目标检测应用场景—数据集【NO.16】交通标志检测

 写在前面：数据集对应应用场景，不同的应用场景有不同的检测难点以及对应改进方法，本系列整理汇总领域内的数据集，方便大家下载数据集，若无法下载可关注后私信领取。关注免费领取整理好的数据集资料！今天分享一个非常好的非常小众的研究方向，有应用创新，可有利于发小论文和大论文，有需要的朋友可私信我，我这边可以辅导毕设、课程设计等。

中国交通标志识别数据库里的公开数据集：Chinese Traffic Signs_数据集-阿里云天池

话不多说直接上图：图像数据集样本图如下所示 ，分为58个类别  0: Speed limit 5
1: Speed limit 15
2: Speed limit 30
3: Speed limit 40
4: Speed limit 50
5: Speed limit 60
6: Speed limit 70
7: Speed limit 80
8: No straight ahead and left turn
9: No straight ahead and right turn
10: No straight ahead
11: No left turn
12: No left turn and right turn
13: No right turn
14: No overtaking
15: No U-turn
16: Except for buses
17: No horn
18: End of speed limit 40
19: End of speed limit 50
20: Straight ahead and right turn
21: Straight ahead
22: Left turn
23: Left turn or right turn
24: Right turn
25: Keep left
26: Keep right
27: Roundabout ahead
28: Motor vehicles only
29: Sound horn
30: Non-motorized vehicles
31: U-turn allowed
32: Left and right turn
33: Watch for traffic lights
34: Beware of danger
35: Watch for pedestrians
36: Beware of non-motorized vehicles
37: Watch for children
38: Beware of sharp right turn
39: Beware of sharp turn
40: Downhill road
41: Uphill road
42: Slow down
43: T-intersection
44: T-intersection
45: Village sign
46: Reverse curve
47: Unmanned railway crossing
48: Construction zone
49: Continuous curves
50: Manned railway crossing
51: Accident-prone area
52: Yield
53: No entry
54: No parking
55: No entry for vehicles
56: Slow down and yield
57: Stop for inspection。可以通过利用目标检测算法检测出交通标志的地方，可以应用在一些交通行业当中等。

 

 

 训练集的数据集一共4198张图片。原数据集格式为xml格式，已转换为YOLO格式。

非常好的非常小众的研究方向，有应用创新，可有利于发小论文和大论文，有需要的朋友可私信我。