237 lines
7.5 KiB
Python
237 lines
7.5 KiB
Python
import cv2
|
||
import numpy as np
|
||
import os
|
||
from rknnlite.api import RKNNLite
|
||
|
||
# ====================== 配置区 ======================
|
||
|
||
MODEL_PATH = "point.rknn"
|
||
OUTPUT_DIR = "./output_rknn"
|
||
os.makedirs(OUTPUT_DIR, exist_ok=True)
|
||
|
||
# 固定参考点（像素坐标）
|
||
FIXED_REF_POINT = (535, 605)
|
||
|
||
# mm/px 缩放因子（根据标定数据填写）
|
||
width_mm = 70.0
|
||
width_px = 42
|
||
SCALE_X = width_mm / float(width_px)
|
||
height_mm = 890.0
|
||
height_px = 507
|
||
SCALE_Y = height_mm / float(height_px)
|
||
print(f"Scale factors: SCALE_X={SCALE_X:.3f} mm/px, SCALE_Y={SCALE_Y:.3f} mm/px")
|
||
|
||
# 输入尺寸
|
||
IMG_SIZE = (640, 640)
|
||
|
||
|
||
def letterbox_resize(image, size, bg_color=114):
|
||
target_w, target_h = size
|
||
h, w = image.shape[:2]
|
||
scale = min(target_w / w, target_h / h)
|
||
new_w, new_h = int(w * scale), int(h * scale)
|
||
resized = cv2.resize(image, (new_w, new_h))
|
||
canvas = np.full((target_h, target_w, 3), bg_color, dtype=np.uint8)
|
||
dx, dy = (target_w - new_w) // 2, (target_h - new_h) // 2
|
||
canvas[dy:dy + new_h, dx:dx + new_w] = resized
|
||
return canvas, scale, dx, dy
|
||
|
||
|
||
def safe_sigmoid(x):
|
||
x = np.clip(x, -50, 50)
|
||
return 1.0 / (1.0 + np.exp(-x))
|
||
|
||
|
||
def softmax(x):
|
||
x = x - np.max(x)
|
||
e = np.exp(x)
|
||
return e / e.sum()
|
||
|
||
|
||
def dfl_to_xywh(loc, grid_x, grid_y, stride):
|
||
"""将 DFL 输出解析为 xywh"""
|
||
xywh_ = np.zeros(4)
|
||
xywh = np.zeros(4)
|
||
|
||
# 每个维度 16 bins 做 softmax
|
||
for i in range(4):
|
||
l = loc[i * 16:(i + 1) * 16]
|
||
l = softmax(l)
|
||
xywh_[i] = sum([j * l[j] for j in range(16)])
|
||
|
||
# 对应公式
|
||
xywh_[0] = (grid_x + 0.5) - xywh_[0]
|
||
xywh_[1] = (grid_y + 0.5) - xywh_[1]
|
||
xywh_[2] = (grid_x + 0.5) + xywh_[2]
|
||
xywh_[3] = (grid_y + 0.5) + xywh_[3]
|
||
|
||
# 转成中心点 + 宽高
|
||
xywh[0] = ((xywh_[0] + xywh_[2]) / 2) * stride
|
||
xywh[1] = ((xywh_[1] + xywh_[3]) / 2) * stride
|
||
xywh[2] = (xywh_[2] - xywh_[0]) * stride
|
||
xywh[3] = (xywh_[3] - xywh_[1]) * stride
|
||
|
||
# 转为左上角坐标
|
||
xywh[0] = xywh[0] - xywh[2] / 2
|
||
xywh[1] = xywh[1] - xywh[3] / 2
|
||
return xywh
|
||
|
||
|
||
def parse_pose_outputs(outputs, conf_threshold=0.5, dx=0, dy=0, scale=1.0):
|
||
"""
|
||
完整解析 RKNN YOLO-Pose 输出
|
||
返回 keypoints, class_id, obj_conf, bbox（已映射回原图）
|
||
"""
|
||
boxes = []
|
||
obj_confs = []
|
||
class_ids = []
|
||
|
||
# 遍历前三个输出 tensor (det 输出)
|
||
for idx in range(3):
|
||
det = np.array(outputs[idx])[0] # (C,H,W)
|
||
C, H, W = det.shape
|
||
num_classes = C - 64 # 前64通道为 DFL bbox
|
||
stride = 640 // H
|
||
|
||
for h in range(H):
|
||
for w in range(W):
|
||
for c in range(num_classes):
|
||
conf = safe_sigmoid(det[64 + c, h, w])
|
||
if conf >= conf_threshold:
|
||
loc = det[:64, h, w].astype(np.float32)
|
||
xywh = dfl_to_xywh(loc, w, h, stride)
|
||
boxes.append(xywh)
|
||
obj_confs.append(conf)
|
||
class_ids.append(c)
|
||
|
||
if not obj_confs:
|
||
best_box = np.array([0, 0, 0, 0])
|
||
class_id = -1
|
||
obj_conf = 0.0
|
||
else:
|
||
max_idx = np.argmax(obj_confs)
|
||
best_box = boxes[max_idx]
|
||
class_id = class_ids[max_idx]
|
||
obj_conf = obj_confs[max_idx]
|
||
|
||
# 🔹 bbox 坐标映射回原图
|
||
x, y, w, h = best_box
|
||
x = (x - dx) / scale
|
||
y = (y - dy) / scale
|
||
w = w / scale
|
||
h = h / scale
|
||
best_box = np.array([x, y, w, h])
|
||
|
||
# 🔹 关键点解析
|
||
kpt_output = np.array(outputs[3])[0] # (num_kpts, 3, num_anchor)
|
||
confs = kpt_output[:, 2, :]
|
||
best_anchor_idx = np.argmax(np.mean(confs, axis=0))
|
||
kpt_data = kpt_output[:, :, best_anchor_idx]
|
||
|
||
keypoints = []
|
||
for i in range(kpt_data.shape[0]):
|
||
x_img, y_img, vis_conf_raw = kpt_data[i]
|
||
vis_prob = safe_sigmoid(vis_conf_raw)
|
||
x_orig = (x_img - dx) / scale
|
||
y_orig = (y_img - dy) / scale
|
||
keypoints.append([x_orig, y_orig, vis_prob])
|
||
|
||
return np.array(keypoints), class_id, obj_conf, best_box
|
||
|
||
|
||
def compute_offset(keypoints, fixed_point, scale_x, scale_y):
|
||
if len(keypoints) < 2:
|
||
return None
|
||
p1, p2 = keypoints[0], keypoints[1]
|
||
cx = (p1[0] + p2[0]) / 2.0
|
||
cy = (p1[1] + p2[1]) / 2.0
|
||
dx_mm = (cx - fixed_point[0]) * scale_x
|
||
dy_mm = (cy - fixed_point[1]) * scale_y
|
||
return cx, cy, dx_mm, dy_mm
|
||
|
||
|
||
def visualize_result(image, keypoints, bbox, fixed_point, offset_info, save_path):
|
||
vis = image.copy()
|
||
colors = [(0, 0, 255), (0, 255, 255)]
|
||
cx, cy, dx_mm, dy_mm = offset_info
|
||
fx, fy = map(int, fixed_point)
|
||
|
||
# 绘制关键点
|
||
for i, (x, y, conf) in enumerate(keypoints[:2]):
|
||
if conf > 0.5:
|
||
cv2.circle(vis, (int(x), int(y)), 6, colors[i], -1)
|
||
if len(keypoints) >= 2:
|
||
cv2.line(vis,
|
||
(int(keypoints[0][0]), int(keypoints[0][1])),
|
||
(int(keypoints[1][0]), int(keypoints[1][1])),
|
||
(0, 255, 0), 2)
|
||
|
||
# 绘制 bbox
|
||
x, y, w, h = bbox
|
||
cv2.rectangle(vis, (int(x), int(y)), (int(x + w), int(y + h)), (255, 0, 0), 2)
|
||
|
||
# 绘制中心点
|
||
cv2.circle(vis, (int(cx), int(cy)), 10, (0, 255, 0), 3)
|
||
cv2.circle(vis, (fx, fy), 12, (255, 255, 0), 3)
|
||
cv2.arrowedLine(vis, (fx, fy), (int(cx), int(cy)), (255, 255, 0), 2, tipLength=0.05)
|
||
cv2.putText(vis, f"DeltaX={dx_mm:+.1f}mm", (fx + 30, fy - 30),
|
||
cv2.FONT_HERSHEY_SIMPLEX, 0.8, (0, 255, 255), 2)
|
||
cv2.putText(vis, f"DeltaY={dy_mm:+.1f}mm", (fx + 30, fy + 30),
|
||
cv2.FONT_HERSHEY_SIMPLEX, 0.8, (0, 255, 255), 2)
|
||
|
||
cv2.imwrite(save_path, vis)
|
||
|
||
|
||
def calculate_offset_from_image(image_path, visualize=False):
|
||
orig = cv2.imread(image_path)
|
||
if orig is None:
|
||
return {'success': False, 'message': f'Failed to load image: {image_path}'}
|
||
|
||
img_resized, scale, dx, dy = letterbox_resize(orig, IMG_SIZE)
|
||
infer_img = np.expand_dims(img_resized[..., ::-1], 0).astype(np.uint8)
|
||
|
||
rknn = RKNNLite(verbose=False)
|
||
ret = rknn.load_rknn(MODEL_PATH)
|
||
if ret != 0:
|
||
return {'success': False, 'message': 'Failed to load RKNN model'}
|
||
|
||
try:
|
||
rknn.init_runtime(core_mask=RKNNLite.NPU_CORE_0)
|
||
outputs = rknn.inference([infer_img])
|
||
finally:
|
||
rknn.release()
|
||
|
||
try:
|
||
keypoints, class_id, obj_conf, bbox = parse_pose_outputs(outputs, dx=dx, dy=dy, scale=scale)
|
||
except Exception as e:
|
||
return {'success': False, 'message': f'Parse error: {str(e)}'}
|
||
|
||
offset_info = compute_offset(keypoints, FIXED_REF_POINT, SCALE_X, SCALE_Y)
|
||
if offset_info is None:
|
||
return {'success': False, 'message': 'Not enough keypoints'}
|
||
|
||
cx, cy, dx_mm, dy_mm = offset_info
|
||
|
||
if visualize:
|
||
vis_save_path = os.path.join(OUTPUT_DIR, f"result_{os.path.basename(image_path)}")
|
||
visualize_result(orig, keypoints, bbox, FIXED_REF_POINT, offset_info, vis_save_path)
|
||
|
||
return {'success': True, 'dx_mm': dx_mm, 'dy_mm': dy_mm,
|
||
'cx': cx, 'cy': cy, 'class_id': class_id,
|
||
'obj_conf': obj_conf, 'bbox': bbox,
|
||
'message': 'Success'}
|
||
|
||
|
||
# ====================== 使用示例 ======================
|
||
if __name__ == "__main__":
|
||
image_path = "11.jpg"
|
||
result = calculate_offset_from_image(image_path, visualize=True)
|
||
|
||
if result['success']:
|
||
print(f"Center point: ({result['cx']:.1f}, {result['cy']:.1f})")
|
||
print(f"Offset: DeltaX={result['dx_mm']:+.2f} mm, DeltaY={result['dy_mm']:+.2f} mm")
|
||
print(f"Class ID: {result['class_id']}, Confidence: {result['obj_conf']:.3f}")
|
||
print(f"BBox: {result['bbox']}")
|
||
else:
|
||
print("Error:", result['message'])
|