Spaces:
Runtime error
Runtime error
Upload simulation_modules.py with huggingface_hub
Browse files- simulation_modules.py +850 -0
simulation_modules.py
ADDED
@@ -0,0 +1,850 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# ===================================================================
|
2 |
+
# ملف: utils.py (نسخة v2.0 - متوافقة مع الخرائط الديناميكية)
|
3 |
+
# ===================================================================
|
4 |
+
import numpy as np
|
5 |
+
from collections import deque
|
6 |
+
import cv2
|
7 |
+
import math
|
8 |
+
from typing import Dict, List
|
9 |
+
import torch
|
10 |
+
from torchvision import transforms
|
11 |
+
import os
|
12 |
+
import json
|
13 |
+
import gzip
|
14 |
+
import torch
|
15 |
+
from torch.utils.data import Dataset
|
16 |
+
from dataclasses import dataclass
|
17 |
+
from typing import Dict, Tuple, Optional
|
18 |
+
import time
|
19 |
+
from dataclasses import dataclass
|
20 |
+
from typing import Dict, Tuple, Optional
|
21 |
+
|
22 |
+
|
23 |
+
def find_peak_box_and_classify(data):
|
24 |
+
"""
|
25 |
+
[v2.0] تجد القمم في شبكة الكشف وتستخرج معلومات الكائنات.
|
26 |
+
هذه النسخة مرنة وتعمل مع أي حجم خريطة.
|
27 |
+
"""
|
28 |
+
# ✅ إصلاح: الحصول على أبعاد الخريطة ديناميكيًا
|
29 |
+
H, W, _ = data.shape
|
30 |
+
|
31 |
+
# ✅ إصلاح: إضافة إطار الأمان يعتمد على حجم الخريطة الفعلي
|
32 |
+
det_data = np.zeros((H + 2, W + 2, 7))
|
33 |
+
det_data[1:H+1, 1:W+1] = data
|
34 |
+
|
35 |
+
detected_objects, object_counts = [], {"car": 0, "bike": 0, "pedestrian": 0, "unknown": 0}
|
36 |
+
|
37 |
+
# ✅ إصلاح: حلقة التكرار تمر على حجم الخريطة الفعلي
|
38 |
+
for i in range(1, H + 1):
|
39 |
+
for j in range(1, W + 1):
|
40 |
+
confidence = det_data[i, j, 0]
|
41 |
+
# تم رفع حد الثقة قليلاً لتقليل الاكتشافات الخاطئة
|
42 |
+
if confidence > 0.2:
|
43 |
+
# البحث عن القمة المحلية
|
44 |
+
if (confidence >= det_data[i,j-1,0] and confidence >= det_data[i,j+1,0] and
|
45 |
+
confidence >= det_data[i-1,j,0] and confidence >= det_data[i+1,j,0]):
|
46 |
+
|
47 |
+
# استخراج البيانات
|
48 |
+
length, width = det_data[i,j,4], det_data[i,j,5]
|
49 |
+
|
50 |
+
# تصنيف مبسط يعتمد على الأبعاد
|
51 |
+
if length > 3.5: obj_type = 'car'
|
52 |
+
elif length > 1.5 and width > 0.5: obj_type = 'bike'
|
53 |
+
else: obj_type = 'pedestrian'
|
54 |
+
|
55 |
+
object_counts[obj_type] += 1
|
56 |
+
|
57 |
+
# ✅ إصلاح: إحداثيات الشبكة الآن نسبة إلى الخريطة الأصلية (i-1, j-1)
|
58 |
+
detected_objects.append({
|
59 |
+
'grid_coords': (i - 1, j - 1),
|
60 |
+
'raw_data': det_data[i, j],
|
61 |
+
'type': obj_type
|
62 |
+
})
|
63 |
+
return detected_objects, object_counts
|
64 |
+
|
65 |
+
def check_for_nearby_obstacle(meta_data, grid_conf, max_dist=15.0, threshold=0.4):
|
66 |
+
"""
|
67 |
+
[v2.0] تتحقق من وجود أي عائق قريب.
|
68 |
+
"""
|
69 |
+
detected_objects, _ = find_peak_box_and_classify(meta_data)
|
70 |
+
if not detected_objects:
|
71 |
+
return False
|
72 |
+
|
73 |
+
# استخدام إعدادات الشبكة من القاموس
|
74 |
+
x_res = grid_conf['x_res']
|
75 |
+
y_res = grid_conf['y_res']
|
76 |
+
x_min = grid_conf['x_min']
|
77 |
+
y_min = grid_conf['y_min']
|
78 |
+
|
79 |
+
for obj_dict in detected_objects:
|
80 |
+
if obj_dict['raw_data'][0] > threshold:
|
81 |
+
raw_data = obj_dict['raw_data']
|
82 |
+
grid_i, grid_j = obj_dict['grid_coords']
|
83 |
+
offset_x, offset_y = raw_data[1], raw_data[2]
|
84 |
+
|
85 |
+
# ✅ إصلاح: استخدام نظام الإحداثيات الصحيح (Y-أمام, X-يسار)
|
86 |
+
# لاحظ أن grid_j يرتبط بـ x_rel و grid_i يرتبط بـ y_rel
|
87 |
+
x_rel = (grid_j * x_res) + x_min + (x_res / 2) + offset_x
|
88 |
+
y_rel = (grid_i * y_res) + y_min + (y_res / 2) + offset_y
|
89 |
+
|
90 |
+
distance = np.linalg.norm([x_rel, y_rel])
|
91 |
+
|
92 |
+
if distance < max_dist:
|
93 |
+
print(f"🛑 تم اكتشاف عائق قريب على مسافة {distance:.2f} متر.")
|
94 |
+
return True, distance
|
95 |
+
|
96 |
+
# print("✅ لا توجد عائق قريب.")
|
97 |
+
|
98 |
+
return False
|
99 |
+
|
100 |
+
|
101 |
+
# ===================================================================
|
102 |
+
# وحدات مساعدة (Helpers) - وظائف نقية ومفصولة
|
103 |
+
# ===================================================================
|
104 |
+
|
105 |
+
def convert_grid_to_relative_xy(grid_i, grid_j, grid_conf):
|
106 |
+
"""
|
107 |
+
يحول إحداثيات الشبكة (i, j) إلى إحداثيات مترية نسبية للسيارة (y-أمام, x-جانب).
|
108 |
+
|
109 |
+
Args:
|
110 |
+
grid_i (int): مؤشر الصف في الشبكة (المحور الأمامي).
|
111 |
+
grid_j (int): مؤشر العمود في الشبكة (المحور الجانبي).
|
112 |
+
grid_conf (dict): قاموس يحتوي على أبعاد ودقة الشبكة.
|
113 |
+
|
114 |
+
Returns:
|
115 |
+
tuple[float, float]: الإحداثيات النسبية (relative_y, relative_x) لمركز الخلية.
|
116 |
+
"""
|
117 |
+
x_res = grid_conf['x_res']
|
118 |
+
y_res = grid_conf['y_res']
|
119 |
+
x_min = grid_conf['x_min']
|
120 |
+
y_min = grid_conf['y_min']
|
121 |
+
|
122 |
+
# حساب مركز الخلية
|
123 |
+
relative_x = (grid_j * x_res) + x_min + (x_res / 2.0)
|
124 |
+
relative_y = (grid_i * y_res) + y_min + (y_res / 2.0)
|
125 |
+
|
126 |
+
return relative_y, relative_x
|
127 |
+
|
128 |
+
# ===================================================================
|
129 |
+
# الفئات الرئيسية (Core Classes)
|
130 |
+
# ===================================================================
|
131 |
+
|
132 |
+
class TrackedObject:
|
133 |
+
"""
|
134 |
+
يمثل كائنًا واحدًا يتم تتبعه عبر الزمن.
|
135 |
+
يحتفظ بمعرف فريد وتاريخ لمواقعه.
|
136 |
+
"""
|
137 |
+
def __init__(self, obj_id, initial_detection, frame_num):
|
138 |
+
self.id = obj_id
|
139 |
+
self.type = initial_detection['type']
|
140 |
+
self.history = deque(maxlen=20) # تخزين آخر 20 موقعًا
|
141 |
+
self.last_frame_seen = frame_num
|
142 |
+
self.last_confidence = 0.0
|
143 |
+
|
144 |
+
self.update(initial_detection, frame_num)
|
145 |
+
|
146 |
+
def update(self, detection, frame_num):
|
147 |
+
"""تحديث حالة الكائن ببيانات اكتشاف جديدة."""
|
148 |
+
self.history.append(detection['global_pos'])
|
149 |
+
self.last_frame_seen = frame_num
|
150 |
+
self.last_confidence = detection['raw_data'][0]
|
151 |
+
|
152 |
+
@property
|
153 |
+
def last_pos(self):
|
154 |
+
"""الحصول على آخر موقع مسجل للكائن."""
|
155 |
+
return self.history[-1] if self.history else None
|
156 |
+
|
157 |
+
def __repr__(self):
|
158 |
+
"""تمثيل نصي مفيد لتصحيح الأخطاء."""
|
159 |
+
return f"Track(ID={self.id}, Type='{self.type}', LastSeen={self.last_frame_seen})"
|
160 |
+
|
161 |
+
|
162 |
+
class Tracker:
|
163 |
+
"""
|
164 |
+
يتتبع الكائنات المكتشفة في خرائط BEV عبر الإطارات.
|
165 |
+
يستخدم نظام إحداثيات عالمي للمطابقة والحفاظ على هوية الكائنات.
|
166 |
+
"""
|
167 |
+
def __init__(self, grid_conf, match_threshold=2.5, prune_age=5):
|
168 |
+
"""
|
169 |
+
Args:
|
170 |
+
grid_conf (dict): إعدادات الشبكة لتحويل الإحداثيات.
|
171 |
+
match_threshold (float): أقصى مسافة (بالمتر) لاعتبار كائن متطابقًا مع مسار.
|
172 |
+
prune_age (int): عدد الإطارات التي يجب انتظارها قبل حذف مسار غير نشط.
|
173 |
+
"""
|
174 |
+
self.grid_conf = grid_conf
|
175 |
+
self.match_threshold = match_threshold
|
176 |
+
self.prune_age = prune_age
|
177 |
+
|
178 |
+
self.tracks = {} # قاموس لتخزين المسارات باستخدام ID كـ مفتاح
|
179 |
+
self.next_track_id = 0
|
180 |
+
|
181 |
+
def process_frame(self, bev_map, ego_pos, ego_theta, frame_num):
|
182 |
+
"""
|
183 |
+
الدالة الرئيسية: تعالج إطارًا واحدًا، تكتشف الكائنات، وتحدث المسارات.
|
184 |
+
"""
|
185 |
+
# 1. اكتشاف الكائنات من خريطة BEV
|
186 |
+
# (نفترض أن find_peak_box_and_classify معرفة في مكان آخر)
|
187 |
+
detections, _ = find_peak_box_and_classify(bev_map)
|
188 |
+
|
189 |
+
# 2. تحويل مواقع الكائنات إلى إحداثيات عالمية
|
190 |
+
self._add_global_positions(detections, ego_pos, ego_theta)
|
191 |
+
|
192 |
+
# 3. مطابقة الاكتشافات بالمسارات الموجودة
|
193 |
+
matches, unmatched_detections = self._match_detections_to_tracks(detections)
|
194 |
+
|
195 |
+
# 4. تحديث المسارات المتطابقة
|
196 |
+
for track_id, detection_idx in matches.items():
|
197 |
+
self.tracks[track_id].update(detections[detection_idx], frame_num)
|
198 |
+
|
199 |
+
# 5. إنشاء مسارات جديدة للاكتشافات غير المتطابقة
|
200 |
+
for detection_idx in unmatched_detections:
|
201 |
+
self._create_new_track(detections[detection_idx], frame_num)
|
202 |
+
|
203 |
+
# 6. حذف المسارات القديمة (التي لم يتم رؤيتها منذ فترة)
|
204 |
+
self._prune_tracks(frame_num)
|
205 |
+
|
206 |
+
return list(self.tracks.values())
|
207 |
+
|
208 |
+
def _add_global_positions(self, detections, ego_pos, ego_theta):
|
209 |
+
"""يضيف مفتاح 'global_pos' لكل اكتشاف في القائمة."""
|
210 |
+
R = np.array([[np.cos(ego_theta), -np.sin(ego_theta)],
|
211 |
+
[np.sin(ego_theta), np.cos(ego_theta)]])
|
212 |
+
|
213 |
+
for det in detections:
|
214 |
+
grid_i, grid_j = det['grid_coords']
|
215 |
+
raw_data = det['raw_data']
|
216 |
+
|
217 |
+
# حساب الموقع النسبي (y-أمام, x-جانب)
|
218 |
+
relative_y, relative_x = convert_grid_to_relative_xy(grid_i, grid_j, self.grid_conf)
|
219 |
+
relative_x += raw_data[1] # الإزاحة الجانبية بالأمتار
|
220 |
+
relative_y += raw_data[2] # الإزاحة الأمامية بالأمتار
|
221 |
+
|
222 |
+
# التحويل إلى إحداثيات عالمية
|
223 |
+
global_offset = R.dot(np.array([relative_y, relative_x]))
|
224 |
+
det['global_pos'] = ego_pos + global_offset
|
225 |
+
|
226 |
+
def _match_detections_to_tracks(self, detections):
|
227 |
+
"""مطابقة بسيطة وجشعة (Greedy Matching) بين الاكتشافات والمسارات."""
|
228 |
+
matches = {}
|
229 |
+
unmatched_detections = set(range(len(detections)))
|
230 |
+
|
231 |
+
if not self.tracks or not detections:
|
232 |
+
return matches, unmatched_detections
|
233 |
+
|
234 |
+
active_track_ids = list(self.tracks.keys())
|
235 |
+
|
236 |
+
for track_id in active_track_ids:
|
237 |
+
track = self.tracks[track_id]
|
238 |
+
min_dist = self.match_threshold
|
239 |
+
best_det_idx = -1
|
240 |
+
|
241 |
+
for i in range(len(detections)):
|
242 |
+
if i not in unmatched_detections: continue
|
243 |
+
|
244 |
+
dist = np.linalg.norm(track.last_pos - detections[i]['global_pos'])
|
245 |
+
if dist < min_dist:
|
246 |
+
min_dist = dist
|
247 |
+
best_det_idx = i
|
248 |
+
|
249 |
+
if best_det_idx != -1:
|
250 |
+
matches[track_id] = best_det_idx
|
251 |
+
unmatched_detections.remove(best_det_idx)
|
252 |
+
|
253 |
+
return matches, unmatched_detections
|
254 |
+
|
255 |
+
def _create_new_track(self, detection, frame_num):
|
256 |
+
"""إنشاء وإضافة مسار جديد إلى القاموس."""
|
257 |
+
new_id = self.next_track_id
|
258 |
+
new_track = TrackedObject(new_id, detection, frame_num)
|
259 |
+
self.tracks[new_id] = new_track
|
260 |
+
self.next_track_id += 1
|
261 |
+
|
262 |
+
def _prune_tracks(self, current_frame_num):
|
263 |
+
"""حذف المسارات التي لم يتم تحديثها منذ فترة طويلة."""
|
264 |
+
ids_to_delete = []
|
265 |
+
for track_id, track in self.tracks.items():
|
266 |
+
if current_frame_num - track.last_frame_seen > self.prune_age:
|
267 |
+
ids_to_delete.append(track_id)
|
268 |
+
|
269 |
+
for track_id in ids_to_delete:
|
270 |
+
del self.tracks[track_id]
|
271 |
+
|
272 |
+
|
273 |
+
class PIDController:
|
274 |
+
|
275 |
+
def __init__(self, K_P=1.0, K_I=0.0, K_D=0.0, n=20):
|
276 |
+
self._K_P, self._K_I, self._K_D = K_P, K_I, K_D
|
277 |
+
self._window = deque([0 for _ in range(n)], maxlen=n)
|
278 |
+
def step(self, error):
|
279 |
+
self._window.append(error)
|
280 |
+
integral = np.mean(self._window) if len(self._window) > 1 else 0.0
|
281 |
+
derivative = (self._window[-1] - self._window[-2]) if len(self._window) > 1 else 0.0
|
282 |
+
return self._K_P * error + self._K_I * integral + self._K_D * derivative
|
283 |
+
def reset(self):
|
284 |
+
self._window.clear()
|
285 |
+
self._window.extend([0 for _ in range(self._window.maxlen)])
|
286 |
+
|
287 |
+
|
288 |
+
|
289 |
+
class InterfuserController:
|
290 |
+
"""
|
291 |
+
[النسخة 34.0] المتحكم ذو الذاكرة (Memory-Enhanced).
|
292 |
+
يعالج مشكلة "التردد المميت" عن طريق:
|
293 |
+
1. عدم الثقة بتقديرات السرعة الأولية للكائنات المكتشفة حديثًا.
|
294 |
+
2. إضافة "فترة سماح" (ذاكرة قصيرة المدى) للحفاظ على الحذر بعد فقدان أثر
|
295 |
+
المركبة التي يتم متابعتها، مما يمنع التسارع غير الآمن.
|
296 |
+
"""
|
297 |
+
def __init__(self, config_dict):
|
298 |
+
self.config = config_dict
|
299 |
+
c = self.config['controller_params']
|
300 |
+
self.freq = self.config.get('frequency', 10.0) # تردد النظام
|
301 |
+
|
302 |
+
# 1. إعداد وحدات التحكم الأساسية
|
303 |
+
self.turn_controller = PIDController(c['turn_KP'], c['turn_KI'], c['turn_KD'], c['turn_n'])
|
304 |
+
self.speed_controller = PIDController(c['speed_KP'], c['speed_KI'], c['speed_KD'], c['speed_n'])
|
305 |
+
|
306 |
+
# 2. دمج المتتبع
|
307 |
+
self.tracker = Tracker(grid_conf=config_dict.get('grid_conf', {}),
|
308 |
+
match_threshold=c.get('tracker_match_thresh', 2.5),
|
309 |
+
prune_age=c.get('tracker_prune_age', 5))
|
310 |
+
|
311 |
+
# 3. إعداد متغيرات الحالة
|
312 |
+
self.current_target_speed = 0.0
|
313 |
+
self.last_steer = 0
|
314 |
+
|
315 |
+
# [تعديل] إضافة متغيرات الذاكرة والسلوك الحذر
|
316 |
+
self.last_followed_track_id = -1
|
317 |
+
self.follow_grace_period_timer = 0
|
318 |
+
|
319 |
+
# متغيرات الحالة الأخرى
|
320 |
+
self.stop_sign_timer, self.red_light_block_timer = 0, 0
|
321 |
+
self.stop_steps_counter, self.forced_move_timer = 0, 0
|
322 |
+
|
323 |
+
def run_step(self, speed, waypoints, junction, traffic_light, stop_sign, bev_map, ego_pos, ego_theta, frame_num):
|
324 |
+
active_tracks = self.tracker.process_frame(bev_map, ego_pos, ego_theta, frame_num)
|
325 |
+
self._update_system_states(speed, traffic_light)
|
326 |
+
steer = self._get_steering_command(waypoints, speed)
|
327 |
+
|
328 |
+
final_goal_speed, reason = self._get_goal_speed(
|
329 |
+
speed, waypoints, traffic_light, stop_sign, junction, active_tracks, ego_pos
|
330 |
+
)
|
331 |
+
|
332 |
+
self._apply_speed_smoothing(final_goal_speed)
|
333 |
+
throttle, brake = self._get_longitudinal_control(speed, self.current_target_speed)
|
334 |
+
|
335 |
+
final_reason = reason if brake else "Cruising"
|
336 |
+
if self.forced_move_timer > 0:
|
337 |
+
throttle, brake, final_reason = self.config['controller_params']['forced_throttle'], False, "Forced Move"
|
338 |
+
|
339 |
+
return steer, throttle, brake, {'target_speed': self.current_target_speed, 'brake_reason': final_reason, 'active_tracks': len(active_tracks)}
|
340 |
+
|
341 |
+
def _get_goal_speed(self, current_speed, waypoints, traffic_light, stop_sign, junction, active_tracks, ego_pos):
|
342 |
+
c = self.config['controller_params']
|
343 |
+
max_speed = c['max_speed']
|
344 |
+
|
345 |
+
# القاعدة 1: قواعد السلامة الثابتة
|
346 |
+
if (traffic_light > c['light_threshold']): return 0.0, "Red Light"
|
347 |
+
if self._is_stop_sign_active(stop_sign, junction): return 0.0, "Stop Sign"
|
348 |
+
|
349 |
+
# القاعدة 2: منطق تجنب العوائق الديناميكي مع الذاكرة
|
350 |
+
obstacle_speed_limit, obstacle_reason, is_following = self._obstacle_avoidance_logic(active_tracks, ego_pos, current_speed)
|
351 |
+
if is_following:
|
352 |
+
return obstacle_speed_limit, obstacle_reason
|
353 |
+
|
354 |
+
# القاعدة 3: منطق الملاحة (نقاط المسار)
|
355 |
+
return self._navigation_logic(waypoints, max_speed)
|
356 |
+
|
357 |
+
def _obstacle_avoidance_logic(self, active_tracks, ego_pos, current_speed):
|
358 |
+
"""[دالة جديدة] تحتوي على كل منطق التعامل مع العوائق."""
|
359 |
+
c = self.config['controller_params']
|
360 |
+
obstacle_speed_limit = c['max_speed']
|
361 |
+
is_following_a_track = False
|
362 |
+
|
363 |
+
if self.follow_grace_period_timer > 0:
|
364 |
+
self.follow_grace_period_timer -= 1
|
365 |
+
|
366 |
+
for track in active_tracks:
|
367 |
+
distance = np.linalg.norm(track.last_pos - ego_pos)
|
368 |
+
|
369 |
+
if distance < c.get('critical_distance', 4.0):
|
370 |
+
self.last_followed_track_id = -1
|
371 |
+
return 0.0, f"Critical Obstacle (ID: {track.id})", True
|
372 |
+
|
373 |
+
if distance < c.get('follow_distance', 12.0):
|
374 |
+
is_following_a_track = True
|
375 |
+
self.last_followed_track_id = track.id
|
376 |
+
self.follow_grace_period_timer = c.get('follow_grace_period', int(2 * self.freq)) # ثانيتان من الذاكرة
|
377 |
+
|
378 |
+
track_speed = self._estimate_track_speed(track)
|
379 |
+
|
380 |
+
# [منطق جديد] تجاهل التقديرات الأولية غير الموثوقة
|
381 |
+
if track_speed < 0.1 and len(track.history) < 3:
|
382 |
+
# لا تثق بالتقدير، حافظ على سرعتك الحالية مؤقتًا لمنع الفرملة غير الضرورية
|
383 |
+
target_speed = current_speed
|
384 |
+
else:
|
385 |
+
target_speed = track_speed * c.get('speed_match_factor', 0.9)
|
386 |
+
|
387 |
+
if target_speed < obstacle_speed_limit:
|
388 |
+
obstacle_speed_limit = target_speed
|
389 |
+
|
390 |
+
if is_following_a_track:
|
391 |
+
return obstacle_speed_limit, f"Following ID {self.last_followed_track_id}", True
|
392 |
+
|
393 |
+
# [منطق جديد] إذا لم نعد نرى السيارة ولكن الذاكرة نشطة
|
394 |
+
if self.follow_grace_period_timer > 0:
|
395 |
+
cautious_speed = c.get('cautious_speed', 5.0) # سرعة منخفضة حذرة
|
396 |
+
return min(current_speed, cautious_speed), f"Cautious (Lost Track {self.last_followed_track_id})", True
|
397 |
+
|
398 |
+
return -1, "No Obstacle", False # إشارة لعدم وجود عائق
|
399 |
+
|
400 |
+
def _navigation_logic(self, waypoints, max_speed):
|
401 |
+
"""[دالة جديدة] تحتوي على منطق الاقتراب من الهدف النهائي."""
|
402 |
+
if not waypoints.any(): return max_speed, "Cruising (No Waypoints)"
|
403 |
+
|
404 |
+
APPROACH_ZONE, STOPPING_ZONE, MIN_APPROACH_SPEED = 15.0, 3.0, 2.5
|
405 |
+
distance_to_target = np.linalg.norm(waypoints[-1].numpy()) # .numpy() للأمان
|
406 |
+
|
407 |
+
if distance_to_target > APPROACH_ZONE:
|
408 |
+
return max_speed, "Cruising"
|
409 |
+
elif distance_to_target > STOPPING_ZONE:
|
410 |
+
ratio = (distance_to_target - STOPPING_ZONE) / (APPROACH_ZONE - STOPPING_ZONE)
|
411 |
+
return MIN_APPROACH_SPEED + ratio * (max_speed - MIN_APPROACH_SPEED), "Approaching Target"
|
412 |
+
else:
|
413 |
+
return 0.0, "Stopping at Target"
|
414 |
+
|
415 |
+
def _apply_speed_smoothing(self, final_goal_speed):
|
416 |
+
"""[دالة جديدة] تحتوي على منطق تنعيم السرعة."""
|
417 |
+
c = self.config['controller_params']
|
418 |
+
accel_rate = c.get('accel_rate', 0.1)
|
419 |
+
decel_rate = c.get('decel_rate', 0.2)
|
420 |
+
|
421 |
+
if final_goal_speed > self.current_target_speed:
|
422 |
+
self.current_target_speed = min(self.current_target_speed + accel_rate, final_goal_speed)
|
423 |
+
elif final_goal_speed < self.current_target_speed:
|
424 |
+
self.current_target_speed = max(self.current_target_speed - decel_rate, final_goal_speed)
|
425 |
+
|
426 |
+
def _estimate_track_speed(self, track):
|
427 |
+
if len(track.history) < 2: return 0.0
|
428 |
+
dist_moved = np.linalg.norm(track.history[-1] - track.history[-2])
|
429 |
+
return dist_moved * self.freq # السرعة = المسافة * التردد
|
430 |
+
|
431 |
+
# --- باقي الدوال المساعدة تبقى كما هي ---
|
432 |
+
def _is_stop_sign_active(self, stop_sign, junction): # ...
|
433 |
+
c = self.config['controller_params']
|
434 |
+
is_active = self.stop_sign_timer > 0
|
435 |
+
if self.stop_sign_timer > 0: self.stop_sign_timer -= 1
|
436 |
+
elif stop_sign > c['stop_threshold']: self.stop_sign_timer = c['stop_sign_duration']
|
437 |
+
if junction < 0.1: self.stop_sign_timer = 0
|
438 |
+
return is_active
|
439 |
+
|
440 |
+
def _get_longitudinal_control(self, current_speed, target_speed): # ...
|
441 |
+
c = self.config['controller_params']
|
442 |
+
speed_error = target_speed - current_speed
|
443 |
+
control_signal = self.speed_controller.step(speed_error)
|
444 |
+
if control_signal > 0:
|
445 |
+
throttle, brake = np.clip(control_signal, 0.0, c['max_throttle']), False
|
446 |
+
else:
|
447 |
+
throttle, brake = 0.0, abs(control_signal) > c['brake_sensitivity']
|
448 |
+
if target_speed < c['min_speed']: throttle, brake = 0.0, True
|
449 |
+
return throttle, brake
|
450 |
+
|
451 |
+
def _update_system_states(self, speed, traffic_light): # ...
|
452 |
+
c = self.config['controller_params']
|
453 |
+
if speed < 0.1: self.stop_steps_counter += 1
|
454 |
+
else: self.stop_steps_counter = 0
|
455 |
+
if self.stop_steps_counter > c['max_stop_time']:
|
456 |
+
self.forced_move_timer, self.stop_steps_counter = c['forced_move_duration'], 0
|
457 |
+
if self.forced_move_timer > 0: self.forced_move_timer -= 1
|
458 |
+
if self.red_light_block_timer > 0: self.red_light_block_timer -= 1
|
459 |
+
elif speed < 0.1 and traffic_light > c['light_threshold']:
|
460 |
+
self.red_light_block_timer = c['max_red_light_time']
|
461 |
+
else:
|
462 |
+
self.red_light_block_timer = 0
|
463 |
+
|
464 |
+
def _get_steering_command(self, waypoints, speed):
|
465 |
+
if not waypoints.any() or speed < 0.2: return 0.0
|
466 |
+
# .numpy() للأمان عند التعامل مع NumPy
|
467 |
+
aim_point = (waypoints[1].numpy() + waypoints[0].numpy()) / 2.0
|
468 |
+
angle_rad = np.arctan2(aim_point[0], aim_point[1])
|
469 |
+
steer = self.turn_controller.step(np.degrees(angle_rad) / -90.0)
|
470 |
+
steer = self.last_steer * 0.4 + steer * 0.6
|
471 |
+
self.last_steer = steer
|
472 |
+
return np.clip(steer, -1.0, 1.0)
|
473 |
+
|
474 |
+
|
475 |
+
def unnormalize_image(tensor: torch.Tensor) -> np.ndarray:
|
476 |
+
"""
|
477 |
+
يعكس عملية تطبيع الصورة في PyTorch ويحولها إلى صورة BGR
|
478 |
+
جاهزة للعرض باستخدام OpenCV.
|
479 |
+
|
480 |
+
Args:
|
481 |
+
tensor: موتر (Tensor) الصورة المُطبع، بالشكل (C, H, W).
|
482 |
+
|
483 |
+
Returns:
|
484 |
+
صورة NumPy array بتنسيق BGR، بالشكل (H, W, C).
|
485 |
+
"""
|
486 |
+
# 1. تحديد قيم المتوسط والانحراف المعياري المستخدمة في التطبيع الأصلي
|
487 |
+
# (هذه هي القيم القياسية لمجموعة بيانات ImageNet)
|
488 |
+
mean = torch.tensor([0.485, 0.456, 0.406], device=tensor.device)
|
489 |
+
std = torch.tensor([0.229, 0.224, 0.225], device=tensor.device)
|
490 |
+
|
491 |
+
# 2. عكس العملية الحسابية: (tensor * std) + mean
|
492 |
+
# يتم تغيير شكل mean و std لتتناسب مع أبعاد الموتر (C, H, W)
|
493 |
+
tensor = tensor * std[:, None, None] + mean[:, None, None]
|
494 |
+
|
495 |
+
# 3. قص القيم للتأكد من أنها ضمن النطاق [0, 1]
|
496 |
+
# قد تتجاوز العملية الحسابية هذا النطاق بشكل طفيف بسبب أخطاء التقريب.
|
497 |
+
tensor = torch.clamp(tensor, 0, 1)
|
498 |
+
|
499 |
+
# 4. تحويل الموتر إلى مصفوفة NumPy
|
500 |
+
img_np = tensor.cpu().numpy()
|
501 |
+
|
502 |
+
# 5. تغيير ترتيب الأبعاد من (C, H, W) إلى (H, W, C)
|
503 |
+
# PyTorch: (القنوات، الارتفاع، العرض)
|
504 |
+
# NumPy/OpenCV: (الارتفاع، العرض، القنوات)
|
505 |
+
img_np = np.transpose(img_np, (1, 2, 0))
|
506 |
+
|
507 |
+
# 6. تحويل نطاق الألوان من [0, 1] إلى [0, 255] وتغيير النوع
|
508 |
+
img_np = (img_np * 255).astype(np.uint8)
|
509 |
+
|
510 |
+
# 7. تحويل قنوات الألوان من RGB إلى BGR
|
511 |
+
# معظم مكتبات التعلم العميق (بما في ذلك PyTorch) تستخدم RGB.
|
512 |
+
# مكتبة OpenCV تتوقع صيغة BGR.
|
513 |
+
img_bgr = cv2.cvtColor(img_np, cv2.COLOR_RGB2BGR)
|
514 |
+
|
515 |
+
return img_bgr
|
516 |
+
# =========================================================
|
517 |
+
# الخطوة 1: دالة world_to_pixel (تبقى كما هي)
|
518 |
+
# =========================================================
|
519 |
+
def world_to_pixel(world_points: np.ndarray, grid_size_pixels: tuple, grid_size_meters: tuple) -> np.ndarray:
|
520 |
+
pixel_per_meter_x = grid_size_pixels[0] / grid_size_meters[0]
|
521 |
+
pixel_per_meter_y = grid_size_pixels[1] / grid_size_meters[1]
|
522 |
+
pixel_x = (grid_size_pixels[0] / 2) + (world_points[:, 1] * pixel_per_meter_x)
|
523 |
+
pixel_y = grid_size_pixels[1] - (world_points[:, 0] * pixel_per_meter_y)
|
524 |
+
return np.vstack((pixel_x, pixel_y)).T
|
525 |
+
|
526 |
+
# ===================================================================
|
527 |
+
# الخطوة 2: دالة render_bev (محدثة للمرحلة الثانية: الخرائط المستقبلية)
|
528 |
+
# ===================================================================
|
529 |
+
def render_bev(
|
530 |
+
active_tracks: List,
|
531 |
+
predicted_waypoints: np.ndarray,
|
532 |
+
ego_pos_global: np.ndarray,
|
533 |
+
ego_theta_global: float,
|
534 |
+
pixels_per_meter: int = 10,
|
535 |
+
grid_size_meters: tuple = (40, 40),
|
536 |
+
future_time_steps: tuple = (1.0, 2.0) # أزمنة التنبؤ المستقبلية
|
537 |
+
) -> Dict[str, np.ndarray]:
|
538 |
+
"""
|
539 |
+
[المرحلة الثانية]
|
540 |
+
تنشئ خريطة للحظة الحالية (t0) وخرائط تنبؤ مستقبلية (t1, t2).
|
541 |
+
"""
|
542 |
+
side_m, fwd_m = grid_size_meters
|
543 |
+
width_px, height_px = int(side_m * pixels_per_meter), int(fwd_m * pixels_per_meter)
|
544 |
+
|
545 |
+
# مصفوفة التحويل من إحداثيات العالم إلى إحداثيات المركبة
|
546 |
+
R_world_to_ego = np.array([[math.cos(ego_theta_global), math.sin(ego_theta_global)],
|
547 |
+
[-math.sin(ego_theta_global), math.cos(ego_theta_global)]])
|
548 |
+
|
549 |
+
# إنشاء قواميس للخرائط الفارغة
|
550 |
+
bev_maps = {
|
551 |
+
't0': np.zeros((height_px, width_px, 3), dtype=np.uint8),
|
552 |
+
't1': np.zeros((height_px, width_px, 3), dtype=np.uint8),
|
553 |
+
't2': np.zeros((height_px, width_px, 3), dtype=np.uint8)
|
554 |
+
}
|
555 |
+
|
556 |
+
# --- 1. رسم الكائنات المتتبعة في الحاضر والمستقبل ---
|
557 |
+
for track in active_tracks:
|
558 |
+
# --- أ. استخراج البيانات الحالية للكائن ---
|
559 |
+
pos_global = track.last_pos
|
560 |
+
yaw_rad_global = getattr(track, 'last_yaw', 0)
|
561 |
+
speed = getattr(track, 'speed', 0)
|
562 |
+
extent = getattr(track, 'last_extent', (1.0, 2.0)) # (الطول، العرض)
|
563 |
+
|
564 |
+
# --- ب. حساب الخصائص النسبية والرسم ---
|
565 |
+
relative_yaw_rad = yaw_rad_global - ego_theta_global
|
566 |
+
angle_deg = -90 - math.degrees(relative_yaw_rad)
|
567 |
+
width_px_obj = extent[1] * 2 * pixels_per_meter
|
568 |
+
length_px_obj = extent[0] * 2 * pixels_per_meter
|
569 |
+
box_size_px = (float(width_px_obj), float(length_px_obj))
|
570 |
+
|
571 |
+
# --- ج. رسم الكائن في الوقت الحالي (t0) ---
|
572 |
+
relative_pos_t0 = R_world_to_ego.dot(pos_global - ego_pos_global)
|
573 |
+
center_pixel_t0 = world_to_pixel(np.array([relative_pos_t0]), (width_px, height_px), grid_size_meters)[0]
|
574 |
+
box_points_t0 = cv2.boxPoints(((float(center_pixel_t0[0]), float(center_pixel_t0[1])), box_size_px, angle_deg))
|
575 |
+
cv2.drawContours(bev_maps['t0'], [box_points_t0.astype(np.int32)], 0, (0, 0, 255), 2) # أحمر
|
576 |
+
|
577 |
+
# --- د. التنبؤ بالمستقبل ورسمه ---
|
578 |
+
for i, t in enumerate(future_time_steps):
|
579 |
+
# حساب الإزاحة بناءً على السرعة والاتجاه والزمن
|
580 |
+
# (هذه العمليات تتم في الإطار المرجعي العالمي)
|
581 |
+
offset = np.array([math.cos(yaw_rad_global), math.sin(yaw_rad_global)]) * speed * t
|
582 |
+
future_pos_global = pos_global + offset
|
583 |
+
|
584 |
+
# تحويل الموقع المستقبلي إلى إحداثيات نسبية ثم إلى بكسلات
|
585 |
+
relative_pos_future = R_world_to_ego.dot(future_pos_global - ego_pos_global)
|
586 |
+
center_pixel_future = world_to_pixel(np.array([relative_pos_future]), (width_px, height_px), grid_size_meters)[0]
|
587 |
+
|
588 |
+
# رسم الصندوق المستقبلي (نفس الحجم والزاوية، ولكن بموقع ولون مختلفين)
|
589 |
+
key = f't{i+1}' # 't1', 't2', etc.
|
590 |
+
box_points_future = cv2.boxPoints(((float(center_pixel_future[0]), float(center_pixel_future[1])), box_size_px, angle_deg))
|
591 |
+
cv2.drawContours(bev_maps[key], [box_points_future.astype(np.int32)], 0, (255, 0, 128), 2) # بنفسجي
|
592 |
+
|
593 |
+
# --- 2. رسم مركبة الأنا (يجب رسمها على كل الخرائط) ---
|
594 |
+
ego_center_pixel = (width_px / 2, height_px - 5)
|
595 |
+
ego_size_px = (1.8 * pixels_per_meter, 4.0 * pixels_per_meter)
|
596 |
+
ego_box = cv2.boxPoints((ego_center_pixel, ego_size_px, -90))
|
597 |
+
for key in bev_maps:
|
598 |
+
cv2.drawContours(bev_maps[key], [ego_box.astype(np.int32)], 0, (0, 255, 255), -1) # أصفر
|
599 |
+
|
600 |
+
# --- 3. رسم نقاط المسار المتوقعة (فقط على خريطة الحاضر t0) ---
|
601 |
+
if predicted_waypoints.size > 0:
|
602 |
+
waypoints_pixels = world_to_pixel(predicted_waypoints, (width_px, height_px), grid_size_meters)
|
603 |
+
cv2.polylines(bev_maps['t0'], [waypoints_pixels.astype(np.int32)], isClosed=False, color=(0, 255, 0), thickness=3)
|
604 |
+
|
605 |
+
return bev_maps
|
606 |
+
|
607 |
+
# =========================================================
|
608 |
+
# (دالة world_to_pixel تبقى كما هي من الرد السابق)
|
609 |
+
def world_to_pixel(world_points, grid_size_pixels, grid_size_meters):
|
610 |
+
pixel_per_meter_x = grid_size_pixels[0] / grid_size_meters[0]
|
611 |
+
pixel_per_meter_y = grid_size_pixels[1] / grid_size_meters[1]
|
612 |
+
|
613 |
+
pixel_x = (world_points[:, 1] * pixel_per_meter_x) + (grid_size_pixels[0] / 2)
|
614 |
+
pixel_y = (grid_size_pixels[1]) - (world_points[:, 0] * pixel_per_meter_y)
|
615 |
+
return np.vstack((pixel_x, pixel_y)).T
|
616 |
+
|
617 |
+
# ==========================================================
|
618 |
+
# الدالة الثانية: generate_bev_image (النسخة النهائية)
|
619 |
+
# ==========================================================
|
620 |
+
def generate_bev_image(sample, grid_size_pixels=(400, 400), grid_size_meters=(20, 20)):
|
621 |
+
"""
|
622 |
+
تأخذ عينة بيانات وتنشئ صورة BEV خام باستخدام OpenCV.
|
623 |
+
:return: صورة (numpy array) بتنسيق BGR.
|
624 |
+
"""
|
625 |
+
side_meters, forward_meters = grid_size_meters
|
626 |
+
width_px, height_px = grid_size_pixels
|
627 |
+
|
628 |
+
bev_image = np.zeros((height_px, width_px, 3), dtype=np.uint8)
|
629 |
+
|
630 |
+
route_data = sample['measurements'].get('route', [])
|
631 |
+
if route_data:
|
632 |
+
route_world = np.array([[point[0], point[1]] for point in route_data])
|
633 |
+
route_pixels = world_to_pixel(route_world, (width_px, height_px), (side_meters, forward_meters))
|
634 |
+
cv2.polylines(bev_image, [route_pixels.astype(np.int_)], isClosed=False, color=(0, 255, 0), thickness=2)
|
635 |
+
|
636 |
+
objects_data = sample['objects']
|
637 |
+
for obj in objects_data:
|
638 |
+
obj_class = obj.get('class', 'غير معروف')
|
639 |
+
if obj_class in ['weather', 'ego_info', 'static']: continue
|
640 |
+
|
641 |
+
pos = obj.get('position', [0,0,0])
|
642 |
+
ext = obj.get('extent', [0,0,0])
|
643 |
+
yaw = obj.get('yaw', 0)
|
644 |
+
|
645 |
+
center_world = np.array([[pos[0], pos[1]]])
|
646 |
+
center_pixel = world_to_pixel(center_world, (width_px, height_px), (side_meters, forward_meters))[0]
|
647 |
+
|
648 |
+
length_m, width_m = ext[0]*2, ext[1]*2
|
649 |
+
|
650 |
+
width_px_obj = width_m * (width_px / side_meters)
|
651 |
+
length_px_obj = length_m * (height_px / forward_meters)
|
652 |
+
|
653 |
+
color = (255, 255, 0) if obj_class == 'ego_car' else (0, 0, 255)
|
654 |
+
|
655 |
+
center_tuple = (float(center_pixel[0]), float(center_pixel[1]))
|
656 |
+
# ملاحظة: OpenCV تتوقع (العرض، الطول) في size_tuple
|
657 |
+
size_tuple = (float(width_px_obj), float(length_px_obj))
|
658 |
+
angle_deg = np.degrees(yaw) # عكس الزاوية لتناسب OpenCV
|
659 |
+
|
660 |
+
box = cv2.boxPoints((center_tuple, size_tuple, angle_deg))
|
661 |
+
box = box.astype(np.int_)
|
662 |
+
cv2.drawContours(bev_image, [box], 0, color, 2)
|
663 |
+
|
664 |
+
return bev_image
|
665 |
+
|
666 |
+
|
667 |
+
# # ==========================================================
|
668 |
+
# # الكلاس الثالث: الواجهة الاحترافية للعرض
|
669 |
+
# # ==========================================================
|
670 |
+
|
671 |
+
|
672 |
+
# ==============================================================================
|
673 |
+
# الكلاس الأول: إعدادات العرض
|
674 |
+
# ==============================================================================
|
675 |
+
@dataclass
|
676 |
+
class DisplayConfig:
|
677 |
+
width: int = 1920
|
678 |
+
height: int = 1080
|
679 |
+
camera_ratio: float = 0.65
|
680 |
+
panel_margin: int = 30
|
681 |
+
section_spacing: int = 25
|
682 |
+
min_section_height: int = 180
|
683 |
+
|
684 |
+
# ==============================================================================
|
685 |
+
# الكلاس الثاني: الواجهة الاحترافية للعرض (النسخة المحسنة)
|
686 |
+
# ==============================================================================
|
687 |
+
class DisplayInterface:
|
688 |
+
def __init__(self, config: Optional[DisplayConfig] = None):
|
689 |
+
self.config = config if config else DisplayConfig()
|
690 |
+
self._init_colors_and_fonts()
|
691 |
+
self._error_log = []
|
692 |
+
|
693 |
+
def _init_colors_and_fonts(self):
|
694 |
+
self.colors = {'panel_bg':(28,30,34), 'text':(240,240,240), 'text_header':(0,165,255), 'separator':(75,75,75), 'throttle':(100,200,100), 'brake':(30,30,240), 'steer_bar':(0,190,255), 'steer_neutral':(100,100,100), 'light_red':(30,30,240), 'light_green':(100,200,100), 'stop_sign':(200,100,255), 'gauge_bg':(50,50,50), 'gauge_needle':(30,30,240), 'error':(30,30,240)}
|
695 |
+
self.fonts = {'header': cv2.FONT_HERSHEY_DUPLEX, 'normal': cv2.FONT_HERSHEY_SIMPLEX}
|
696 |
+
|
697 |
+
def run_interface(self, data: Dict) -> np.ndarray:
|
698 |
+
"""الدالة الرئيسية التي تنشئ وتعيد لوحة التحكم النهائية."""
|
699 |
+
try:
|
700 |
+
dashboard = np.zeros((self.config.height, self.config.width, 3), dtype=np.uint8)
|
701 |
+
cam_w = int(self.config.width * self.config.camera_ratio)
|
702 |
+
|
703 |
+
# --- 1. رسم المكونات الرئيسية ---
|
704 |
+
self._draw_camera_view(dashboard, data.get('camera_view'), cam_w)
|
705 |
+
self._draw_info_overlay(dashboard, data, cam_w)
|
706 |
+
|
707 |
+
# --- 2. رسم لوحة المعلومات الجانبية ---
|
708 |
+
dashboard[:, cam_w:] = self.colors['panel_bg']
|
709 |
+
current_y = self.config.panel_margin
|
710 |
+
current_y = self._draw_bev_maps_section(dashboard, data, cam_w, current_y)
|
711 |
+
self._draw_controls_section(dashboard, data, cam_w, current_y + self.config.section_spacing)
|
712 |
+
|
713 |
+
return dashboard
|
714 |
+
except Exception as e:
|
715 |
+
self._log_error(str(e))
|
716 |
+
return self._create_error_display(str(e))
|
717 |
+
|
718 |
+
# --------------------------------------------------------------------------
|
719 |
+
# الدوال المساعدة للرسم
|
720 |
+
# --------------------------------------------------------------------------
|
721 |
+
|
722 |
+
def _draw_camera_view(self, db: np.ndarray, view: np.ndarray, cam_w: int):
|
723 |
+
if view is not None:
|
724 |
+
db[:, :cam_w] = cv2.resize(view, (cam_w, self.config.height))
|
725 |
+
|
726 |
+
def _draw_info_overlay(self, db: np.ndarray, data: Dict, cam_w: int):
|
727 |
+
"""يرسم شريط المعلومات الشفاف فوق عرض الكاميرا."""
|
728 |
+
overlay = np.zeros_like(db); panel_h=90; alpha=0.6
|
729 |
+
cv2.rectangle(overlay, (0, self.config.height-panel_h), (cam_w, self.config.height), self.colors['panel_bg'], -1)
|
730 |
+
db[:,:cam_w] = cv2.addWeighted(db[:,:cam_w], 1, overlay[:,:cam_w], alpha, 0)
|
731 |
+
|
732 |
+
base_y = self.config.height - panel_h + 35
|
733 |
+
self._draw_text(db, f"Frame: {data.get('frame_num', 'N/A')}", (20, base_y), 1.2)
|
734 |
+
# ============
|
735 |
+
counts=data.get('object_counts',{});
|
736 |
+
self._draw_text(db, f"Detections: C:{counts.get('car',0)}", (20, base_y+35), 1.0, font_type='normal')
|
737 |
+
# self._draw_text(db, f"Detections: C:{counts.get('car',0)} B:{counts.get('bike',0)} P:{counts.get('pedestrian',0)}",
|
738 |
+
# (20, base_y + 35), self.fonts['normal'], 1.0)
|
739 |
+
|
740 |
+
right_x=cam_w-320
|
741 |
+
light_prob = data.get('light_prob', 0.0)
|
742 |
+
light_color = self.colors['light_red'] if light_prob > 0.5 else self.colors['light_green']
|
743 |
+
cv2.circle(db, (right_x, base_y), 12, light_color, -1)
|
744 |
+
self._draw_text(db, f"Light: {light_prob:.2f}", (right_x+25, base_y+8), 1.0, font_type='normal')
|
745 |
+
|
746 |
+
stop_prob = data.get('stop_prob', 0.0)
|
747 |
+
stop_color = self.colors['stop_sign'] if stop_prob > 0.5 else self.colors['steer_neutral']
|
748 |
+
cv2.circle(db, (right_x, base_y+35), 12, stop_color, -1)
|
749 |
+
self._draw_text(db, f"Stop: {stop_prob:.2f}", (right_x+25, base_y+43), 1.0, font_type='normal')
|
750 |
+
|
751 |
+
def _draw_bev_maps_section(self, db: np.ndarray, data: Dict, info_x: int, start_y: int) -> int:
|
752 |
+
"""يرسم قسم خرائط BEV (الحالية والمستقبلية)."""
|
753 |
+
x = info_x + self.config.panel_margin
|
754 |
+
panel_w = (self.config.width - info_x) - 2 * self.config.panel_margin
|
755 |
+
|
756 |
+
main_bev_h = 350; future_bev_h = 150
|
757 |
+
|
758 |
+
# الخريطة الرئيسية
|
759 |
+
bev_t0 = cv2.resize(data.get('map_t0', np.zeros((1,1,3))), (panel_w, main_bev_h))
|
760 |
+
self._draw_text(bev_t0, "BEV t+0.0s", (10, 30), 1.2, color=self.colors['text_header'])
|
761 |
+
db[start_y:start_y+main_bev_h, x:x+panel_w] = bev_t0
|
762 |
+
|
763 |
+
# الخرائط المستقبلية
|
764 |
+
future_y = start_y + main_bev_h + 10
|
765 |
+
future_bev_w = panel_w // 2
|
766 |
+
|
767 |
+
bev_t1 = cv2.resize(data.get('map_t1', np.zeros((1,1,3))), (future_bev_w, future_bev_h))
|
768 |
+
self._draw_text(bev_t1, "t+1.0s", (10, 20), 0.7, font_type='normal')
|
769 |
+
|
770 |
+
bev_t2 = cv2.resize(data.get('map_t2', np.zeros((1,1,3))), (future_bev_w, future_bev_h))
|
771 |
+
self._draw_text(bev_t2, "t+2.0s", (10, 20), 0.7, font_type='normal')
|
772 |
+
|
773 |
+
db[future_y:future_y+future_bev_h, x:x+future_bev_w] = bev_t1
|
774 |
+
db[future_y:future_y+future_bev_h, x+future_bev_w:x+panel_w] = bev_t2
|
775 |
+
|
776 |
+
separator_y = future_y + future_bev_h + self.config.section_spacing
|
777 |
+
cv2.line(db, (info_x, separator_y), (self.config.width, separator_y), self.colors['separator'], 2)
|
778 |
+
return separator_y
|
779 |
+
|
780 |
+
def _draw_controls_section(self, db: np.ndarray, data: Dict, info_x: int, start_y: int):
|
781 |
+
"""يرسم قسم عناصر التحكم بالمركبة."""
|
782 |
+
x = info_x + self.config.panel_margin
|
783 |
+
self._draw_text(db, "VEHICLE CONTROL", (x, start_y+30), 1.2, color=self.colors['text_header'])
|
784 |
+
|
785 |
+
self._draw_gauge_display(db, data, x, start_y + 80)
|
786 |
+
self._draw_control_bars(db, data, x + 180, start_y + 80)
|
787 |
+
|
788 |
+
def _draw_gauge_display(self, db: np.ndarray, data: Dict, x: int, y: int):
|
789 |
+
"""يرسم عداد السرعة."""
|
790 |
+
radius = 65
|
791 |
+
self._draw_speed_gauge(db, x+radius, y+radius, radius, data.get('speed',0), data.get('target_speed',0), 60.0)
|
792 |
+
|
793 |
+
def _draw_control_bars(self, db: np.ndarray, data: Dict, x: int, y: int):
|
794 |
+
"""يرسم مؤشرات التوجيه، الوقود، والمكابح."""
|
795 |
+
bar_w = self.config.width - x - self.config.panel_margin
|
796 |
+
|
797 |
+
self._draw_text(db, f"Steer: {data.get('steer', 0.0):.2f}", (x, y), 0.8, font_type='normal')
|
798 |
+
self._draw_steer_indicator(db, x, y+15, bar_w, 20, data.get('steer', 0.0))
|
799 |
+
|
800 |
+
self._draw_text(db, f"Throttle: {data.get('throttle', 0.0):.2f}", (x, y+50), 0.8, font_type='normal')
|
801 |
+
self._draw_bar(db, x, y+65, bar_w, 20, data.get('throttle', 0.0), 1.0, self.colors['throttle'])
|
802 |
+
|
803 |
+
brake_on = data.get('brake', False)
|
804 |
+
brake_color = self.colors['brake'] if brake_on else self.colors['text']
|
805 |
+
self._draw_text(db, f"Brake: {'ON' if brake_on else 'OFF'}", (x, y+100), 0.8, brake_color, font_type='normal')
|
806 |
+
self._draw_bar(db, x, y+115, bar_w, 20, float(brake_on), 1.0, self.colors['brake'])
|
807 |
+
|
808 |
+
# --------------------------------------------------------------------------
|
809 |
+
# الدوال الأساسية للرسم (Primitives)
|
810 |
+
# --------------------------------------------------------------------------
|
811 |
+
|
812 |
+
def _draw_text(self, img, text, pos, size, color=None, font_type='header', thickness=1):
|
813 |
+
color = color if color is not None else self.colors['text']
|
814 |
+
font = self.fonts[font_type]
|
815 |
+
cv2.putText(img, text, (pos[0]+1, pos[1]+1), font, size, (0,0,0), thickness+1, cv2.LINE_AA)
|
816 |
+
cv2.putText(img, text, pos, font, size, color, thickness, cv2.LINE_AA)
|
817 |
+
|
818 |
+
def _draw_bar(self, img, x, y, w, h, val, max_val, color):
|
819 |
+
val=np.clip(val,0,max_val); ratio=val/max_val
|
820 |
+
cv2.rectangle(img,(x,y),(x+w,y+h),self.colors['steer_neutral'],-1)
|
821 |
+
cv2.rectangle(img,(x,y),(x+int(w*ratio),y+h),color,-1)
|
822 |
+
cv2.rectangle(img,(x,y),(x+w,y+h),self.colors['text'],1)
|
823 |
+
|
824 |
+
def _draw_steer_indicator(self, img, x, y, w, h, val):
|
825 |
+
center_x=x+w//2; val=np.clip(val,-1.0,1.0)
|
826 |
+
cv2.rectangle(img,(x,y),(x+w,y+h),self.colors['steer_neutral'],-1)
|
827 |
+
cv2.line(img,(center_x,y),(center_x,y+h),self.colors['text'],1)
|
828 |
+
indicator_x=center_x+int(val*(w//2*0.95))
|
829 |
+
cv2.line(img,(indicator_x,y),(indicator_x,y+h),self.colors['steer_bar'],6)
|
830 |
+
|
831 |
+
def _draw_speed_gauge(self, img, cx, cy, r, spd, t_spd, m_spd):
|
832 |
+
cv2.circle(img,(cx,cy),r,self.colors['gauge_bg'],-1); cv2.circle(img,(cx,cy),r,self.colors['text'],2)
|
833 |
+
ratio=np.clip(spd/m_spd,0,1); angle=math.radians(135+ratio*270)
|
834 |
+
ex=cx+int(r*0.85*math.cos(angle)); ey=cy+int(r*0.85*math.sin(angle))
|
835 |
+
cv2.line(img,(cx,cy),(ex,ey),self.colors['gauge_needle'],3); cv2.circle(img,(cx,cy),5,self.colors['gauge_needle'],-1)
|
836 |
+
(w,h),_=cv2.getTextSize(f"{spd:.1f}",self.fonts['header'],1.5,3)
|
837 |
+
self._draw_text(img,f"{spd:.1f}",(cx-w//2,cy+10),1.5); self._draw_text(img,"km/h",(cx-25,cy+35),0.5, font_type='normal')
|
838 |
+
self._draw_text(img,f"Target: {t_spd:.1f}",(cx-50,cy-r-10),0.7)
|
839 |
+
|
840 |
+
def _create_error_display(self, error_msg: str):
|
841 |
+
error_display = np.zeros((self.config.height, self.config.width, 3), dtype=np.uint8)
|
842 |
+
error_display[:] = self.colors['panel_bg']
|
843 |
+
self._draw_text(error_display, "SYSTEM ERROR", (self.config.width//2-200, self.config.height//2-50), 1.5, self.colors['error'])
|
844 |
+
self._draw_text(error_display, error_msg, (50, self.config.height//2+20), 0.8, self.colors['steer_bar'], font_type='normal')
|
845 |
+
return error_display
|
846 |
+
|
847 |
+
def _log_error(self, error_msg: str):
|
848 |
+
self._error_log.append(f"{time.strftime('%Y-%m-%d %H:%M:%S')} - {error_msg}")
|
849 |
+
if len(self._error_log) > 50: self._error_log.pop(0)
|
850 |
+
|