YunYang1994/tensorflow-yolov3 IndexError: list index out of range 解决办法

原项目地址：YunYang1994/tensorflow-yolov3
报错描述：
详细描述：
寻找错误原因：
总结：

原项目地址：YunYang1994/tensorflow-yolov3

报错描述：

运行识别程序 video_demo.py 时，程序报错：

D:\Yolov3_Tensorflow\python\python.exe D:/Yolov3_Tensorflow/tensorflow-yolov3/video_demo.py
2019-10-10 08:26:58.675604: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2
2019-10-10 08:26:59.363176: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1411] Found device 0 with properties:
name: GeForce MX250 major: 6 minor: 1 memoryClockRate(GHz): 1.582
pciBusID: 0000:01:00.0
totalMemory: 2.00GiB freeMemory: 1.62GiB
2019-10-10 08:26:59.363377: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1490] Adding visible gpu devices: 0
2019-10-10 08:27:00.278345: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] Device interconnect StreamExecutor with strength 1 edge matrix:
2019-10-10 08:27:00.278458: I tensorflow/core/common_runtime/gpu/gpu_device.cc:977]      0
2019-10-10 08:27:00.278519: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990] 0:   N
2019-10-10 08:27:00.278730: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1103] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 1374 MB memory) -> physical GPU (device: 0, name: GeForce MX250, pci bus id: 0000:01:00.0, compute capability: 6.1)
2019-10-10 08:27:18.867149: W tensorflow/core/common_runtime/bfc_allocator.cc:215] Allocator (GPU_0_bfc) ran out of memory trying to allocate 1.13GiB. The caller indicates that this is not a failure, but may mean that there could be performance gains if more memory were available.
2019-10-10 08:27:18.894705: W tensorflow/core/common_runtime/bfc_allocator.cc:215] Allocator (GPU_0_bfc) ran out of memory trying to allocate 1.14GiB. The caller indicates that this is not a failure, but may mean that there could be performance gains if more memory were available.
2019-10-10 08:27:18.898060: W tensorflow/core/common_runtime/bfc_allocator.cc:215] Allocator (GPU_0_bfc) ran out of memory trying to allocate 2.13GiB. The caller indicates that this is not a failure, but may mean that there could be performance gains if more memory were available.
2019-10-10 08:27:19.032698: W tensorflow/core/common_runtime/bfc_allocator.cc:215] Allocator (GPU_0_bfc) ran out of memory trying to allocate 1.60GiB. The caller indicates that this is not a failure, but may mean that there could be performance gains if more memory were available.
2019-10-10 08:27:19.109551: W tensorflow/core/common_runtime/bfc_allocator.cc:215] Allocator (GPU_0_bfc) ran out of memory trying to allocate 1.56GiB. The caller indicates that this is not a failure, but may mean that there could be performance gains if more memory were available.
2019-10-10 08:27:19.123356: W tensorflow/core/common_runtime/bfc_allocator.cc:215] Allocator (GPU_0_bfc) ran out of memory trying to allocate 2.06GiB. The caller indicates that this is not a failure, but may mean that there could be performance gains if more memory were available.
Traceback (most recent call last):File "D:/Yolov3_Tensorflow/tensorflow-yolov3/video_demo.py", line 141, in <module>image = utils.draw_bbox(frame, bboxes)File "D:\Yolov3_Tensorflow\tensorflow-yolov3\core\utils.py", line 83, in draw_bboxbbox_color = colors[class_ind]
IndexError: list index out of rangeProcess finished with exit code 1

详细描述：

程序运行成功会打开识别或者视频流进行识别并画出识别框，在画面静止的状态下，运行貌似稳定并能准确显示识别框，当画面有较大幅度动作时，会立即出现上述报错。

寻找错误原因：

1、在检查的过程中偶然看见程序说 ./data/classes/voc.names 文件不存在，当时没在意。因为查看文件夹，voc.names 文件确实存在。

2、后来查看 video_demo.py 文件，看到两个参数，突然想起什么，input_size = 416 可能是识别时识别框预设大小，而num_classes = 80 依稀记得是识别分类的数量。

于是打开 voc.names 查看，发现里面只有20个类，而打开 coco.names ，发现有80个类，这应该多少有点问题：

3、于是打开 config.py 文件，将 __C.YOLO.CLASSES 参数由 "./data/classes/voc.names" 改成了 "./data/classes/coco.names"。

4、运行程序，没有再出现程序报错，识别也良好！

总结：

识别使用的权重文件一定要与我们识别类的 .names 文件相对应（猜测权重文件中记录的只是类序号，它转换成文字表示需要 .names 文件支持），同时我们在 video_demo.py 中填写的 num_classes 参数也要与 .names 文件中的类数量一致。
（我本次使用的权重文件为 yolov3_coco_demo.ckpt ）