带标注的麻将识别数据集，六千多张图片，识别率96.7%，可识别麻将的点数和类型，如1条，8萬，2饼東南西北中發白板等，支持yolo，coco json，pascal voc xml格式

带标注的麻将识别数据集六千多张图片识别率96.7%可识别麻将的点数和类型如1条8萬2饼東南西北中發白板等支持yolococo jsonpascal voc xml格式数据集比较大压缩后大约1.8G如果你想要1000张左右的图片可使用这个数据集带标注的麻将识别数据集识别率97.5%可识别麻将的点数和类型如1条8萬2饼東南西北中發白板等支持yolococo jsonpascal voc xml格式数据集详细介绍模型训练指标数据集拆分训练集6294图片验证集788图片测试集568图片预处理自动定向 应用增强每个训练样本的输出结果 3翻转 水平、垂直旋转角度 介于 -15° 和 15° 之间剪切 水平方向±15°垂直方向±15°色调 介于-25°和25°之间亮度 介于 -25% 和 25% 之间模糊 最大 2 像素标注信息说明饼是D 比如1饼就是1D条是B 比如2条就是2B幺鸡就是1B萬是C 比如九万就是9C发财GD白板WD東 EW南 SW西 WW北 NW红中 RD季节牌 春 1S 夏2S 秋 3S 冬 4S花卉牌 梅 1F、兰2F、竹 3F、菊 4F数据集图片和标注信息模型训练代码模型训练比较简单参考我的代码https://gitcode.net/pbymw8iwm/Python_299818这里不做详细说明里边包含了yolococo jsonpascal voc xml格式的训练以及推理代码。模型测试推理结果{predictions: [{x: 1248.5,y: 42,width: 45,height: 64,confidence: 0.873,class: 3B,class_id: 10,detection_id: c4c73e63-49b8-4dce-8bab-258f01e19562},{x: 358,y: 607.5,width: 68,height: 81,confidence: 0.813,class: 3D,class_id: 12,detection_id: dcca4cf9-0f92-4166-9efe-5a1e1376e110},{x: 844,y: 331.5,width: 52,height: 109,confidence: 0.795,class: 3B,class_id: 10,detection_id: d8dcd902-1bc3-4e34-a4b8-693ddf19f75b},{x: 792,y: 360,width: 48,height: 110,confidence: 0.766,class: 2B,class_id: 5,detection_id: 5fea1bf7-a9e5-4206-8b99-fe51e47a16e1},{x: 941.5,y: 290,width: 43,height: 100,confidence: 0.745,class: 5B,class_id: 20,detection_id: 65098f7b-33ab-4578-a275-08bfe190e30a},{x: 421.5,y: 587,width: 63,height: 118,confidence: 0.705,class: 5D,class_id: 22,detection_id: 1215458f-9b56-4591-8ec8-2df305537243},{x: 123,y: 265.5,width: 140,height: 43,confidence: 0.685,class: 9D,class_id: 34,detection_id: 72de7c1e-0737-4a75-9aca-479f5cd5e8cb},{x: 981,y: 254.5,width: 40,height: 103,confidence: 0.681,class: RD,class_id: 38,detection_id: 14db0843-ec04-4a29-a57c-692b32a5bd8d},{x: 895,y: 308,width: 48,height: 116,confidence: 0.665,class: 3B,class_id: 10,detection_id: a7b3a04b-5442-4962-affb-6e2e5cd92336},{x: 180,y: 240.5,width: 132,height: 41,confidence: 0.647,class: 9D,class_id: 34,detection_id: 216ab332-6ff6-4a8f-b1a8-7d3a1ce875b5},{x: 150.5,y: 314,width: 129,height: 42,confidence: 0.638,class: 9C,class_id: 33,detection_id: 7b6383e7-721a-4548-b90d-6d8451ad94ea},{x: 590,y: 477.5,width: 54,height: 109,confidence: 0.638,class: 2C,class_id: 6,detection_id: 1ba09df6-fdaa-4f3a-aa46-6fa980328838},{x: 535.5,y: 516,width: 53,height: 118,confidence: 0.635,class: 6D,class_id: 25,detection_id: 9ed5647a-2f26-4683-bbee-dee8715a05ac},{x: 478,y: 550,width: 54,height: 128,confidence: 0.634,class: 4D,class_id: 17,detection_id: 6ecd6ac3-7a7d-4e4f-a520-3399dbf05823},{x: 696,y: 422.5,width: 46,height: 105,confidence: 0.537,class: 4C,class_id: 16,detection_id: ec21fc14-bd8f-4f51-a28d-e7e46705144a},{x: 536,y: 516.5,width: 52,height: 123,confidence: 0.521,class: 4D,class_id: 17,detection_id: bed7467c-715a-4367-9eb3-b119626d0f51}]}推理结果{predictions: [{x: 2573.5,y: 1559.5,width: 225,height: 291,confidence: 0.9,class: 5B,class_id: 20,detection_id: 25b36a3f-85d4-493e-8711-6be07d9e69f7},{x: 610,y: 1548,width: 238,height: 310,confidence: 0.898,class: 6C,class_id: 24,detection_id: 17b5f4af-6c07-4be1-8599-9b2067c2e1e5},{x: 3307,y: 1556.5,width: 230,height: 315,confidence: 0.896,class: 4D,class_id: 17,detection_id: 6195e289-fc5f-4981-8bc6-bbb196745d4b},{x: 1107,y: 1556,width: 232,height: 306,confidence: 0.891,class: 8B,class_id: 29,detection_id: c49ac580-4376-4d51-b0dd-b4e0589e08b1},{x: 856,y: 1551,width: 234,height: 306,confidence: 0.889,class: 8B,class_id: 29,detection_id: 7754faf8-2372-40bf-ad3e-8c7bd930b8e4},{x: 3063,y: 1555,width: 228,height: 304,confidence: 0.888,class: 7D,class_id: 28,detection_id: 99d83bab-7147-479a-b2dd-94f24ea02ead},{x: 2091,y: 1555.5,width: 224,height: 303,confidence: 0.882,class: 7D,class_id: 28,detection_id: b74d6de8-1f25-4bb9-9264-5b947182c40d},{x: 2818.5,y: 1556,width: 237,height: 298,confidence: 0.88,class: 2B,class_id: 5,detection_id: 4eeab96a-accd-410e-959a-f2d4fdfd5e32},{x: 1849,y: 1557,width: 230,height: 304,confidence: 0.876,class: 1D,class_id: 2,detection_id: 69f5f0f8-86db-4ced-a2e4-6c8e4b35c341},{x: 3540,y: 1551,width: 230,height: 306,confidence: 0.868,class: 2B,class_id: 5,detection_id: 6f9766ee-88cb-4a07-8fa5-fb4cd0870b76},{x: 1607.5,y: 1554.5,width: 233,height: 307,confidence: 0.833,class: 2D,class_id: 7,detection_id: ee84345a-31fc-4d93-a911-663f3b394a74},{x: 366,y: 1550.5,width: 234,height: 313,confidence: 0.833,class: 3C,class_id: 11,detection_id: 038fed00-90b5-4c1e-8e09-3f7783ed3b46},{x: 1359.5,y: 1555.5,width: 243,height: 309,confidence: 0.823,class: 9D,class_id: 34,detection_id: dc74cf60-3a54-4507-9ee5-c8c657782f21},{x: 2329.5,y: 1556.5,width: 231,height: 301,confidence: 0.807,class: 3D,class_id: 12,detection_id: fde1d3f9-8f02-4344-b61d-31251a57a1d1}]}数据集下载地址yolo 12:https://download.csdn.net/download/pbymw8iwm/92747088yolo 11:https://download.csdn.net/download/pbymw8iwm/92747081yolo 5:https://download.csdn.net/download/pbymw8iwm/92747082yolo 8:https://download.csdn.net/download/pbymw8iwm/92747083yolo 9:https://download.csdn.net/download/pbymw8iwm/92747084coco jsonhttps://download.csdn.net/download/pbymw8iwm/92747085yolo26:https://download.csdn.net/download/pbymw8iwm/92747086yolo darknethttps://download.csdn.net/download/pbymw8iwm/92747087