adder-codec-rs 0.2.7

Encoder/transcoder/decoder for ADΔER (Address, Decimation, Δt Event Representation) streams. Currently, only implemented for raw (uncompressed) streams. Includes a transcoder for casting either framed or event video into an ADΔER representation in a manner which preserves the temporal resolution of the source.
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
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
use opencv::core::{Mat, Size, CV_8U, CV_8UC3};
use std::error::Error;
use std::fmt;

use bumpalo::Bump;
use std::path::Path;
use std::sync::mpsc::{channel, Sender};

use crate::raw::stream::{Error as StreamError, Raw};
use crate::{raw, Codec, Coord, Event, PlaneSize, SourceType, TimeMode, D};
use opencv::highgui;
use opencv::imgproc::resize;
use opencv::prelude::*;

use crate::framer::scale_intensity::FrameValue;
use crate::transcoder::event_pixel_tree::Mode::Continuous;
use crate::transcoder::event_pixel_tree::{DeltaT, Intensity32, Mode, PixelArena};
use crate::SourceCamera;
use davis_edi_rs::util::reconstructor::ReconstructionError;
use ndarray::{Array3, Axis};
use rayon::iter::IntoParallelIterator;
use rayon::iter::ParallelIterator;
use rayon::iter::{IndexedParallelIterator, IntoParallelRefMutIterator};
use rayon::ThreadPool;

#[derive(Debug)]
pub enum SourceError {
    /// Could not open source file
    Open,

    /// ADDER parameters are invalid for the given source
    BadParams,

    StartOutOfBounds,

    /// Source buffer is empty
    BufferEmpty,

    /// Source buffer channel is closed
    BufferChannelClosed,

    /// No data from next spot in buffer
    NoData,

    /// Data not initialized
    UninitializedData,

    /// OpenCV error
    OpencvError(opencv::Error),

    StreamError(raw::stream::Error),

    /// EDI error
    EdiError(ReconstructionError),
}

impl fmt::Display for SourceError {
    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
        write!(f, "Source error")
    }
}

impl From<SourceError> for Box<dyn std::error::Error> {
    fn from(value: SourceError) -> Self {
        value.to_string().into()
    }
}

impl From<opencv::Error> for SourceError {
    fn from(value: opencv::Error) -> Self {
        SourceError::OpencvError(value)
    }
}
impl From<StreamError> for SourceError {
    fn from(value: StreamError) -> Self {
        SourceError::StreamError(value)
    }
}

#[derive(PartialEq, Eq, Clone, Copy, Debug)]
pub enum FramedViewMode {
    Intensity,
    D,
    DeltaT,
}

pub struct VideoState {
    pub plane: PlaneSize,
    pub(crate) pixel_tree_mode: Mode,
    pub chunk_rows: usize,
    pub in_interval_count: u32,
    pub(crate) c_thresh_pos: u8,
    pub(crate) c_thresh_neg: u8,
    pub(crate) delta_t_max: u32,
    pub(crate) ref_time: u32,
    pub(crate) ref_time_divisor: f64,
    pub(crate) tps: DeltaT,
    pub(crate) write_out: bool,
    pub(crate) show_display: bool,
    pub(crate) show_live: bool,
}

impl Default for VideoState {
    fn default() -> Self {
        VideoState {
            plane: PlaneSize::default(),
            pixel_tree_mode: Mode::Continuous,
            chunk_rows: 64,
            in_interval_count: 1,
            c_thresh_pos: 0,
            c_thresh_neg: 0,
            delta_t_max: 7650,
            ref_time: 255,
            ref_time_divisor: 1.0,
            tps: 7650,
            write_out: false,
            show_display: false,
            show_live: false,
        }
    }
}

pub trait VideoBuilder {
    fn contrast_thresholds(self, c_thresh_pos: u8, c_thresh_neg: u8) -> Self;

    fn c_thresh_pos(self, c_thresh_pos: u8) -> Self;

    fn c_thresh_neg(self, c_thresh_neg: u8) -> Self;

    fn chunk_rows(self, chunk_rows: usize) -> Self;

    fn time_parameters(
        self,
        tps: DeltaT,
        ref_time: DeltaT,
        delta_t_max: DeltaT,
    ) -> Result<Self, Box<dyn Error>>
    where
        Self: std::marker::Sized;

    fn write_out(
        self,
        output_filename: String,
        source_camera: SourceCamera,
        time_mode: TimeMode,
    ) -> Result<Box<Self>, Box<dyn std::error::Error>>;

    fn show_display(self, show_display: bool) -> Self;
}

// impl VideoBuilder for Video {}

/// Attributes common to ADΔER transcode process
pub struct Video {
    pub state: VideoState,
    pub(crate) event_pixel_trees: Array3<PixelArena>,
    pub instantaneous_frame: Mat,
    pub instantaneous_view_mode: FramedViewMode,
    pub event_sender: Sender<Vec<Event>>,
    pub(crate) stream: Raw,
}

impl Video {
    /// Initialize the Video with default parameters.
    pub(crate) fn new(plane: PlaneSize, pixel_tree_mode: Mode) -> Result<Video, Box<dyn Error>> {
        let mut state = VideoState {
            pixel_tree_mode,
            ..Default::default()
        };

        let mut data = Vec::new();
        for y in 0..plane.height {
            for x in 0..plane.width {
                for c in 0..plane.channels {
                    let px = PixelArena::new(
                        1.0,
                        Coord {
                            x,
                            y,
                            c: match &plane.channels {
                                1 => None,
                                _ => Some(c),
                            },
                        },
                    );
                    data.push(px);
                }
            }
        }

        let event_pixel_trees: Array3<PixelArena> =
            Array3::from_shape_vec((plane.h_usize(), plane.w_usize(), plane.c_usize()), data)?;
        let mut instantaneous_frame = Mat::default();
        match plane.channels {
            1 => unsafe {
                instantaneous_frame.create_rows_cols(plane.h() as i32, plane.w() as i32, CV_8U)?;
            },
            _ => unsafe {
                instantaneous_frame.create_rows_cols(
                    plane.h() as i32,
                    plane.w() as i32,
                    CV_8UC3,
                )?;
            },
        }

        state.plane = plane;
        let instantaneous_view_mode = FramedViewMode::Intensity;
        let (event_sender, _) = channel();
        let stream = Raw::new();
        Ok(Video {
            state,
            event_pixel_trees,
            instantaneous_frame,
            instantaneous_view_mode,
            event_sender,
            stream,
        })
    }

    pub fn c_thresh_pos(mut self, c_thresh_pos: u8) -> Self {
        self.state.c_thresh_pos = c_thresh_pos;
        self
    }

    pub fn c_thresh_neg(mut self, c_thresh_neg: u8) -> Self {
        self.state.c_thresh_neg = c_thresh_neg;
        self
    }

    pub fn chunk_rows(mut self, chunk_rows: usize) -> Self {
        self.state.chunk_rows = chunk_rows;
        self
    }

    pub fn time_parameters(
        mut self,
        tps: DeltaT,
        ref_time: DeltaT,
        delta_t_max: DeltaT,
    ) -> Result<Self, Box<dyn Error>> {
        if self.stream.has_output_stream() {
            return Err(
                "Cannot change time parameters after output stream has been initialized".into(),
            );
        }
        if ref_time > f32::MAX as u32 {
            eprintln!(
                "Reference time {} is too large. Keeping current value of {}.",
                ref_time, self.state.ref_time
            );
            return Ok(self);
        }
        if tps > f32::MAX as u32 {
            eprintln!(
                "Time per sample {} is too large. Keeping current value of {}.",
                tps, self.state.tps
            );
            return Ok(self);
        }
        if delta_t_max > f32::MAX as u32 {
            eprintln!(
                "Delta t max {} is too large. Keeping current value of {}.",
                delta_t_max, self.state.delta_t_max
            );
            return Ok(self);
        }
        if delta_t_max < ref_time {
            eprintln!(
                "Delta t max {} is smaller than reference time {}. Keeping current value of {}.",
                delta_t_max, ref_time, self.state.delta_t_max
            );
            return Ok(self);
        }
        self.state.delta_t_max = delta_t_max;
        self.state.ref_time = ref_time;
        self.state.tps = tps;
        Ok(self)
    }

    pub fn write_out(
        mut self,
        output_filename: String,
        source_camera: Option<SourceCamera>,
        time_mode: Option<TimeMode>,
    ) -> Result<Self, Box<dyn std::error::Error>> {
        self.state.write_out = true;

        let path = Path::new(&output_filename);
        self.stream.open_writer(path)?;
        self.stream.encode_header(
            self.state.plane.clone(),
            self.state.tps,
            self.state.ref_time,
            self.state.delta_t_max,
            2,
            source_camera,
            time_mode,
        )?;

        self.event_pixel_trees.par_map_inplace(|px| {
            px.time_mode(time_mode);
        });
        Ok(self)
    }

    pub fn show_display(mut self, show_display: bool) -> Self {
        self.state.show_display = show_display;
        self
    }

    /// Close and flush the stream writer.
    /// # Errors
    /// Returns an error if the stream writer cannot be closed cleanly.
    pub fn end_write_stream(&mut self) -> Result<(), Box<dyn Error>> {
        self.stream.close_writer()
    }

    #[allow(clippy::needless_pass_by_value)]
    pub(crate) fn integrate_matrix(
        &mut self,
        matrix: Mat,
        time_spanned: f32,
        view_interval: u32,
    ) -> std::result::Result<Vec<Vec<Event>>, SourceError> {
        let frame_arr: &[u8] = match matrix.data_bytes() {
            Ok(v) => v,
            Err(e) => {
                return Err(SourceError::OpencvError(e));
            }
        };
        if self.state.in_interval_count == 0 {
            self.set_initial_d(frame_arr);
        }

        self.state.in_interval_count += 1;

        if self.state.in_interval_count % view_interval == 0 {
            self.state.show_live = true;
        } else {
            self.state.show_live = false;
        }

        let px_per_chunk: usize = self.state.chunk_rows * self.state.plane.area_wc();

        // Important: if framing the events simultaneously, then the chunk division must be
        // exactly the same as it is for the framer
        let big_buffer: Vec<Vec<Event>> = self
            .event_pixel_trees
            .axis_chunks_iter_mut(Axis(0), self.state.chunk_rows)
            .into_par_iter()
            .enumerate()
            .map(|(chunk_idx, mut chunk)| {
                let mut buffer: Vec<Event> = Vec::with_capacity(px_per_chunk);
                let bump = Bump::new();
                let base_val = bump.alloc(0);
                let px_idx = bump.alloc(0);
                let frame_val = bump.alloc(0);
                let frame_val_intensity32 = bump.alloc(0.0);

                for (chunk_px_idx, px) in chunk.iter_mut().enumerate() {
                    *px_idx = chunk_px_idx + px_per_chunk * chunk_idx;

                    *frame_val_intensity32 = (f64::from(frame_arr[*px_idx])
                        * self.state.ref_time_divisor)
                        as Intensity32;
                    *frame_val = *frame_val_intensity32 as u8;

                    integrate_for_px(
                        px,
                        base_val,
                        frame_val,
                        *frame_val_intensity32, // In this case, frame val is the same as intensity to integrate
                        time_spanned,
                        &mut buffer,
                        &self.state,
                    );
                }
                buffer
            })
            .collect();

        if self.state.write_out {
            self.stream.encode_events_events(&big_buffer)?;
        }

        let db = match self.instantaneous_frame.data_bytes_mut() {
            Ok(v) => v,
            Err(e) => {
                return Err(SourceError::OpencvError(e));
            }
        };

        // TODO: When there's full support for various bit-depth sources, modify this accordingly
        let practical_d_max =
            fast_math::log2_raw(255.0 * (self.state.delta_t_max / self.state.ref_time) as f32);
        db.par_iter_mut().enumerate().for_each(|(idx, val)| {
            let y = idx / self.state.plane.area_wc();
            let x = (idx % self.state.plane.area_wc()) / self.state.plane.c_usize();
            let c = idx % self.state.plane.c_usize();
            *val = match self.event_pixel_trees[[y, x, c]].arena[0].best_event {
                Some(event) => u8::get_frame_value(
                    &event.into(),
                    SourceType::U8,
                    self.state.ref_time as DeltaT,
                    practical_d_max,
                    self.state.delta_t_max,
                    self.instantaneous_view_mode,
                ),
                None => *val,
            };
        });

        if self.state.show_live {
            show_display("instance", &self.instantaneous_frame, 1, self)?;
        }

        Ok(big_buffer)
    }

    fn set_initial_d(&mut self, frame_arr: &[u8]) {
        self.event_pixel_trees.par_map_inplace(|px| {
            let idx = px.coord.y as usize * self.state.plane.area_wc()
                + px.coord.x as usize * self.state.plane.c_usize()
                + px.coord.c.unwrap_or(0) as usize;
            let intensity = frame_arr[idx];
            let d_start = f32::from(intensity).log2().floor() as D;
            px.arena[0].set_d(d_start);
            px.base_val = intensity;
        });
    }

    /// Get `ref_time`
    pub fn get_ref_time(&self) -> u32 {
        self.state.ref_time
    }

    /// Get `delta_t_max`
    pub fn get_delta_t_max(&self) -> u32 {
        self.state.delta_t_max
    }

    /// Get `tps`
    pub fn get_tps(&self) -> u32 {
        self.state.tps
    }

    /// Set a new value for `delta_t_max`
    pub fn update_delta_t_max(&mut self, dtm: u32) {
        // Validate new value
        self.state.delta_t_max = self.state.ref_time.max(dtm);
    }

    /// Set a new value for `c_thresh_pos`
    pub fn update_adder_thresh_pos(&mut self, c: u8) {
        self.state.c_thresh_pos = c;
    }

    /// Set a new value for `c_thresh_neg`
    pub fn update_adder_thresh_neg(&mut self, c: u8) {
        self.state.c_thresh_neg = c;
    }
}

pub fn integrate_for_px(
    px: &mut PixelArena,
    base_val: &mut u8,
    frame_val: &u8,
    intensity: Intensity32,
    time_spanned: f32,
    buffer: &mut Vec<Event>,
    state: &VideoState,
) {
    if px.need_to_pop_top {
        buffer.push(px.pop_top_event(intensity, state.pixel_tree_mode, state.ref_time));
    }

    *base_val = px.base_val;

    if *frame_val < base_val.saturating_sub(state.c_thresh_neg)
        || *frame_val > base_val.saturating_add(state.c_thresh_pos)
    {
        px.pop_best_events(buffer, state.pixel_tree_mode, state.ref_time);
        px.base_val = *frame_val;

        // If continuous mode and the D value needs to be different now
        if let Continuous = state.pixel_tree_mode {
            match px.set_d_for_continuous(intensity) {
                None => {}
                Some(event) => buffer.push(event),
            };
        }
    }

    px.integrate(
        intensity,
        time_spanned.into(),
        state.pixel_tree_mode,
        state.delta_t_max,
        state.ref_time,
    );

    if px.need_to_pop_top {
        buffer.push(px.pop_top_event(intensity, state.pixel_tree_mode, state.ref_time));
    }
}

/// If `video.show_display`, shows the given [`Mat`] in an `OpenCV` window
/// with the given name.
///
/// # Errors
/// Returns an [`OpencvError`] if the window cannot be shown, or the [`Mat`] cannot be scaled as
/// needed.
pub fn show_display(window_name: &str, mat: &Mat, wait: i32, video: &Video) -> opencv::Result<()> {
    if video.state.show_display {
        show_display_force(window_name, mat, wait)?;
    }
    Ok(())
}

/// Shows the given [`Mat`] in an `OpenCV` window with the given name.
/// This function is the same as [`show_display`], except that it does not check
/// [`Video::show_display`].
/// This function is useful for debugging.
/// # Errors
/// Returns an [`OpencvError`] if the window cannot be shown, or the [`Mat`] cannot be scaled as
/// needed.
pub fn show_display_force(window_name: &str, mat: &Mat, wait: i32) -> opencv::Result<()> {
    let mut tmp = Mat::default();

    if mat.rows() == 940 {
        highgui::imshow(window_name, mat)?;
    } else {
        let factor = mat.rows() as f32 / 940.0;
        resize(
            mat,
            &mut tmp,
            Size {
                width: (mat.cols() as f32 / factor) as i32,
                height: 940,
            },
            0.0,
            0.0,
            0,
        )?;
        highgui::imshow(window_name, &tmp)?;
    }

    highgui::wait_key(wait)?;
    Ok(())
}

pub trait Source {
    /// Intake one input interval worth of data from the source stream into the ADΔER model as
    /// intensities.
    fn consume(
        &mut self,
        view_interval: u32,
        thread_pool: &ThreadPool,
    ) -> Result<Vec<Vec<Event>>, SourceError>;

    fn get_video_mut(&mut self) -> &mut Video;

    fn get_video_ref(&self) -> &Video;

    fn get_video(self) -> Video;
}