Object Detection & Segmentation with RabbitDetect

Object detection and segmentation tutorial with rabbitdetect dataset(update to 2023b version addon YOLOX SOLOv2)

RabbitDetect

Bulit on 2022/02 by Fred Liu
Major update 2023.05.17
New update 2023.12.07(YOLOX,SOLOv2)
Youtube Link

版本:MATALB: update to 2023b,minimum vervion 2022a.
需要工具箱: Deeplearning , Image Processing, Computer Vision, Parallel Computing
需要支援包: YOLOv3,YOLOv4 Package & pretrain modle Package
Computer Vision Toolbox Model for YOLO v3 Object Detection
Computer Vision Toolbox Model for YOLO v4 Object Detection
YOLOX:
Computer Vision Toolbox Automated Visual Inspection Library
MASK-RCNN
Computer Vision Toolbox Model for Mask R-CNN Instance Segmentation
SOLOv2
Computer Vision Toolbox Model for SOLOv2 Instance Segmentation

---

首先請閱讀setup_readme.m (First to read setup_readme )

因為內建資料庫資料較少，因此在訓練一些模型上效果可能較差，範例提供整體流程，但實作請換較大型的資料庫使用。 Due to the limited amount of data in the built-in database, the performance of some models may be poorer during training. The example provides the overall process, but for implementation, it is recommended to use a larger database.

基於MATLAB 物件偵測於rabbit dataset
(MATLAB Object Detection with rabbit dataset)

---

1.資料請下載(data download):Rabbit_myself_416.zip or Rabbit_myself_608.zip
2.已訓練模型(Pretain_model):model\Modeldownload
3.演算法(algorithm):src_main\FasterRCNN,SSD,YLOLv2,YOLOv3,YOLOv4,YOLOX
4.標記檔案(label data):Rabbit_myself_608.mat
5.src_input: XMLinput , Jsoninput

---

使用流程(Use the process):

---

1.首先請閱讀setup_readme.m (First to read setup_readme )

2.標記影像：可使用image labeler標記　or 載入Rabbit_myselft_608標記資料(可使用Change_gTruthPath.m)
(Label Image:use image labeler to label or download "Rabbit_myselft_608" label dataset.
It can use "Change_gTruthPath.m" to change path.)

3.模型：可以自行透過演算法訓練(src_main)，也使用Pre-trained進行測試(model)
(Model:you can train through the algorithm by yourself, and also use Pre-trained for testing)

---

基於MATLAB 語意分割於rabbit dataset
(MATLAB semantic segmentation with rabbit dataset)

---

1.資料請下載(data download):Rabbit_myself_416 or Rabbit_myself_608
2.已訓練模型(Pretain_model):model\Modeldownload
3.演算法(algorithm): SP_DeepLabv3
4.標記檔案(label data):gTruth_Pixel_2.mat
5.src_input:JsonSegInput.m,readFcn.m,readFcn2.m

---

基於MATLAB 實例分割於coco dataset

---

1.資料請下載(data download):https://github.com/cocodataset/cocoapi
Download:"2014 Train images" and "2014 Train/Val annotations" links

2.已訓練模型(Pretain_model): NaN
3.演算法(algorithm): SP_MaskRCNN,SP_SOLOv2
4.標記檔案(label data):Using coco dataset/ gTruth_Instance.mat 5.src_input:Polygon2mask_bbox.m