panlabel 0.7.0

The universal annotation converter
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
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235

//! Integration tests for Unity Perception / SOLO JSON support.

mod common;

use common::write_bmp;
use panlabel::ir::io_unity_perception_json::{
    from_unity_perception_json_str, read_unity_perception_json, write_unity_perception_json,
    ATTR_INSTANCE_ID, ATTR_SKIPPED_ANNOTATIONS,
};
use panlabel::ir::{Annotation, BBoxXYXY, Category, Dataset, Image, Pixel};
use serde_json::Value;
use std::path::Path;

fn assert_close(actual: f64, expected: f64) {
    assert!(
        (actual - expected).abs() < 1e-9,
        "expected {actual} to be close to {expected}"
    );
}

#[test]
fn reads_fixture_with_bbox_values_and_skipped_non_bbox_annotation() {
    let dataset = read_unity_perception_json(Path::new(
        "tests/fixtures/sample_valid.unity_perception.json",
    ))
    .expect("read Unity fixture");

    assert_eq!(dataset.images.len(), 2);
    assert_eq!(dataset.categories.len(), 2);
    assert_eq!(dataset.annotations.len(), 2);
    assert_eq!(dataset.images[0].file_name, "img1.png");
    assert_eq!(dataset.images[0].width, 100);
    assert_eq!(dataset.images[1].file_name, "img2.png");
    assert_eq!(
        dataset.info.attributes.get(ATTR_SKIPPED_ANNOTATIONS),
        Some(&"1".to_string())
    );

    let first = dataset
        .annotations
        .iter()
        .find(|ann| ann.attributes.get(ATTR_INSTANCE_ID).map(String::as_str) == Some("10"))
        .expect("first bbox");
    assert_close(first.bbox.xmin(), 10.0);
    assert_close(first.bbox.ymin(), 20.0);
    assert_close(first.bbox.xmax(), 50.0);
    assert_close(first.bbox.ymax(), 50.0);

    let origin_dimension = dataset
        .annotations
        .iter()
        .find(|ann| ann.attributes.get(ATTR_INSTANCE_ID).map(String::as_str) == Some("crate-1"))
        .expect("origin/dimension bbox");
    assert_close(origin_dimension.bbox.xmin(), 5.0);
    assert_close(origin_dimension.bbox.ymin(), 4.0);
    assert_close(origin_dimension.bbox.xmax(), 30.0);
    assert_close(origin_dimension.bbox.ymax(), 40.0);
}

#[test]
fn reads_narrow_legacy_captures_json() {
    let json = r#"{
        "version": "0.0.1",
        "captures": [{
            "id": "capture-1",
            "sequence_id": "seq-1",
            "step": 7,
            "timestamp": 12.5,
            "filename": "legacy.png",
            "format": "png",
            "annotations": [{
                "id": "bbox-def",
                "annotation_definition": "bbox-def",
                "values": [{
                    "label_id": 3,
                    "label_name": "box",
                    "instance_id": "inst-1",
                    "x": 2,
                    "y": 3,
                    "width": 8,
                    "height": 9
                }]
            }]
        }]
    }"#;

    let dataset = from_unity_perception_json_str(json).expect("read legacy captures JSON");
    assert_eq!(dataset.images.len(), 1);
    assert_eq!(dataset.images[0].file_name, "legacy.png");
    assert_eq!(dataset.images[0].width, 10);
    assert_eq!(dataset.images[0].height, 12);
    assert_eq!(dataset.annotations.len(), 1);
}

#[test]
fn directory_input_reads_solo_frames_and_annotation_definitions() {
    let temp = tempfile::tempdir().expect("create temp dir");
    std::fs::create_dir_all(temp.path().join("sequence.0")).expect("create sequence dir");
    write_bmp(&temp.path().join("images/img.bmp"), 320, 240);
    std::fs::write(
        temp.path().join("annotation_definitions.json"),
        r#"{
            "annotationDefinitions": [{
                "@type": "type.unity.com/unity.solo.BoundingBox2DAnnotationDefinition",
                "id": "bbox-def",
                "spec": [{"label_id": 5, "label_name": "mapped"}]
            }]
        }"#,
    )
    .expect("write definitions");
    std::fs::write(
        temp.path().join("sequence.0/step0.frame_data.json"),
        r#"{
            "frame": 0,
            "sequence": 0,
            "step": 0,
            "captures": [{
                "id": "camera",
                "filename": "img.bmp",
                "annotations": [{
                    "@type": "type.unity.com/unity.solo.BoundingBox2DAnnotation",
                    "id": "bbox-def",
                    "values": [{"label_id": 5, "instance_id": 1, "x": 1, "y": 2, "width": 30, "height": 40}]
                }]
            }]
        }"#,
    )
    .expect("write frame");

    let dataset = read_unity_perception_json(temp.path()).expect("read directory");
    assert_eq!(dataset.images[0].width, 320);
    assert_eq!(dataset.images[0].height, 240);
    assert_eq!(dataset.categories[0].name, "mapped");
    assert_eq!(dataset.annotations.len(), 1);
}

#[test]
fn directory_output_writes_minimal_solo_layout_and_roundtrips() {
    let temp = tempfile::tempdir().expect("create temp dir");
    let output_dir = temp.path().join("unity-out");
    let dataset = Dataset {
        images: vec![Image::new(1u64, "img.png", 100, 80)],
        categories: vec![Category::new(1u64, "person")],
        annotations: vec![Annotation::new(
            1u64,
            1u64,
            1u64,
            BBoxXYXY::<Pixel>::from_xyxy(10.0, 20.0, 50.0, 70.0),
        )],
        ..Default::default()
    };

    write_unity_perception_json(&output_dir, &dataset).expect("write directory");
    assert!(output_dir.join("annotation_definitions.json").is_file());
    assert!(output_dir
        .join("sequence.0/step0.frame_data.json")
        .is_file());
    assert!(output_dir.join("images/README.txt").is_file());

    let raw = std::fs::read_to_string(output_dir.join("sequence.0/step0.frame_data.json"))
        .expect("read frame");
    let value: Value = serde_json::from_str(&raw).expect("parse frame");
    assert_eq!(
        value["captures"][0]["annotations"][0]["@type"],
        "type.unity.com/unity.solo.BoundingBox2DAnnotation"
    );

    let restored = read_unity_perception_json(&output_dir).expect("read output directory");
    assert_eq!(restored.images.len(), 1);
    assert_eq!(restored.annotations.len(), 1);
    assert_close(restored.annotations[0].bbox.xmin(), 10.0);
    assert_close(restored.annotations[0].bbox.ymax(), 70.0);
}

#[test]
fn empty_bbox_values_are_not_reported_as_skipped_annotations() {
    let temp = tempfile::tempdir().expect("create temp dir");
    let output_dir = temp.path().join("unity-empty");
    let dataset = Dataset {
        images: vec![Image::new(1u64, "empty.png", 100, 80)],
        categories: vec![Category::new(1u64, "person")],
        annotations: vec![],
        ..Default::default()
    };

    write_unity_perception_json(&output_dir, &dataset).expect("write directory");
    let restored = read_unity_perception_json(&output_dir).expect("read output directory");
    assert_eq!(restored.images.len(), 1);
    assert_eq!(restored.annotations.len(), 0);
    assert!(!restored
        .info
        .attributes
        .contains_key(ATTR_SKIPPED_ANNOTATIONS));
}

#[test]
fn writer_removes_stale_panlabel_frame_files() {
    let temp = tempfile::tempdir().expect("create temp dir");
    let output_dir = temp.path().join("unity-stale");
    std::fs::create_dir_all(output_dir.join("sequence.0")).expect("create sequence dir");
    std::fs::write(
        output_dir.join("sequence.0/step99.frame_data.json"),
        r#"{"captures":[{"filename":"stale.png","dimension":[1,1],"annotations":[]}]}"#,
    )
    .expect("write stale frame");

    let dataset = Dataset {
        images: vec![Image::new(1u64, "fresh.png", 100, 80)],
        categories: vec![Category::new(1u64, "person")],
        annotations: vec![],
        ..Default::default()
    };

    write_unity_perception_json(&output_dir, &dataset).expect("write directory");
    assert!(!output_dir
        .join("sequence.0/step99.frame_data.json")
        .exists());
    let restored = read_unity_perception_json(&output_dir).expect("read output directory");
    assert_eq!(restored.images.len(), 1);
    assert_eq!(restored.images[0].file_name, "fresh.png");
}

#[test]
fn rejects_file_output_as_ambiguous() {
    let temp = tempfile::tempdir().expect("create temp dir");
    let dataset = Dataset {
        images: vec![Image::new(1u64, "img.png", 100, 80)],
        categories: vec![Category::new(1u64, "person")],
        annotations: vec![],
        ..Default::default()
    };
    let err = write_unity_perception_json(&temp.path().join("out.json"), &dataset)
        .expect_err("file output should be rejected");
    assert!(err.to_string().contains("directory datasets only"));
}