def reconstruct(Iorig, I, Y, out_size, threshold=.9):
net_stride = 2**4
side = ((208. + 40.) / 2.) / net_stride # 7.75, 'alpha' in the paper
Probs = Y[..., 0]
Affines = Y[..., 2:]
rx, ry = Y.shape[:2]
ywh = Y.shape[1::-1]
iwh = np.array(I.shape[1::-1], dtype=float).reshape((2, 1))
xx, yy = np.where(Probs > threshold)
WH = getWH(I.shape)
MN = WH / net_stride
vxx = vyy = 0.5 #alpha
base = lambda vx, vy: np.matrix([[-vx, -vy, 1.], [vx, -vy, 1.],
[vx, vy, 1.], [-vx, vy, 1.]]).T
labels = []
for i in range(len(xx)):
y, x = xx[i], yy[i]
affine = Affines[y, x]
prob = Probs[y, x]
mn = np.array([float(x) + .5, float(y) + .5])
A = np.reshape(affine, (2, 3))
A[0, 0] = max(A[0, 0], 0.)
A[1, 1] = max(A[1, 1], 0.)
pts = np.array(A * base(vxx, vyy)) #*alpha
pts_MN_center_mn = pts * side
pts_MN = pts_MN_center_mn + mn.reshape((2, 1))
pts_prop = pts_MN / MN.reshape((2, 1))
labels.append(DLabel(0, pts_prop, prob))
final_labels = nms(labels, .1)
TLps = []
if len(final_labels):
final_labels.sort(key=lambda x: x.prob(), reverse=True)
for i, label in enumerate(final_labels):
t_ptsh = getRectPts(0, 0, out_size[0], out_size[1])
ptsh = np.concatenate((label.pts * getWH(Iorig.shape).reshape(
(2, 1)), np.ones((1, 4))))
H = find_T_matrix(ptsh, t_ptsh)
Ilp = cv2.warpPerspective(Iorig, H, out_size, borderValue=.0)
TLps.append(Ilp)
return final_labels, TLps
def __pts2xys(self,pts): N = pts.shape[1] pts = pts*getWH(self.Iorig.shape).reshape((2,1)) pts = np.concatenate((pts,np.ones((1,N)))) pts = np.squeeze(np.array(np.matmul(self.R,pts))) pts = pts[:2]/pts[2] return pts
def reset_view(self): self.cx, self.cy = .5,.5 wh = np.array([self.width,self.height],dtype=float) self.zoom_factor = (wh/getWH(self.Iorig.shape)).min() self.mouse_center = np.array([.5,.5]) self.angles = np.array([0.,0.,0.]) self._setPerspective()
def updatePerspectiveMatrix(self): zf = self.zoom_factor w,h = getWH(self.Iorig.shape) self.dx = self.cx*w*zf - self.width/2. self.dy = self.cy*h*zf - self.height/2. R = np.eye(3) R = np.matmul(R,np.matrix([[zf,0,-self.dx],[0,zf,-self.dy],[0,0,1]],dtype=float)) R = np.matmul(R,perspective_transform((w,h),angles=self.angles)) self.R = R self.Rinv = np.linalg.inv(R)
def rectifyToPts(self,pts): if pts.shape[1] != 4: return ptsh = pts*getWH(self.Iorig.shape).reshape((2,1)) ptsh = np.concatenate((ptsh,np.ones((1,4)))) to_pts = self.__pts2xys(pts) wi,hi = (to_pts.min(1)[:2]).tolist() wf,hf = (to_pts.max(1)[:2]).tolist() to_pts = np.matrix([[wi,wf,wf,wi],[hi,hi,hf,hf],[1,1,1,1]]) self.R = find_T_matrix(ptsh,to_pts) self.Rinv = np.linalg.inv(self.R) self._setPerspective(update=False)
def augment_sample(I,pts,dim):
maxsum,maxangle = 120,np.array([80.,80.,45.])
angles = np.random.rand(3)*maxangle
if angles.sum() > maxsum:
angles = (angles/angles.sum())*(maxangle/maxangle.sum())
I = im2single(I)
iwh = getWH(I.shape)
whratio = random.uniform(2.,4.)
wsiz = random.uniform(dim*.2,dim*1.)
hsiz = wsiz/whratio
dx = random.uniform(0.,dim - wsiz)
dy = random.uniform(0.,dim - hsiz)
pph = getRectPts(dx,dy,dx+wsiz,dy+hsiz)
pts = pts*iwh.reshape((2,1))
T = find_T_matrix(pts2ptsh(pts),pph)
H = perspective_transform((dim,dim),angles=angles)
H = np.matmul(H,T)
Iroi,pts = project(I,H,pts,dim)
hsv_mod = np.random.rand(3).astype('float32')
hsv_mod = (hsv_mod - .5)*.3
hsv_mod[0] *= 360
Iroi = hsv_transform(Iroi,hsv_mod)
Iroi = np.clip(Iroi,0.,1.)
pts = np.array(pts)
if random.random() > .5:
Iroi,pts = flip_image_and_pts(Iroi,pts)
tl,br = pts.min(1),pts.max(1)
llp = Label(0,tl,br)
return Iroi,llp,pts
def mouse_callback(self,event,x,y,flags,param): mc = np.array([x,y],dtype=float) mc = np.matmul(self.Rinv,np.append(mc,1.)) mc = np.squeeze(np.array(mc)) self.mouse_center = (mc[:2]/mc[2])/getWH(self.Iorig.shape)
def __pt2xy(self,pt): pt = np.squeeze(np.array(np.matmul(self.R,np.append(pt*getWH(self.Iorig.shape),1.)))) pt = pt[:2]/pt[2] return tuple(pt.astype(int).tolist())