PaperlayerView exact revision on arXiv

Source-linked AI summary

Ternary Visible Light Communication Using Event-Based Vision Sensors

Sotaro Kuremoto, Junya Hara, Hiroshi Higashi, Yuichi Tanaka

arXiv:2609.36905v1eess.IVeess.SP

TL;DR

Conventional camera frame rates limit OCC throughput, and existing EVS-based OCC leaves event polarity unused. This paper encodes ternary symbols on an LCD marker read by an EVS and uses frame-transition detection for demodulation. The proposed method reaches 336.5 kbps and outperforms the existing binary method in throughput and BER.

  • Problem

    Conventional camera frame rates limit OCC throughput, while existing EVS-based OCC uses event presence or absence and leaves event polarity unused.

    Source: §1 · p. 1

  • Method

    An LCD encodes increases, decreases, and unchanged brightness as ternary symbols for an EVS, with frame-transition-based demodulation and finder patterns for synchronization and localization.

    Source: §1 · p. 1, §3.5 · p. 3

  • Results

    336.5 kbps at sc = 10 was Ternary-FTD-FP’s highest mean throughput; both Ternary-FTD variants outperformed Binary-OR for every sc.

    Source: §4.3 · p. 4

  • Takeaways & Limitations

    Experiments show the proposed EVS/LCD ternary method achieved higher throughput and lower BER than the existing binary method.

    Source: §5 · p. 4

    The paper assumes a static setting in which the marker and sensor do not move relative to each other.

    Source: §3 · p. 2

Abstract

from arXiv · show

This paper proposes a ternary visible light communication method using an event-based vision sensor (EVS) and a liquid crystal display (LCD). The throughput of optical camera communication (OCC) is limited by the frame rates of conventional frame-based cameras, and EVSs are expected to overcome this limitation because their pixels asynchronously trigger events with polarity in response to brightness changes, with a temporal resolution on the order of microseconds. However, existing event-based OCC relies on binary signaling that uses only the presence or absence of events and leaves the polarity unused. The proposed method maps the three brightness transitions (increase, decrease, and no change) to ternary symbols, thereby increasing the information carried per symbol. At the transmitter, the LCD displays a marker in which the transmitted data are encoded; at the receiver, the EVS captures the marker and records the brightness changes as events with positive and negative polarities. To suppress errors caused by the timing mismatch between display updates at the transmitter and demodulation at the receiver, we also propose a demodulation algorithm based on frame-transition detection. Experimental results demonstrate that the proposed method achieves a throughput of 336.5 kbps, approximately 150% higher than that of its binary counterpart.

1. INTRODUCTION

Conventional camera frame rates limit OCC throughput, while binary EVS-based OCC leaves event polarity unused. This work encodes ternary symbols with an LCD and EVS, adds synchronization and localization measures, and reports higher throughput and lower BER than the existing binary method.

1. INTRODUCTION

  • OCC sends information from a display or LED to an image sensor, and visible light does not pass through obstacles.The paper identifies indoor positioning and underwater communication as possible applications.Source: §1 · p. 1
  • Existing EVS-based OCC encodes bits by event presence or absence but leaves polarity unused, although polarity modulation could increase information per symbol.EVSs detect brightness increases and decreases as positive- and negative-polarity events, respectively.Source: §1 · p. 1
  • The LCD encodes data as increases, decreases, or no change in marker-cell brightness, which the EVS records as events.The three transitions represent ternary symbols.Source: §1 · p. 1
  • Frame-transition-based demodulation uses marker brightness changes as a synchronization signal to prevent symbol-decision errors from LCD characteristics.Four-corner finder patterns support subpixel-accurate reference-point localization to suppress localization errors.Source: §1 · p. 1
  • 336.5 kbps was the proposed method’s highest mean throughput, and it achieved higher throughput and lower BER than the existing binary method.Finder patterns further reduced BER.Source: §1 · p. 1

2. BINARY OCC USING EVS AND LCD

The existing EVS-based OCC system uses asynchronous brightness-change events to detect and decode a dynamic binary marker. Its receiver filters events, locates the marker from event periodicity, and combines partial cell decisions across progressive LCD updates.

2.1. Event-Based Vision Sensor

  • EVS pixels detect brightness changes independently and asynchronously, producing events with positive or negative polarity when changes cross a contrast threshold.The sensor’s temporal resolution is typically on the order of 10^-6 s.Source: §2.1 · p. 1

2.2. Dynamic Marker and Its Decoding

  • The binary dynamic marker contains a data payload, an interior orientation locator, and an exterior locator that distinguishes it from the background.The proposed marker also has finder patterns and separators, while retaining the payload and interior locator.Source: §2.2 · p. 1, Figure 1 · p. 2
  • Alternating data and blank frames encode each payload cell as 255 or 0, with a maximum throughput of fNd/2 bps.The transmitter displays frames at f fps, so data frames occur at f/2 fps.Source: §2.2 · p. 2
  • The receiver filters events, detects the marker using same-polarity event intervals near 2/f, and decodes cells from positive events.Because LCD rows refresh progressively, it combines partial cell decisions from repeated decoding times using logical OR.Source: §2.2 · p. 2
  • The surface of active events returns the timestamp of the latest event of a given polarity at each pixel before evaluation time T1.The receiver evaluates this surface over successive times to compute same-polarity event intervals.Source: §2.2 · p. 2

3. TERNARY OCC USING EVS AND LCD

The method replaces bit-based OCC with trit-based encoding that uses brightness increases, decreases, and no change. Its marker design and frame-transition-based decoding address localization and timing challenges in the EVS–LCD setup.

3. TERNARY OCC USING EVS AND LCD

  • The proposed ternary method retains a similar encode-decode flow to while adding trit-based encoding, a specially designed marker, and a decoder for ternary symbols and robust detection.The paper assumes the marker and sensor remain static relative to each other.Source: §3 · p. 2, Figure 2 · p. 2

3.1. Data Encoding

  • Three-bit groups map one-to-one to two trits, so each pair of payload cells carries three bits, or 1.5 bits per cell.Eight of the nine possible trit pairs are used; (+1, +1) is unused.Source: §3.1 · p. 2

3.2. Marker Structure

  • Each payload cell encodes +1, 0, or −1 through a brightness increase, no change, or decrease; corner finder patterns and the interior locator support orientation and detection.Separators keep payload events from interfering with the finder-pattern scan ratio.Source: §3.2 · pp. 2–3

3.3. Marker Localization by Finder Pattern Detection

  • Scanning event regions for the QR-code 1 : 1 : 3 : 1 : 1 ratio in both directions identifies four finder-pattern centers with subpixel accuracy.Their fixed cell coordinates provide known reference points on the marker.Source: §3.3 · p. 3

3.4. Ternary Symbol Decision

  • The receiver rectifies the marker using the four finder-pattern centers, counts per-cell events by polarity, and applies thresholds to assign tentative trits.A recent event receives a normalized signal value close to 1.Source: §3.4 · p. 3

3.5. Demodulation Based on Frame-Transition Detection

  • Frame-transition-based demodulation addresses decision errors when asynchronous decoding mixes events from successive LCD transitions.LCD updates progress from the marker’s top row to its bottom row.Source: §3.5 · p. 3

3.5.1. Frame-Transition Detection

  • Orange-colored cells that change at every transition provide event counts for detecting each row’s blank-to-data and data-to-blank transition times.The sign of the majority event count indicates transition direction, and threshold crossings set the transition times.Source: §3.5.1 · p. 3

3.5.2. Output Trit Decision

  • The decoder combines tentative trits from both transitions of a data frame, assigning +1 or −1 for the corresponding polarity sequences and 0 otherwise.The timing window Δt1 is set by the EVS response time.Source: §3.5.2 · p. 3

4. EXPERIMENTS

Experiments compare ternary EVS communication with binary EVS communication across cell sizes, using throughput and BER as metrics. The full method combines frame-transition detection with finder-pattern localization.

4.1. Experimental Setup

  • The setup used an IMX636 EVS and a 60 Hz LCD, 70 cm apart, with the marker displayed at 60 fps; brightness levels were 130, 90, and 40.Filters and demodulation used a 1 ms refractory window, a 5-pixel spatial radius, a 10 ms time window, minimum event count 3, and θw = 0.5, θb = 0.1, θc = 0.5, n = 9.Source: §4.1 · p. 4

4.2. Evaluation Protocol

  • Throughput was R = (1 − BER)Rmax, with α = 1 bit per payload cell for Binary-OR and α = 3/2 for ternary methods.Methods included Binary-OR, Ternary-FTD, and Ternary-FTD-FP; Ternary-OR differed from Binary-OR only in encoding.Source: §4.2 · p. 4
  • Five trials per condition and cell size used random bit sequences during the final 5 s of each 10 s trial; tested sc values were 30, 25, 20, 15, and 10.Payload excluded the four finder patterns, separators, and interior locator.Source: §4.2 · p. 4

4.3. Results and Discussion

  • 336.5 kbps was the highest mean throughput, achieved by Ternary-FTD-FP at sc = 10; both ternary methods outperformed Binary-OR at every tested sc.The throughput comparison is shown in Fig. 5(a).Source: Figure 5 · p. 4, §4.3 · p. 4
  • Ternary-FTD had lower BER than both OR-based methods at every sc; Ternary-FTD-FP lowered BER further except at sc = 30.BER generally increased as sc decreased, as each cell covered fewer pixels; frame-transition detection avoids errors from events spanning both transitions, and finder-pattern centers reduce marker-position error.Source: §4.3 · p. 4, Figure 5 · p. 4
  • The proposed method reached 0.41 bps/pixel at 60 fps, comparable to frame-camera OCC systems at 120 fps, though experimental setups differed.Grid-patterned OCC and ChromaCode reported 0.023 and 0.36 bps/pixel, respectively; the proposed method's BER was 3.2 × 10^-2 versus ChromaCode's 5 × 10^-2.Source: §4.3 · p. 4

5. CONCLUSION

The paper proposes ternary visible light communication using an EVS and an LCD, with frame-transition demodulation and finder-pattern localization. Experiments report higher throughput and lower BER than the existing binary method.

5. CONCLUSION

  • Each payload cell carries a ternary symbol through event polarity in the proposed EVS-and-LCD method.The three brightness transitions encode the symbols.Source: §5 · p. 4
  • Frame-transition-based demodulation suppresses errors caused by transmitter–receiver asynchrony.Source: §5 · p. 4
  • QR-code finder patterns enable subpixel-accurate localization of reference points; experiments found higher throughput and lower BER than the existing binary method.Source: §5 · p. 4
Loading 2609.36905v1…