tracktor 0.4.1

Multi-target tracking with random finite sets
Documentation 
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
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155

//! Basic LMB (Labeled Multi-Bernoulli) Tracking Example
//!
//! Demonstrates single-sensor multi-target tracking with track identity
//! preservation using the LMB filter.

use tracktor::filters::lmb::{LabeledBirthModel, LmbFilter, LmbTrack, extract_lmb_estimates};
use tracktor::prelude::*;

// Labeled birth model implementation
struct SimpleBirthModel {
    birth_locations: Vec<(f64, StateVector<f64, 4>, StateCovariance<f64, 4>)>,
}

impl SimpleBirthModel {
    fn new() -> Self {
        let cov = StateCovariance::from_matrix(nalgebra::matrix![
            100.0, 0.0, 0.0, 0.0;
            0.0, 100.0, 0.0, 0.0;
            0.0, 0.0, 25.0, 0.0;
            0.0, 0.0, 0.0, 25.0
        ]);

        Self {
            birth_locations: vec![
                (0.03, StateVector::from_array([10.0, 10.0, 0.0, 0.0]), cov),
                (0.03, StateVector::from_array([190.0, 190.0, 0.0, 0.0]), cov),
            ],
        }
    }
}

impl LabeledBirthModel<f64, 4> for SimpleBirthModel {
    fn birth_tracks(&self, label_gen: &mut LabelGenerator) -> Vec<BernoulliTrack<f64, 4>> {
        self.birth_locations
            .iter()
            .map(|(existence, mean, cov)| {
                let label = label_gen.next_label();
                let state = GaussianState::new(1.0, *mean, *cov);
                BernoulliTrack::new(label, *existence, state)
            })
            .collect()
    }

    fn expected_birth_count(&self) -> f64 {
        self.birth_locations.iter().map(|(e, _, _)| e).sum()
    }
}

fn main() {
    println!("LMB Filter: Labeled Multi-Target Tracking");
    println!("==========================================\n");

    // Create models
    let transition = ConstantVelocity2D::new(1.0, 0.99);
    let observation = PositionSensor2D::new(5.0, 0.9);
    let clutter = UniformClutter2D::new(5.0, (0.0, 200.0), (0.0, 200.0));
    let birth = SimpleBirthModel::new();

    // Create LMB filter
    let filter: LmbFilter<f64, _, _, _, _, 4, 2> =
        LmbFilter::new(transition, observation, clutter, birth);

    // Initialize with known targets
    let initial_tracks = vec![
        LmbTrack::new(
            Label::new(0, 0),
            0.95,
            GaussianState::new(
                1.0,
                StateVector::from_array([50.0, 50.0, 2.0, 1.0]),
                StateCovariance::from_matrix(nalgebra::matrix![
                    10.0, 0.0, 0.0, 0.0;
                    0.0, 10.0, 0.0, 0.0;
                    0.0, 0.0, 5.0, 0.0;
                    0.0, 0.0, 0.0, 5.0
                ]),
            ),
        ),
        LmbTrack::new(
            Label::new(0, 1),
            0.95,
            GaussianState::new(
                1.0,
                StateVector::from_array([100.0, 100.0, -1.0, 2.0]),
                StateCovariance::from_matrix(nalgebra::matrix![
                    10.0, 0.0, 0.0, 0.0;
                    0.0, 10.0, 0.0, 0.0;
                    0.0, 0.0, 5.0, 0.0;
                    0.0, 0.0, 0.0, 5.0
                ]),
            ),
        ),
    ];

    let mut state = filter.initial_state_from(initial_tracks);

    println!(
        "Initial: {} tracks, {:.2} expected targets\n",
        state.num_tracks(),
        state.expected_target_count()
    );

    // Simulated measurements over time
    let measurements_per_step = [
        vec![[52.0, 51.0], [99.0, 102.0], [150.0, 30.0]], // Both + clutter
        vec![[54.0, 52.0], [45.0, 180.0]],                // One detected + clutter
        vec![[56.0, 53.0], [97.0, 106.0]],                // Both detected
        vec![[58.0, 54.0], [95.0, 110.0], [20.0, 20.0]],  // Both + clutter
        vec![[60.0, 55.0], [93.0, 114.0]],                // Both detected
    ];

    let dt = 1.0;

    for (t, meas_data) in measurements_per_step.iter().enumerate() {
        let measurements: Vec<Measurement<f64, 2>> = meas_data
            .iter()
            .map(|m| Measurement::from_array(*m))
            .collect();

        println!("Time {}: {} measurements", t, measurements.len());

        // Run filter step
        let (updated, stats) = filter.step(state, &measurements, dt);
        state = updated;

        // Prune low-existence tracks
        state.tracks.prune_by_existence(0.01);

        println!(
            "  Tracks: {}, Expected: {:.2}",
            state.num_tracks(),
            state.expected_target_count()
        );

        if stats.singular_covariance_count > 0 {
            println!(
                "  Warning: {} singular covariances",
                stats.singular_covariance_count
            );
        }

        // Extract and display estimates
        let estimates = extract_lmb_estimates(&state);
        for est in &estimates {
            println!(
                "  Track {:?}: pos=({:.1}, {:.1}), r={:.3}",
                est.label,
                est.state.index(0),
                est.state.index(1),
                est.existence
            );
        }
        println!();
    }
}