据增强的方法###1.直方图均衡化 2.clahe自适应对比度直方图均衡化 3.白平衡 4.亮度增强 5.亮度，饱和度，对比度增强 6.去除图像上的高光部分########7.自适应亮度增强 8.随机遮挡 9.图像高斯模糊 10.压缩图像########11:自定义裁剪，图像大小 w,h，例如 w=400,height=600########12：自定义平移，平移比例 w,h [0-1] 例如w=0.10,h=0,2########13：自定义缩放，调整图像大小 w,h,例如 w=400,h=600########14：图像翻转########15:图像任意旋转，传入旋转角度列表anglelist=[90,-90]#####type=4rootpath = r"E:\data_set\WCP-data\test_wcp"savepath = r"E:\data_set\WCP-data\handle_img"#type 6 去除高光需要给mask蒙版的保存地址masksavepath = r"E:\data_set\WCP-data\mask"#type 11-15 需要传入 file_path:传入类别的信息txt，最好和生成labelmap的顺序一致file_path = r"E:\data_set\wcp.txt"#1.直方图均衡化 2.clahe自适应对比度直方图均衡化 3.白平衡 4.亮度增强 5.亮度，饱和度，对比度增强 6.去除图像上的高光部分if type in range(1,7):imgfiles = os.listdir(rootpath)for i in range(0, len(imgfiles)):path = os.path.join(rootpath, imgfiles[i])print(imgfiles[i])if os.path.isfile(path):if (imgfiles[i].endswith("jpg") or imgfiles[i].endswith("JPG")):#直方图均衡化if type==1:print("hiscolor running....")enhance_img=hisColor_Img(path)newname = 'hiscolor_' + imgfiles[i].split(".")[0]cv2.imwrite(os.path.join(savepath, newname + ".jpg"), enhance_img)#clahe自适应对比度直方图均衡化if type==2:print("clahe running....")#设置clahe直方图均衡化的网格大小ksize=8enhance_img=clahe_Img(path,ksize)newname = 'clahe_' + imgfiles[i].split(".")[0]cv2.imwrite(os.path.join(savepath, newname + ".jpg"), enhance_img)#白平衡if type==3:print("whiteBalance running....")enhance_img=whiteBalance_Img(path)newname = 'whiteBalance_' + imgfiles[i].split(".")[0]cv2.imwrite(os.path.join(savepath, newname + ".jpg"), enhance_img)#亮度增强if type==4:#亮度值，可改变，>1亮度变暗，<1亮度变亮#flag:True: 亮度值为(1-ga,1+ga) False:亮度值为ga,默认为1print("bright running....")ga=0.8flag=Falseenhance_img=bright_Img(path,ga,flag)newname = 'bright_' + imgfiles[i].split(".")[0]tl.vis.save_image(enhance_img, os.path.join(savepath, newname + ".jpg"))#亮度，对比度，饱和度增强if type==5:# ga: ga为亮度值，>1亮度变暗，<1亮度变亮# co: 对比度值，1为原始值# sa: 饱和度值，1为原始值# flag: True: 亮度值为(1 - ga, 1 + ga)，对比度(1 - co, 1 + co)，饱和度(1 - sa, 1 + sa)# False: 亮度值为ga, 对比度co, 饱和度saga = 1co=0.9sa=1flag = Falseprint("illumination running....")enhance_img= illumination_Img(path,ga,co,sa,flag)newname = 'illumination_' + imgfiles[i].split(".")[0]tl.vis.save_image(enhance_img, os.path.join(savepath, newname + ".jpg"))#去除高光if type==6:print("xiufu running....")maskpath = os.path.join(masksavepath, "mask_" + imgfiles[i])cv2.imwrite(maskpath, create_mask(path))enhance_img= xiufu_Img(path, maskpath)newname = 'xiufu_' + imgfiles[i].split(".")[0]cv2.imwrite(os.path.join(savepath, newname + ".jpg"), enhance_img)shutil.copyfile(os.path.join(rootpath, imgfiles[i].split(".")[0] + ".xml"),os.path.join(savepath, newname + ".xml"))#亮度自适应调节if type==7:autobright_Img(rootpath, savepath)#随机遮挡if type==8:erase_Img(rootpath, savepath)#高斯模糊if type==9:#设置高斯模糊核，越大越模糊ksize=(9,11,13,17)#设置模糊比例rate=0.5blur_Img(rootpath, savepath, ksize,rate)#压缩图像if type==10:#设置压缩图像的阈值，例如 pic_size=500，大于500的才压缩pic_size=500compress_Img(rootpath, savepath,pic_size)#自定义裁剪 or 自定义平移 or自定义缩放 or 图像翻转if type in range(11,15):#在type=1 or type=3 时，w,h尺寸大小，例如 w=600,h=800#在type=2时，w,h为平移的比例，例如w=0.1,h=0.2# w =800# h =1280w=0.1h=0.2classes,classes_dict=get_classdict(file_path)list = os.listdir(rootpath)for i in range(0, len(list)):path = os.path.join(rootpath, list[i])if os.path.isfile(path):if (list[i].endswith("jpg") or list[i].endswith("JPG")):img=list[i]img_xml=list[i].split(".")[0] + ".xml"im=Image.open(path)print(img)if im.size[0]>w and im.size[1]>h:handle_img(rootpath, savepath, classes, classes_dict,img, img_xml,w,h,type)#随机旋转if type==15:#旋转角度列表，旋转比例 例如 anglelist = [90, -90]anglelist =[90,]rate=0.5route_pic_xml(rootpath, savepath,anglelist,rate)
