Export
1. ISAT Json Format
1{
2 "info": {
3 "description": "ISAT", # Always ‘ISAT’ for the software to recoginize.
4 "folder": "xxx/ISAT_with_segment_anything/example/images", # The directory where the images are stored.
5 "name": "000000002592.jpg", # The name (path) of the image file.
6 "width": 640, # The dimensions of the image; depth is 3 for RGB images.
7 "height": 366,
8 "depth": 3,
9 "note": "" # An optional field for any additional notes related to the image.
10 },
11 "objects": [ # Object list.
12 {
13 "category": "table", # The class label of the object.
14 "group": 1, # Instance id.
15 "segmentation": [ # The vertices of a polygon composed of x and y coordinates.
16 [0.0, 168.0],
17 [0.0, 365.0],
18 ...
19 ],
20 "area": 223615.81705484597, # The area covered by the object in pixels.
21 "layer": 1.0, # Layer hierarchy
22 "bbox": [0.0, 0.0, 639.0, 365.0], # The bounding box coordinates in the format [x_min, y_min, x_max, y_max].
23 "iscrowd": false, # A boolean value indicating if the object is part of a crowd.
24 "note": "" # An optional field for any additional notes related to the annotation mask.
25 },
26 {
27 "category": "cup",
28 "group": 2,
29 "segmentation": [
30 [276.0, 36.0],
31 [264.0, 39.0],
32 ...
33 ],
34 "area": 32479.5,
35 "layer": 2.0,
36 "bbox": [223.0, 34.0, 430.0, 244.0],
37 "iscrowd": false,
38 "note": ""
39 },
40 ...
41 ]
42}
2. Export as Common Data Formats.
Click Converter button in Tool in Menu bar to open Converter interface.
Tip
ISAT provides the conversion functionality between the following data formats:
ISAT <==> COCO
ISAT <==> YOLO
ISAT <==> LABELME
ISAT ===> VOC
ISAT ===> VOC(object detection)