init update

.gitignore

@@ -0,0 +1,45 @@
+# 图片格式
+*.jpg
+*.jpeg
+*.png
+*.svg
+*.gif
+
+# 视频格式
+*.mp4
+*.avi
+.ipynb_checkpoints
+*/__pycache__
+
+# 日志格式
+*.log
+*.data
+*.txt
+*.csv
+
+# 参数文件
+*.yaml
+*.json
+
+# 压缩文件格式
+*.zip
+*.tar
+*.tar.gz
+*.rar
+
+# 字体格式
+*.ttc
+*.ttf
+*.otf
+
+*.pt
+*.db
+
+/flagged
+/run
+!requirements.txt
+!cls_name/*
+!model_config/*
+!img_example/*
+
+app copy.py

README.md

@@ -11,3 +11,5 @@ license: gpl-3.0
 ---
 
 Check out the configuration reference at https://huggingface.co/docs/hub/spaces#reference
+
+🚀 项目主页：https://gitee.com/CV_Lab/gradio_yolov5_det

__init__.py

@@ -0,0 +1,2 @@
+__author__ = "曾逸夫（Zeng Yifu）"
+__email__ = "zyfiy1314@163.com"

app.py

@@ -0,0 +1,387 @@
+# Gradio YOLOv5 Det v0.2
+# 创建人：曾逸夫
+# 创建时间：2022-05-01
+# email：zyfiy1314@163.com
+# 项目主页：https://gitee.com/CV_Lab/gradio_yolov5_det
+
+import argparse
+import csv
+import json
+import sys
+from pathlib import Path
+
+import gradio as gr
+import torch
+import yaml
+from PIL import Image, ImageDraw, ImageFont
+
+from util.fonts_opt import is_fonts
+from util.pdf_opt import pdf_generate
+
+ROOT_PATH = sys.path[0]  # 根目录
+
+# 模型路径
+model_path = "ultralytics/yolov5"
+
+# 模型名称临时变量
+model_name_tmp = ""
+
+# 设备临时变量
+device_tmp = ""
+
+# 文件后缀
+suffix_list = [".csv", ".yaml"]
+
+# 字体大小
+FONTSIZE = 25
+
+
+def parse_args(known=False):
+    parser = argparse.ArgumentParser(description="Gradio YOLOv5 Det v0.2")
+    parser.add_argument(
+        "--model_name", "-mn", default="yolov5s", type=str, help="model name"
+    )
+    parser.add_argument(
+        "--model_cfg",
+        "-mc",
+        default="./model_config/model_name_p5_all.yaml",
+        type=str,
+        help="model config",
+    )
+    parser.add_argument(
+        "--cls_name",
+        "-cls",
+        default="./cls_name/cls_name.yaml",
+        type=str,
+        help="cls name",
+    )
+    parser.add_argument(
+        "--nms_conf",
+        "-conf",
+        default=0.5,
+        type=float,
+        help="model NMS confidence threshold",
+    )
+    parser.add_argument(
+        "--nms_iou", "-iou", default=0.45, type=float, help="model NMS IoU threshold"
+    )
+
+    parser.add_argument(
+        "--label_dnt_show",
+        "-lds",
+        action="store_true",
+        default=False,
+        help="label show",
+    )
+    parser.add_argument(
+        "--device", "-dev", default="cpu", type=str, help="cuda or cpu",
+    )
+    parser.add_argument(
+        "--inference_size", "-isz", default=640, type=int, help="model inference size"
+    )
+
+    args = parser.parse_known_args()[0] if known else parser.parse_args()
+    return args
+
+
+# yaml文件解析
+def yaml_parse(file_path):
+    return yaml.safe_load(open(file_path, encoding="utf-8").read())
+
+
+# yaml csv 文件解析
+def yaml_csv(file_path, file_tag):
+    file_suffix = Path(file_path).suffix
+    if file_suffix == suffix_list[0]:
+        # 模型名称
+        file_names = [i[0] for i in list(csv.reader(open(file_path)))]  # csv版
+    elif file_suffix == suffix_list[1]:
+        # 模型名称
+        file_names = yaml_parse(file_path).get(file_tag)  # yaml版
+    else:
+        print(f"{file_path}格式不正确！程序退出！")
+        sys.exit()
+
+    return file_names
+
+
+#  模型加载
+def model_loading(model_name, device):
+
+    # 加载模型
+    model = torch.hub.load(
+        model_path, model_name, force_reload=True, device=device, _verbose=False
+    )
+
+    return model
+
+
+# 检测信息
+def export_json(results, model, img_size):
+
+    return [
+        [
+            {
+                "id": i,
+                "class": int(result[i][5]),
+                # "class_name": model.model.names[int(result[i][5])],
+                "class_name": model_cls_name_cp[int(result[i][5])],
+                "normalized_box": {
+                    "x0": round(result[i][:4].tolist()[0], 6),
+                    "y0": round(result[i][:4].tolist()[1], 6),
+                    "x1": round(result[i][:4].tolist()[2], 6),
+                    "y1": round(result[i][:4].tolist()[3], 6),
+                },
+                "confidence": round(float(result[i][4]), 2),
+                "fps": round(1000 / float(results.t[1]), 2),
+                "width": img_size[0],
+                "height": img_size[1],
+            }
+            for i in range(len(result))
+        ]
+        for result in results.xyxyn
+    ]
+
+
+# 帧转换
+def pil_draw(img, countdown_msg, textFont, xyxy, font_size, label_opt):
+
+    img_pil = ImageDraw.Draw(img)
+
+    img_pil.rectangle(xyxy, fill=None, outline="green")  # 边界框
+
+    if label_opt:
+        text_w, text_h = textFont.getsize(countdown_msg)  # 标签尺寸
+        img_pil.rectangle(
+            (xyxy[0], xyxy[1], xyxy[0] + text_w, xyxy[1] + text_h),
+            fill="green",
+            outline="green",
+        )  # 标签背景
+        img_pil.multiline_text(
+            (xyxy[0], xyxy[1]),
+            countdown_msg,
+            fill=(205, 250, 255),
+            font=textFont,
+            align="center",
+        )
+
+    return img
+
+
+# YOLOv5图片检测函数
+def yolo_det(
+    img, device, model_name, inference_size, conf, iou, label_opt, model_cls, opt
+):
+
+    global model, model_name_tmp, device_tmp
+
+    if model_name_tmp != model_name:
+        # 模型判断，避免反复加载
+        model_name_tmp = model_name
+        model = model_loading(model_name_tmp, device)
+    elif device_tmp != device:
+        device_tmp = device
+        model = model_loading(model_name_tmp, device)
+
+    # -----------模型调参-----------
+    model.conf = conf  # NMS 置信度阈值
+    model.iou = iou  # NMS IOU阈值
+    model.max_det = 1000  # 最大检测框数
+    model.classes = model_cls  # 模型类别
+
+    results = model(img, size=inference_size)  # 检测
+
+    img_size = img.size  # 帧尺寸
+
+    # 加载字体
+    textFont = ImageFont.truetype(str(f"{ROOT_PATH}/fonts/SimSun.ttf"), size=FONTSIZE)
+
+    det_img = img.copy()
+
+    for result in results.xyxyn:
+        for i in range(len(result)):
+            id = int(i)  # 实例ID
+            obj_cls_index = int(result[i][5])  # 类别索引
+            obj_cls = model_cls_name_cp[obj_cls_index]  # 类别
+
+            # ------------边框坐标------------
+            x0 = float(result[i][:4].tolist()[0])
+            y0 = float(result[i][:4].tolist()[1])
+            x1 = float(result[i][:4].tolist()[2])
+            y1 = float(result[i][:4].tolist()[3])
+
+            # ------------边框实际坐标------------
+            x0 = int(img_size[0] * x0)
+            y0 = int(img_size[1] * y0)
+            x1 = int(img_size[0] * x1)
+            y1 = int(img_size[1] * y1)
+
+            conf = float(result[i][4])  # 置信度
+            # fps = f"{(1000 / float(results.t[1])):.2f}"  # FPS
+
+            det_img = pil_draw(
+                img,
+                f"{id}-{obj_cls}:{conf:.2f}",
+                textFont,
+                [x0, y0, x1, y1],
+                FONTSIZE,
+                label_opt,
+            )
+
+    det_json = export_json(results, model, img.size)[0]  # 检测信息
+
+    # JSON格式化
+    det_json_format = json.dumps(
+        det_json, sort_keys=True, indent=4, separators=(",", ":"), ensure_ascii=False
+    )
+
+    # -------pdf-------
+    report = "./Det_Report.pdf"
+    if "pdf" in opt:
+        pdf_generate(f"{det_json_format}", report)
+    else:
+        report = None
+
+    if "json" not in opt:
+        det_json = None
+
+    return det_img, det_json, report
+
+
+def main(args):
+    gr.close_all()
+
+    global model, model_cls_name_cp
+
+    slider_step = 0.05  # 滑动步长
+
+    nms_conf = args.nms_conf
+    nms_iou = args.nms_iou
+    label_opt = args.label_dnt_show
+    model_name = args.model_name
+    model_cfg = args.model_cfg
+    cls_name = args.cls_name
+    device = args.device
+    inference_size = args.inference_size
+
+    is_fonts(f"{ROOT_PATH}/fonts")  # 检查字体文件
+
+    # 模型加载
+    model = model_loading(model_name, device)
+
+    model_names = yaml_csv(model_cfg, "model_names")
+    model_cls_name = yaml_csv(cls_name, "model_cls_name")
+
+    model_cls_name_cp = model_cls_name.copy()  # 类别名称
+
+    # -------------------输入组件-------------------
+    inputs_img = gr.inputs.Image(type="pil", label="原始图片")
+    inputs_device = gr.inputs.Dropdown(
+        choices=["0", "cpu"], default=device, type="value", label="设备"
+    )
+    inputs_model = gr.inputs.Dropdown(
+        choices=model_names, default=model_name, type="value", label="模型"
+    )
+    inputs_size = gr.inputs.Radio(
+        choices=[320, 640], default=inference_size, label="推理尺寸"
+    )
+    input_conf = gr.inputs.Slider(
+        0, 1, step=slider_step, default=nms_conf, label="置信度阈值"
+    )
+    inputs_iou = gr.inputs.Slider(
+        0, 1, step=slider_step, default=nms_iou, label="IoU 阈值"
+    )
+    inputs_label = gr.inputs.Checkbox(default=(not label_opt), label="标签显示")
+    inputs_clsName = gr.inputs.CheckboxGroup(
+        choices=model_cls_name, default=model_cls_name, type="index", label="类别"
+    )
+    inputs_opt = gr.inputs.CheckboxGroup(
+        choices=["pdf", "json"], default=["pdf"], type="value", label="操作"
+    )
+
+    # 输入参数
+    inputs = [
+        inputs_img,  # 输入图片
+        inputs_device,  # 设备
+        inputs_model,  # 模型
+        inputs_size,  # 推理尺寸
+        input_conf,  # 置信度阈值
+        inputs_iou,  # IoU阈值
+        inputs_label,  # 标签显示
+        inputs_clsName,  # 类别
+        inputs_opt,  # 检测操作
+    ]
+
+    # 输出参数
+    outputs_img = gr.outputs.Image(type="pil", label="检测图片")
+    outputs02_json = gr.outputs.JSON(label="检测信息")
+    outputs03_pdf = gr.outputs.File(label="下载检测报告")
+
+    outputs = [outputs_img, outputs02_json, outputs03_pdf]
+
+    # 标题
+    title = "基于Gradio的YOLOv5通用目标检测系统v0.2"
+    # 描述
+    description = "<div align='center'>可自定义目标检测模型、安装简单、使用方便</div><div align='center'>Customizable object detection model, easy to install and easy to use</div>"
+
+    # 示例图片
+    examples = [
+        [
+            "./img_example/bus.jpg",
+            "cpu",
+            "yolov5s",
+            640,
+            0.6,
+            0.5,
+            True,
+            ["人", "公交车"],
+            ["pdf"],
+        ],
+        [
+            "./img_example/Millenial-at-work.jpg",
+            "cpu",
+            "yolov5l",
+            320,
+            0.5,
+            0.45,
+            True,
+            ["人", "椅子", "杯子", "笔记本电脑"],
+            ["json"],
+        ],
+        [
+            "./img_example/zidane.jpg",
+            "cpu",
+            "yolov5m",
+            640,
+            0.25,
+            0.5,
+            False,
+            ["人", "领带"],
+            ["pdf", "json"],
+        ],
+    ]
+
+    # 接口
+    gr.Interface(
+        fn=yolo_det,
+        inputs=inputs,
+        outputs=outputs,
+        title=title,
+        description=description,
+        examples=examples,
+        theme="seafoam",
+        # live=True, # 实时变更输出
+        flagging_dir="run",  # 输出目录
+        # flagging_options=["good", "generally", "bad"],
+        # allow_flagging="auto",
+        # ).launch(inbrowser=True, auth=['admin', 'admin'])
+    ).launch(
+        inbrowser=True,  # 自动打开默认浏览器
+        show_tips=True,  # 自动显示gradio最新功能
+        # favicon_path="./icon/logo.ico",
+    )
+
+
+if __name__ == "__main__":
+    args = parse_args()
+    main(args)

cls_name/cls_name.csv

@@ -0,0 +1,80 @@
+人
+自行车
+汽车
+摩托车
+飞机
+公交车
+火车
+卡车
+船
+红绿灯
+消防栓
+停止标志
+停车收费表
+长凳
+鸟
+猫
+狗
+马
+羊
+牛
+象
+熊
+斑马
+长颈鹿
+背包
+雨伞
+手提包
+领带
+手提箱
+飞盘
+滑雪板
+单板滑雪
+运动球
+风筝
+棒球棒
+棒球手套
+滑板
+冲浪板
+网球拍
+瓶子
+红酒杯
+杯子
+叉子
+刀
+勺
+碗
+香蕉
+苹果
+三明治
+橙子
+西兰花
+胡萝卜
+热狗
+比萨
+甜甜圈
+蛋糕
+椅子
+长椅
+盆栽
+床
+餐桌
+马桶
+电视
+笔记本电脑
+鼠标
+遥控器
+键盘
+手机
+微波炉
+烤箱
+烤面包机
+洗碗槽
+冰箱
+书
+时钟
+花瓶
+剪刀
+泰迪熊
+吹风机
+牙刷

cls_name/cls_name.yaml

@@ -0,0 +1,7 @@
+model_cls_name: ['人', '自行车', '汽车', '摩托车', '飞机', '公交车', '火车', '卡车', '船', '红绿灯', '消防栓', '停止标志',
+                '停车收费表', '长凳', '鸟', '猫', '狗', '马', '羊', '牛', '象', '熊', '斑马', '长颈鹿', '背包', '雨伞', '手提包', '领带',
+                '手提箱', '飞盘', '滑雪板', '单板滑雪', '运动球', '风筝', '棒球棒', '棒球手套', '滑板', '冲浪板', '网球拍', '瓶子', '红酒杯',
+                '杯子', '叉子', '刀', '勺', '碗', '香蕉', '苹果', '三明治', '橙子', '西兰花', '胡萝卜', '热狗', '比萨', '甜甜圈', '蛋糕',
+                '椅子', '长椅', '盆栽', '床', '餐桌', '马桶', '电视', '笔记本电脑', '鼠标', '遥控器', '键盘', '手机', '微波炉', '烤箱',
+                '烤面包机', '洗碗槽', '冰箱', '书', '时钟', '花瓶', '剪刀', '泰迪熊', '吹风机', '牙刷'
+            ]

model_config/model_name_p5_all.csv

@@ -0,0 +1,5 @@
+yolov5n
+yolov5s
+yolov5m
+yolov5l
+yolov5x

model_config/model_name_p5_all.yaml

@@ -0,0 +1 @@
+model_names: ["yolov5n", "yolov5s", "yolov5m", "yolov5l", "yolov5x"]

model_config/model_name_p5_n.csv

@@ -0,0 +1 @@
+yolov5n

model_config/model_name_p5_n.yaml

@@ -0,0 +1 @@
+model_names: ["yolov5n"]

model_config/model_name_p6_all.csv

@@ -0,0 +1,5 @@
+yolov5n6
+yolov5s6
+yolov5m6
+yolov5l6
+yolov5x6

model_config/model_name_p6_all.yaml

@@ -0,0 +1 @@
+model_names: ["yolov5n6", "yolov5s6", "yolov5m6", "yolov5l6", "yolov5x6"]

model_download/yolov5_model_p5_all.sh

@@ -0,0 +1,8 @@
+cd ./yolov5
+
+# 下载YOLOv5模型
+wget -c -t 0 https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5n.pt
+wget -c -t 0 https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5s.pt
+wget -c -t 0 https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5m.pt
+wget -c -t 0 https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5l.pt
+wget -c -t 0 https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5x.pt

model_download/yolov5_model_p5_n.sh

@@ -0,0 +1,4 @@
+cd ./yolov5
+
+# 下载YOLOv5模型
+wget -c -t 0 https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5n.pt

model_download/yolov5_model_p6_all.sh

@@ -0,0 +1,8 @@
+cd ./yolov5
+
+# 下载YOLOv5模型
+wget -c -t 0 https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5n6.pt
+wget -c -t 0 https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5s6.pt
+wget -c -t 0 https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5m6.pt
+wget -c -t 0 https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5l6.pt
+wget -c -t 0 https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5x6.pt

requirements.txt

@@ -0,0 +1,39 @@
+# Base ----------------------------------------
+matplotlib>=3.2.2
+numpy>=1.18.5
+opencv-python-headless>=4.5.5.64
+Pillow>=7.1.2
+PyYAML>=5.3.1
+requests>=2.23.0
+scipy>=1.4.1
+torch>=1.7.0
+torchvision>=0.8.1
+gradio==2.9b40
+tqdm>=4.41.0
+wget>=3.2
+rich>=12.2.0
+fpdf>=1.7.2
+
+# Logging -------------------------------------
+tensorboard>=2.4.1
+# wandb
+
+# Plotting ------------------------------------
+pandas>=1.1.4
+seaborn>=0.11.0
+
+# Export --------------------------------------
+# coremltools>=4.1  # CoreML export
+# onnx>=1.9.0  # ONNX export
+# onnx-simplifier>=0.3.6  # ONNX simplifier
+# scikit-learn==0.19.2  # CoreML quantization
+# tensorflow>=2.4.1  # TFLite export
+# tensorflowjs>=3.9.0  # TF.js export
+# openvino-dev  # OpenVINO export
+
+# Extras --------------------------------------
+# albumentations>=1.0.3
+# Cython  # for pycocotools https://github.com/cocodataset/cocoapi/issues/172
+# pycocotools>=2.0  # COCO mAP
+# roboflow
+thop  # FLOPs computation

util/fonts_opt.py

@@ -0,0 +1,67 @@
+# 字体管理
+# 创建人：曾逸夫
+# 创建时间：2022-05-01
+
+import os
+import sys
+from pathlib import Path
+
+import wget
+from rich.console import Console
+
+ROOT_PATH = sys.path[0]  # 项目根目录
+
+fonts_list = ["SimSun.ttf"]  # 字体列表
+fonts_suffix = ["ttc", "ttf", "otf"]  # 字体后缀
+
+data_url_dict = {
+    "SimSun.ttf": "https://gitee.com/CV_Lab/gradio_yolov5_det/attach_files/1053539/download/SimSun.ttf"
+}
+
+console = Console()
+
+
+# 创建字体库
+def add_fronts(font_diff):
+
+    global font_name
+
+    for k, v in data_url_dict.items():
+        if k in font_diff:
+            font_name = v.split("/")[-1]  # 字体名称
+            Path(f"{ROOT_PATH}/fonts").mkdir(parents=True, exist_ok=True)  # 创建目录
+
+            file_path = f"{ROOT_PATH}/fonts/{font_name}"  # 字体路径
+
+            try:
+                # 下载字体文件
+                wget.download(v, file_path)
+            except Exception as e:
+                print("路径错误！程序结束！")
+                print(e)
+                sys.exit()
+            else:
+                print()
+                console.print(
+                    f"{font_name} [bold green]字体文件下载完成！[/bold green] 已保存至：{file_path}"
+                )
+
+
+# 判断字体文件
+def is_fonts(fonts_dir):
+    if os.path.isdir(fonts_dir):
+        # 如果字体库存在
+        f_list = os.listdir(fonts_dir)  # 本地字体库
+
+        font_diff = list(set(fonts_list).difference(set(f_list)))
+
+        if font_diff != []:
+            # 字体不存在
+            console.print("[bold red]字体不存在，正在加载。。。[/bold red]")
+            add_fronts(font_diff)  # 创建字体库
+        else:
+            console.print(f"{fonts_list}[bold green]字体已存在！[/bold green]")
+    else:
+        # 字体库不存在，创建字体库
+        console.print("[bold red]字体库不存在，正在创建。。。[/bold red]")
+        add_fronts(fonts_list)  # 创建字体库

util/pdf_opt.py

@@ -0,0 +1,77 @@
+# PDF管理
+# 创建人：曾逸夫
+# 创建时间：2022-05-05
+
+from fpdf import FPDF
+
+
+title = "Gradio YOLOv5 Det v0.2"
+
+
+class PDF(FPDF):
+# 参考：https://pyfpdf.readthedocs.io/en/latest/Tutorial/index.html
+    def header(self):
+        # 设置中文字体
+        self.add_font("SimSun", "", "./fonts/SimSun.ttf", uni=True)
+        self.set_font("SimSun", "", 16)
+        # Calculate width of title and position
+        w = self.get_string_width(title) + 6
+        self.set_x((210 - w) / 2)
+        # Colors of frame, background and text
+        self.set_draw_color(255, 255, 255)
+        self.set_fill_color(255, 255, 255)
+        self.set_text_color(0, 0, 0)
+        # Thickness of frame (1 mm)
+        # self.set_line_width(1)
+        # Title
+        self.cell(w, 9, title, 1, 1, "C", 1)
+        # Line break
+        self.ln(10)
+
+    def footer(self):
+        # Position at 1.5 cm from bottom
+        self.set_y(-15)
+        # 设置中文字体
+        self.add_font("SimSun", "", "./fonts/SimSun.ttf", uni=True)
+        self.set_font("SimSun", "", 12)
+        # Text color in gray
+        self.set_text_color(128)
+        # Page number
+        self.cell(0, 10, "Page " + str(self.page_no()), 0, 0, "C")
+
+    def chapter_title(self, num, label):
+        # 设置中文字体
+        self.add_font("SimSun", "", "./fonts/SimSun.ttf", uni=True)
+        self.set_font("SimSun", "", 12)
+        # Background color
+        self.set_fill_color(200, 220, 255)
+        # Title
+        # self.cell(0, 6, 'Chapter %d : %s' % (num, label), 0, 1, 'L', 1)
+        self.cell(0, 6, f"检测结果：", 0, 1, "L", 1)
+        # Line break
+        self.ln(4)
+
+    def chapter_body(self, name):
+
+        # 设置中文字体
+        self.add_font("SimSun", "", "./fonts/SimSun.ttf", uni=True)
+        self.set_font("SimSun", "", 12)
+        # Output justified text
+        self.multi_cell(0, 5, name)
+        # Line break
+        self.ln()
+        self.cell(0, 5, "--------------------------------------")
+
+    def print_chapter(self, num, title, name):
+        self.add_page()
+        self.chapter_title(num, title)
+        self.chapter_body(name)
+
+
+# pdf生成函数
+def pdf_generate(input_file, output_file):
+    pdf = PDF()
+    pdf.set_title(title)
+    pdf.set_author("Zeng Yifu")
+    pdf.print_chapter(1, "A RUNAWAY REEF", input_file)
+    pdf.output(output_file)