update

README.md

@@ -18,6 +18,9 @@ tags:
 | PP-DocLayout_plus-L | pp-docLayout_plus-l.onnx | layout detection |
 | PP-OCRv6_medium_det | pp-ocrV6_medium_det.onnx | text line detection |
 | PP-OCRv6_medium_rec | pp-ocrV6_medium_rec.onnx | text recognition |
+| PP-LCNet_x1_0_table_cls | pp-lcNet_x1_0_table_cls.onnx | table classification |
+| RT-DETR-L_wired_table_cell_det | rt-detr-l_wired_table_cell_det.onnx | table cell detection (wired) |
+| RT-DETR-L_wireless_table_cell_det | rt-detr-l_wireless_table_cell_det.onnx | table cell detection (wireless) |
 
 ## Usage
 
@@ -38,6 +41,7 @@ Test images used by the examples:
 
 - [jp_1.jpg](https://github.com/cimo/One_workspace/blob/main/file/test/jp_1.jpg) = full page
 - [det_box_0294.jpg](https://github.com/cimo/One_workspace/blob/main/file/test/det_box_0294.jpg) = single text line
+- [Table_1.jpg](https://github.com/cimo/One_workspace/blob/main/file/test/Table_1.jpg) = single table
 
 ## PP-DocLayout_plus-L
 
@@ -113,3 +117,73 @@ then the blank at index 0.
 - Character map = index 0 is the blank, 1..18708 are the lines of<br>
 [PP-OCRv6_medium_rec/onnx/dictionary.txt](PP-OCRv6_medium_rec/onnx/dictionary.txt), 18709 is the space.
 - Score = mean probability of the kept time steps.
+
+## PP-LCNet_x1_0_table_cls
+
+- [PP-LCNet_x1_0_table_cls/src/example.py](PP-LCNet_x1_0_table_cls/src/example.py)
+
+```
+python3 PP-LCNet_x1_0_table_cls/src/example.py Table_1.jpg
+```
+
+```
+0.843585 | wired
+```
+
+Note:
+- Input = a single cropped table, not a full page: use the `table` boxes from PP-DocLayout_plus-L.
+- Input = RGB image with the short side resized to 256, center cropped to 224x224, scaled to 0..1<br>
+then normalized with the ImageNet mean / std, CHW with batch dimension.
+- Feed = `x`, the only input, fixed at 224x224 with a dynamic batch.
+- Output = `[1, 2]` already softmaxed, index 0 = `wired`, index 1 = `wireless`.
+- The label picks the cell detection model: `wired` = RT-DETR-L_wired_table_cell_det,<br>
+`wireless` = RT-DETR-L_wireless_table_cell_det.
+
+## RT-DETR-L_wired_table_cell_det
+
+- [RT-DETR-L_wired_table_cell_det/src/example.py](RT-DETR-L_wired_table_cell_det/src/example.py)
+
+```
+python3 RT-DETR-L_wired_table_cell_det/src/example.py Table_1.jpg
+```
+
+```
+0.953868 | [467, 189, 1012, 216]
+0.953357 | [467, 162, 1012, 189]
+0.953275 | [467, 216, 1012, 243]
+0.951546 | [467, 243, 1012, 270]
+0.950149 | [467, 135, 1012, 162]
+```
+
+Note:
+- Input = a single cropped table with the borders drawn: use it when PP-LCNet_x1_0_table_cls<br>
+answers `wired`.
+- Input = RGB image resized to 640x640, float32 0..1, CHW with batch dimension.
+- Feed = `image`, `im_shape` (640, 640), `scale_factor` (640 / height, 640 / width).
+- Output row = `[classId, score, x1, y1, x2, y2]` with coordinates already in the cropped table space,<br>
+the box count is in the second output.
+- 1 label class only, `classId` is always 0 (cell), so the class is not printed.
+- Output = always 300 rows: filter by score (0.3 in the example) and remove the duplicates on your side,<br>
+the example drops a box over 0.5 of intersection over union or over 0.9 of containment.
+- The boxes are the single cells: cluster their edges to rebuild the rows / columns and the spans.
+
+## RT-DETR-L_wireless_table_cell_det
+
+- [RT-DETR-L_wireless_table_cell_det/src/example.py](RT-DETR-L_wireless_table_cell_det/src/example.py)
+
+```
+python3 RT-DETR-L_wireless_table_cell_det/src/example.py Table_1.jpg
+```
+
+```
+0.951309 | [464, 162, 1009, 189]
+0.950602 | [464, 189, 1010, 216]
+0.950491 | [464, 216, 1010, 243]
+0.950054 | [464, 243, 1010, 270]
+0.949992 | [464, 297, 1010, 324]
+```
+
+Note:
+- Same input, feed and output of RT-DETR-L_wired_table_cell_det, only the weights change: use it when<br>
+PP-LCNet_x1_0_table_cls answers `wireless`.
+- Input = a single cropped table without the borders drawn, the cells are separated by the spacing only.