multivox/python/vortexstream.py at main · nerdfever/multivox · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
import numpy as np
from record3d import Record3DStream
import cv2
from threading import Event
import asyncio
import gzip

# this code runs on a PC. It obtains point cloud data from an iPhone running Record3D, connected via a USB cable,
# and streams it to a vortex device running pointvision.py

class VortexStream:
    def __init__(self):
        self.event = Event()
        self.session = None
        self.DEVICE_TYPE__TRUEDEPTH = 0
        self.DEVICE_TYPE__LIDAR = 1
        self.running = False
        self.socket = None

    def on_new_frame(self):
        self.event.set()  # Notify the main thread to stop waiting and process new frame.

    def on_stream_stopped(self):
        self.running = False
        print('Stream stopped')

    def connect_to_device(self, dev_idx):
        print('Searching for devices')
        devs = Record3DStream.get_connected_devices()
        print('{} device(s) found'.format(len(devs)))
        for dev in devs:
            print('\tID: {}\n\tUDID: {}\n'.format(dev.product_id, dev.udid))

        if len(devs) <= dev_idx:
            raise RuntimeError('Cannot connect to device #{}, try different index.'
                               .format(dev_idx))

        dev = devs[dev_idx]
        self.session = Record3DStream()
        self.session.on_new_frame = self.on_new_frame
        self.session.on_stream_stopped = self.on_stream_stopped
        self.session.connect(dev)  # Initiate connection and start capturing

    def make_intrinsic_mat(self, coeffs):
        return np.array([[coeffs.fx,         0, coeffs.tx],
                         [        0, coeffs.fy, coeffs.ty],
                         [        0,         0,         1]])

    def make_inv_intrinsic_mat(self, coeffs):
        return np.array([[1/coeffs.fx,         0, -coeffs.tx/coeffs.fx],
                         [        0, 1/coeffs.fy, -coeffs.ty/coeffs.fy],
                         [        0,         0,         1]])

    async def start(self, host, port):
        reader, writer = await asyncio.open_connection(host, 0x5658)

        '''
        cv2.namedWindow('VoxelsX', cv2.WINDOW_NORMAL)
        cv2.namedWindow('VoxelsY', cv2.WINDOW_NORMAL)
        cv2.namedWindow('VoxelsZ', cv2.WINDOW_NORMAL)
        cv2.resizeWindow('VoxelsX', 512, 256)
        cv2.resizeWindow('VoxelsY', 512, 256)
        cv2.resizeWindow('VoxelsZ', 512, 512)
        '''

        def adjust_gamma(image, gamma):
            invGamma = 1.0 / gamma
            table = np.array([((i / 255.0) ** invGamma) * 255 for i in range(256)]).astype("uint8")
            return cv2.LUT(image, table)


        self.running = True
        try:
            while self.running:
                self.event.wait()
                self.event.clear()
                depth = self.session.get_depth_frame()
                color = self.session.get_rgb_frame()

                downscale = 2
                depth = cv2.resize(depth, (depth.shape[1]//downscale, depth.shape[0]//downscale), interpolation=cv2.INTER_NEAREST)
                color = cv2.resize(color, (depth.shape[1], depth.shape[0]), interpolation=cv2.INTER_NEAREST)

                r = (color[:, :, 0] >> 5) & 0x07
                g = (color[:, :, 1] >> 5) & 0x07
                b = (color[:, :, 2] >> 6) & 0x03
                pixel_t = (r << 5) | (g << 2) | b

                fx = self.session.get_intrinsic_mat().fx
                fy = self.session.get_intrinsic_mat().fy
                tx = self.session.get_intrinsic_mat().tx
                ty = self.session.get_intrinsic_mat().ty
                scale = 256

                rows, cols = depth.shape
                row_indices = np.arange(rows).reshape(-1, 1)
                col_indices = np.arange(cols)
                x_transformed = 64 + ((col_indices * downscale - tx) / fx) * depth * scale
                y_transformed = 32 - ((row_indices * downscale - ty) / fy) * depth * scale

                points = np.stack([x_transformed, (depth - 0.25) * scale, y_transformed, pixel_t], axis=-1).reshape(-1, 4)
                points = points[~np.isnan(points).any(axis=1)]
                points[:, :3] = np.round(points[:, :3])
                points = points[(points[:, 0] >= 0) & (points[:, 0] < 128) & (points[:, 1] >= 0) & (points[:, 1] < 128) & (points[:, 2] >= 0) & (points[:, 2] < 64)]
                points = np.array(points, np.uint8)
                points = np.unique(points[:, :4], axis=0)


                cv2.imshow('Color',  cv2.cvtColor(color, cv2.COLOR_RGB2BGR))
                cv2.imshow('Depth', depth)

                data = gzip.compress(points.tobytes())
                await writer.drain()
                writer.write(b'\xff\xff\xff\xff' + (len(data).to_bytes(4, 'big')))
                writer.write(data)


                pressed = cv2.waitKey(1)
                if pressed == 27 or pressed == ord('q'):
                    break

        finally:

            writer.close()
            await writer.wait_closed()


if __name__ == '__main__':
    app = VortexStream()
    app.connect_to_device(dev_idx=0)
    asyncio.run(app.start('vortex.local', 0x5658))