Table 3.
Proposed methods with their results. Values are approximate due to non-standard tests performed.
| Reference | Camera | Platform | Monocular/Stereo | Max Reported Depth (m) | Accuracy/Error | Latency | Power Consumption |
|---|---|---|---|---|---|---|---|
| INDOOR | |||||||
| [42] | [63] | Embedded Blackfin BF537 DSP |
Stereo | 3.5 | Error: 2.00–45 cm | 5.0–50 ms | 5 W |
| [43] | [15] | Standard CPU (Pentium 4 laptop) | Stereo | 0.060 | Error: 3.67–5.79% | 0.30 ms | 30% CPU load |
| [44] | [15] | CPU | Stereo | - | Accuracy: 95.0% | - | - |
| [51] | [19] | Spartan 6 FPGA | Stereo | 1.0 | Error: <1.50% | 50 ms | - |
| [35] | [15] | FPGA | Stereo | 3.5 | Error: 0.1–0.5 m | 13.7 µs | - |
| [49] | [64] | Single-core Intel i7 1.9 GHz CPU | Stereo | - | Error: 3.10–13.4 cm | - | 17W |
| [48] 1 | [15] | CPU | Stereo | 5.0 | Accuracy: 75–85% | - | - |
| [47] 2 | [15] | CPU | Monocular | - | - | - | - |
| [40] | [15] | SpiNNaker | Stereo | 2.8 | Accuracy: 96.7–62.2% | 2 ms | 90 W |
| [53] | [21] | CPU i7 3.40 GHz | Stereo | 5.0 | Accuracy: 96.0% | 30 ms | 1.99 mW |
| [45] | [7] | CPU | Stereo | 4.5 | Error: 0.110 m | - | - |
| [55] | [21] | Intel i7 3.4 GHz CPU | Stereo | 5.0 | Accuracy: 61.1–92.0% | 2 ms | - |
| [41] | [21] | TrueNorth | Stereo | 0.12 | Error: 5.00–11.6% | 9 ms | 0.058 mW/Pixel |
| [52] | [23] | NVIDIA 960M GPU | Stereo | 30 | Error: 0.36–0.44 m | 40 ms | - |
| [28] | [65] | CPU | Monocular | 5.5 | Error: 4.00–23.0% | 10 ms | 200 mW |
| [57] | [60] | GeForce GTX TITAN X GPU | Stereo | 30 | Error: 0.136–0.184 m | - | - |
| [50] | [65] | Intel Core i7-8750H CPU | Stereo | 6.0 | Error: 0.16–0.19 m | 78 ms | - |
| Outdoor | |||||||
| [46] 3 | [15] | Intel® i7 processor at 2.60 GHz |
Stereo | 30 | Error: 3.97–6.47% | 32 μs/event | - |
| [58] | [60] | CPU | Monocular | 30 | Error: 2.19–5.05 m | - | - |
| [56] | [60] | CPU | Monocular | 30 | Error: 1.42–4.46 m | - | - |
1 Method was mostly tested indoors with one test outdoors. Due to averaged results this reference is designated as being tested indoors. 2 Method assumed to be mostly tested indoors as no experimental results included; placed as being tested indoors. 3 Results are designated as being taken outdoors due to having more fast motion and scene noise.