going somewhere

.gitignore

@@ -1,6 +1,161 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+#poetry.lock
+
+# pdm
+#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
+#pdm.lock
+#   pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
+#   in version control.
+#   https://pdm.fming.dev/#use-with-ide
+.pdm.toml
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
 env
 *.mp4
 .idea
 .DS_Store
 *.pt
 *.png
+*.ipynb

dodgeballtracker/__init__.py

@@ -1,6 +1,8 @@
 from pathlib import Path
 
-from dodgeballtracker.logic import DetectedObjectBox, DodgeballTracker
+from dodgeballtracker.logic import (DetectedObjectBox,
+                                    DodgeballTracker,
+                                    get_court_perspective_transform)
 import numpy as np
 
 
@@ -13,8 +15,10 @@ if __name__ == "__main__":
         DetectedObjectBox.from_viridis(threshold) for threshold in np.arange(0.5, 1, 0.05)
     ]
     input_video = input_dir / "micro.mp4"
-    output_video = output_dir / "output.mp4"
+    output_video = output_dir / "output-lmbn_n_cuhk03_d.mp4"
 
-    dodgeball_tracker = DodgeballTracker(input_video, output_video, detected_box_objects)
+    transformation_matrix = get_court_perspective_transform()
+
+    dodgeball_tracker = DodgeballTracker(input_video, output_video, detected_box_objects, transformation_matrix)
 
     dodgeball_tracker.main()

dodgeballtracker/logic.py

@@ -29,27 +29,7 @@ class DetectedObjectBox:
         return cls(min_threshold, color)
 
 
-class DodgeballTracker:
-
-    def __init__(self,
-                 input_file: Path,
-                 output_file: Path,
-                 detected_box_objects: list[DetectedObjectBox]):
-        self.input_path = input_file
-        self.output_path = output_file
-        self.output_json_path = output_file.with_suffix(".json")
-        self.detected_box_objects = sorted(detected_box_objects, reverse=True)
-        self.transformation_matrix = self.get_court_perspective_transform()
-        self.points = []
-
-    def write_frame(self, writer, frame) -> None:
-        cv2.imshow("Frame", frame)
-        writer.write(frame)
-
-    def extract_box(self, data) -> tuple[int, int, int, int]:
-        return int(data[0]), int(data[1]), int(data[2]), int(data[3])
-
-    def get_court_perspective_transform(self):
+def get_court_perspective_transform():
     """
     """
     empirical_top_left = np.float32([416, 175])
@@ -73,6 +53,28 @@ class DodgeballTracker:
                                    court_bottom_right]) * pixel_per_meter
     return cv2.getPerspectiveTransform(empirical_source, empirical_target)
 
+
+class DodgeballTracker:
+
+    def __init__(self,
+                 input_file: Path,
+                 output_file: Path,
+                 detected_box_objects: list[DetectedObjectBox],
+                 transformation_matrix):
+        self.input_path = input_file
+        self.output_path = output_file
+        self.output_json_path = output_file.with_suffix(".json")
+        self.detected_box_objects = sorted(detected_box_objects, reverse=True)
+        self.transformation_matrix = transformation_matrix
+        self.points = []
+
+    def write_frame(self, writer, frame) -> None:
+        cv2.imshow("Frame", frame)
+        writer.write(frame)
+
+    def extract_box(self, data) -> tuple[int, int, int, int]:
+        return int(data[0]), int(data[1]), int(data[2]), int(data[3])
+
     def print_box(self, data, frame_idx, frame_ts):
         xmin, ymin, xmax, ymax = self.extract_box(data)
         xavg = (xmax + xmin) / 2
@@ -80,16 +82,26 @@ class DodgeballTracker:
         ret = self.transformation_matrix @ v
         x, y = ret[0] / ret[2], ret[1] / ret[2]
         object_id = data[-1]
-        self.points.append({"x": x, "y": y, "frameIndex": frame_idx, "frameTimestamp": frame_ts, "objectId": object_id})
+        self.points.append({"x": x,
+                            "y": y,
+                            "frameIndex": frame_idx,
+                            "frameTimestamp": frame_ts,
+                            "objectId": int(object_id)
+                            })
 
-    def plot_boxe_to_frame(self, data, frame, frame_idx, frame_ts):
+    def print_tracker_data(self, tracker_data, frame_idx, frame_ts):
+        for d in tracker_data:
+            self.print_box(d, frame_idx, frame_ts)
+
+    def plot_boxe_to_frame(self, data, frame):
         confidence = float(data[4])
         for detected_box_object in self.detected_box_objects:
 
             if detected_box_object.min_threshold < confidence:
                 xmin, ymin, xmax, ymax = self.extract_box(data)
                 cv2.rectangle(frame, (xmin, ymin), (xmax, ymax), detected_box_object.color, 2)
-                return data  # [[xmin, ymin, xmax - xmin, ymax - ymin], confidence, int(data[5])]
+                return True
+        return False
 
     def run_model_on_frame(self, model, frame):
         return model(frame)[0]
@@ -108,43 +120,45 @@ class DodgeballTracker:
         cv2.polylines(frame, pts, True, (0, 255, 255), 8)
 
     def persist_points(self):
-        nbr_id = max((p["objectId"] for p in self.points))
-        cm = matplotlib.colormaps['viridis']
-        for p in self.points:
-            value = p["objectId"] / nbr_id
-            color: Color = tuple((int(255 * cm(value)[i]) for i in range(3)))
-            p["color"] = f"rgb({color[0]}, {color[2]}, {color[1]})"
-
         with open(self.output_json_path, "wt") as file_:
             json.dump({"points": self.points}, file_, indent=4)
 
+    def _get_tracker(self):
+
+        return DeepOCSORT(
+            model_weights=Path('lmbn_n_cuhk03_d.pt'),  # which ReID model to use
+            device=torch.device("mps"),
+            fp16=True,
+        )
+
+    def _get_detection_model(self):
+        """
+        load the pre-trained YOLOv8n model
+        """
+        return YOLO("yolov8n.pt")
+
     def main(self):
         # define some constants
 
         video_cap = cv2.VideoCapture(str(self.input_path))
         writer = create_video_writer(video_cap, str(self.output_path))
-        tracker = DeepOCSORT(
-            model_weights=Path('osnet_x0_25_msmt17.pt'),  # which ReID model to use
-            device=torch.device("mps"),
-            fp16=True,
-        )
+        tracker = self._get_tracker()
 
         # load the pre-trained YOLOv8n model
-        model = YOLO("yolov8n.pt")
+        model = self._get_detection_model()
 
         for i, (frame, ts) in enumerate(iter_frames(video_cap)):
             # run the YOLO model on the frame
             res = []
             detections = self.run_model_on_frame(model, frame)
             for data in detections.boxes.data.tolist():
-                selected_data = self.plot_boxe_to_frame(data, frame, i, ts)
-                if selected_data:
-                    res.append(selected_data)
+                is_good_box = self.plot_boxe_to_frame(data, frame)
+                if is_good_box:
+                    res.append(data)
 
             # show the frame to our screen
             data_ = tracker.update(np.array(res), frame)
-            for d in data_:
-                self.print_box(d, i, ts)
+            self.print_tracker_data(data_, i, ts)
 
             self.write_frame(writer, frame)
 
@@ -155,6 +169,3 @@ class DodgeballTracker:
         writer.release()
         cv2.destroyAllWindows()
         self.persist_points()
-
-
-

frontend/index.html

@@ -14,6 +14,15 @@
   <h1>Hello, world!</h1>
 
   <h3 id="first-ts"></h3>
+  <div>
+      <select>
+      <option value="lmbn_n_cuhk03_d">lmbn_n_cuhk03_d</option>
+  </select>
+
+  </div>
+
+
+
   <svg width="900" height="450" id="first" style="background-color: lightgray">
   </svg>
 

frontend/main.js

@@ -1,5 +1,7 @@
 import anime from "https://esm.sh/animejs@3.2.2";
 import m from "https://esm.sh/mithril@2.2.2"
+import colormap from "https://esm.sh/colormap@2.3.2"
+
 
 const second = 1000;
 
@@ -68,23 +70,29 @@ document.querySelector('#seek').onclick = function(){
 
 function generatePoints(points, frames){
     const lst = [];
-    for(const point of points.points){
+    const colors = colormap({
+        colormap: 'viridis',
+        nshades: new Set(points.map(p => p.objectId)).size,
+        format: 'hex',
+        alpha: 1
+    });
+    for(const point of points){
         if(frames.indexOf(point.frameIndex) !== -1){
-            lst.push(m("circle", {cx: point.y, cy: point.x, r: 3, fill: point.color}));
+            lst.push(m("text", {x: point.y, y: point.x, fill: colors[point.objectId]}, point.objectId));
         }
     }
     return m("g", lst);
 }
 
 function getTimeStamp(points, frameId){
-    for(const point of points.points){
+    for(const point of points){
         if(point.frameIndex === frameId){
             return `${point.frameTimestamp / 1000} seconds`;
         }
     }
 }
 
-const url = "/dodgeball-tracker/outputs/output.json";
+const url = "/dodgeball-tracker/outputs/output-lmbn_n_cuhk03_d.json";
 const resp = await fetch(url);
 const points = await resp.json();
 
@@ -97,5 +105,5 @@ function range(min, max){
     return ret
 }
 
-m.render(document.getElementById("first"), generatePoints(points, range(-10, 400)));
-m.render(document.getElementById("first-ts"), m("span", getTimeStamp(points, 80)));
+m.render(document.getElementById("first"), generatePoints(points.points, range(-10, 400)));
+m.render(document.getElementById("first-ts"), m("span", getTimeStamp(points.points, 80)));

outputs/.gitignore

@@ -0,0 +1,2 @@
+*.json
+*.mp4