下面是一个简单的流程,看起来并不复杂,使用一个图片初始化该类,然后根据需求将模型的输出结果prediction送入draw_instance_predictions函数即可,最后调用save保存图片即可。 visualizer = Visualizer(image, self.metadata, instance_mode=self.instance_mode) predictions = self.predictor(image) vis_output = visualizer....
draw_instance_predictions(outputs["instances"].to("cpu")) img = v.get_image()[:,:,[2,1,0]] img = Image.fromarray(img) plt.figure(figsize=(10, 10)) plt.imshow(img) 以下是来自测试集中未见图像的一些预测结果: 4. 评估阶段(Evaluation) 下面的代码用于评估所得模型。 cfg = get_cfg() ...
out = v.draw_instance_predictions(outputs["instances"].to("cpu")) result_image = out.get_image()[:, :, ::-1]# 保存目标检测后的结果save_dir ="image"os.makedirs(save_dir, exist_ok=True) save_path = os.path.join(save_dir,"result.jpg") cv2.imwrite(save_path, result_image)print(...
v = v.draw_instance_predictions(outputs['instances'].to('cpu')) plt.figure(figsize = (14, 10)) plt.imshow(cv2.cvtColor(v.get_image()[:, :, ::-1], cv2.COLOR_BGR2RGB)) plt.savefig('./output.jpg') 2. run_train.py 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 ...
# 可视化工具output=viz.draw_instance_predictions(predictions["instances"].to('cpu')) # 把 GPU 跑的数据结果放到 CPUcv2.imshow("Result",output.get_image()[:,:,::-1])cv2.waitKey()if__name__=='__main__':detector=Detector(model_type='OD') # 这里选择检测模式('OD': 目标检测;'IS':...
(cfg.DATASETS.TRAIN[0]),scale=1.2)out=v.draw_instance_predictions(outputs["instances"].to("cpu"))# cv2.imshow("Predictions", out.get_image()[:, :, ::-1])# cv2.waitKey(0) # Display until a key is pressedcv2.imwrite("./tmp/output.ObjectDetection.jpg",out.get_image()[:,:,::...
instance_mode=ColorMode.IMAGE ) v = v.draw_instance_predictions(outputs["instances"].to("cpu")) #Passing the predictions to CPU from the GPU cv2_imshow(v.get_image()[:, :, ::-1]) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. ...
v = v.draw_instance_predictions(outputs["instances"].to("cpu")) cv2_imshow(v.get_image()[:, :, ::-1]) 这是一个覆盖了预测的样本图像所得到的结果。 总结与思考 你可能已经阅读了我以前的教程,其中介绍了一个类似的对象检测框架,名为 MMdetection,它也构建在 pytorch 上。那么 Detectron2 和它相...
v = v.draw_instance_predictions(outputs["instances"].to("cpu")) cv2_imshow(v.get_image()[:, :, ::-1]) """ # Train on a custom dataset In this section, we show how to train an existing detectron2 model on a custom dataset in a new format. ...
out = v.draw_instance_predictions(outputs["instances"].to("cpu"))cv2_imshow(out.get_image()[:, :, ::-1]) Instance Segmentation 只要將模型定義檔更換為Instance segmentation model便可以了。 model_path = "COCO-InstanceSegmentation/mask_rcnn_X_101_32x8d_FPN_3x.yaml"cfg = get_cfg()cfg....