vhliboptimal_test
Test suite for shape contour detection and image outline recognition
Project Source code
Full Source Code on GitHub: https://github.com/vigatron/vhliboptimal_test
Tested Platforms
| Platform / Board | Processor / MCU | Architecture | Freq |
|---|---|---|---|
| ASUS Vivobook | Intel i5-1135G7 | x86_64 | 2.40 GHz |
| AMD Based Desktop | AMD FX-8300 | x86_64 | 3.30 GHz |
| Orange Pi PC Plus | ARM Cortex-A7 | ARMv7-A | 1.20 GHz |
| Raspberry Pi Model B+ V1.2 | ARM1176JZF-S | ARMv6 | 700 MHz |
| CMB32F407HDMIR3 | STM32F407 | Cortex-M4 | 168 MHz |
| Waveshare Core7XXI | STM32F746 | Cortex-M7 | 216 MHz |
| CMB32H750HDMIR1 | STM32H750 | Cortex-M7 | 480 MHz |
| ESP32-WROOM-32D | ESP32-D0WD | Xtensa LX6 | 240 MHz |
Rev 0.0.8 Notes:
VHTimerStamp integrated as submodule
Rev 0.0.7 Notes:
vhliboptimal integrated as submodule
Time-critical routines placed in ITCMRAM / IRAM
- STM32F407:
- BitFields data arrays optimized for CCMRAM
- Time-critical routines running in FLASH section faster than in SRAM
- ITCMRAM is not available on STM32F407
- STM32F746:
- Optimized for DTCMRAM (BitFields data arrays)
- Selected time-critical routines placed in ITCMRAM
- STM32H750:
- Optimized for DTCMRAM (BitFields data arrays)
- Selected time-critical routines placed in ITCMRAM
- ESP32-D0WD:
- Optimized for ESP32 DRAM (BitFields data arrays)
- Selected time-critical routines placed in IRAM
Test Images Examples
Example 1: Image size: 4096*4096 Letters and Geometric shape (Extreme Size Test)
|
|
Example 2: Image size: 4096*4096 Letters and Geometric shape (Extreme Size Test)
|
|
Example 3: Image size: 1920*1080 (1080p) Shapes
|
|
Example 4: Image size: 1920*1080 (1080p) Synthetic Dense Text Block (Extreme Stress Test)
|
|
Example 5: Image size: 640*480 (VGA) Shapes
|
|
Example 6: Image size: 800*600 (SVGA) Text
|
|
Example 7: Image size: 1280 × 720 (HD 720p) Shapes & Text
|
|
vhliboptimal Benchmark Results
Library rev 0.8.1
Optimized for STM32 F4 / F7 / H7
-
CMB32F407HDMIR3 (STM32F407) GRID 512x512 GRID 256x256 GRID 128x128 -
WaveShare Core7XXI (STM32F746) GRID 512x512 GRID 256x256 GRID 128x128 -
CMB32H750HDMIR1 (STM32H750) GRID 512x512 GRID 256x256 GRID 128x128
Library rev 0.8.0
Added zero-allocation support with FIXED_GRID option for STM32 & ESP32
- Benchmark results on Intel i5-1135G7
- Benchmark results on AMD FX-8300
- Benchmark results on Orange Pi PC Plus
- Benchmark results on Raspberry Pi Model B+ V1.2
- Benchmark results on CMB32F407HDMIR3
- Benchmark results on WaveShare Core7XXI
- Benchmark results on CMB32H750HDMIR1
- Benchmark results on ESP32-WROOM-32D initial revision
- Benchmark results on ESP32-WROOM-32D optimized revision
Library rev 0.7.5
Initial revision for PC and SBC
- Benchmark on Intel i5-1135G7 @ 2.40 GHz
- Benchmark on AMD FX-8300 @ 3.30 GHz
- Benchmark on Orange Pi PC Plus ARM Cortex-A7 @ 1.2 GHz
Dependencies
-
vhlibplatform 0.4.2
https://github.com/vigatron/vhlibplatform -
vhliboptimal 0.8.1
https://github.com/vigatron/vhliboptimal -
vhlibrle7b 0.1.0
https://github.com/vigatron/vhlibrle7b -
vhlibtimerstamp 0.0.2
https://github.com/vigatron/vhlibtimerstamp
Build & Run
GRID Memory Layout Configuration
| Grid Size | Scale Level | Dimensions | Total Memory | Breakdown (Objects + Spans) |
|---|---|---|---|---|
| GRID_512x512 | SCALE_LV = 9 | 512 × 512 | 118,336 bytes | 12,800 bytes (800 objects) + 40,000 bytes (10,000 spans) |
| GRID_256x256 | SCALE_LV = 8 | 256 × 256 | 69,184 bytes | 12,800 bytes (800 objects) + 40,000 bytes (10,000 spans) |
| GRID_128x128 | SCALE_LV = 7 | 128 × 128 | 56,896 bytes | 12,800 bytes (800 objects) + 40,000 bytes (10,000 spans) |
Common camera image resolutions
| Name | Resolution | Aspect ratio |
|---|---|---|
| VGA | 640 × 480 | 4:3 |
| SVGA | 800 × 600 | 4:3 |
| XGA | 1024 × 768 | 4:3 |
| HD 720p | 1280 × 720 | 16:9 |
| SXGA | 1280 × 1024 | 5:4 |
| UXGA | 1600 × 1200 | 4:3 |
| FHD 1080p | 1920 × 1080 | 16:9 |
| QSXGA | 2592 × 1944 | 4:3 |
| 4K UHD | 3840 × 2160 | 16:9 |
Link to resized monochrome .bmp images:
© 2026 V01G04A81 / Viktor Glebov