About Us
High-performance image processing acceleration for ARM
About Zango Tech
Zango Tech is a team with years of hands-on experience in ARM-platform performance engineering — ARM NEON intrinsics tuning, image processing algorithm productization, and Android / embedded delivery pipelines.
We started ACL Pack because of four observations:
OpenCV is too heavy for mobile
A 15–50 MB shared library plus a chain of external dependencies (zlib / libpng / libjpeg / libtiff / protobuf) is a real cost for phone apps, IoT devices, and embedded scenarios.
Many operators aren't fully NEON-optimized on ARM
The OpenCV cv::hal layer is a generic SIMD abstraction, not a hand-written NEON path per operator. Going deeper — dedicated intrinsics for each operator — unlocks order-of-magnitude speedups.
Commercial ARM NEON libraries are near-zero
ppl.cv has zero feature-detection operators, ARM Compute Library is built for internal inference, libyuv is YUV-only, Simd is x86-centric. A commercial NEON library covering the full CV operator suite is a market gap.
Developers worldwide want a high-value commercial choice
Transparent pricing, instant purchase, proper commercial licensing, responsive support — things community open-source projects struggle to deliver consistently.
ACL Pack is our direct answer to all four: single static library, zero dependencies, deep NEON optimization, 163 callable entries, starting at ¥108 via self-service purchase.
About ACL Pack
ACL Pack (Accelerated Computer-vision Library) is a high-performance image processing SDK built for ARM platforms, delivering lightweight, efficient, and plug-and-play computer vision acceleration for mobile and embedded developers.
We focus on deep ARM NEON optimization, shipping as a single static library with zero external dependencies and 5-minute integration. 72.3% of NEON operator samples outperform OpenCV across the full size matrix (4 devices × 7 size tiers, 6,864 measured samples), with M tier peak speedup 78×.
Core Capabilities
163 Callable Entries
Covering filters, color conversion, geometric transforms, arithmetic, analysis, feature detection, transform, math, drawing, and contour analysis
Deep ARM NEON Optimization
Line-by-line NEON vectorization, 72.3% of operator samples faster than OpenCV, peak speedup 78×
Lightweight, Zero Dependencies
Single static library, zero external dependencies, C++17 API, 5-minute integration
Commercial-grade Licensing
Signed license system, symbol obfuscation protection, flexible plans
Use Cases
- ●Camera apps & photo editors — real-time filters, enhancement, effects
- ●AI preprocessing pipelines — normalization, color conversion, resizing before inference
- ●Industrial vision & inspection — edge detection, morphology, feature matching
- ●Embedded devices & IoT — efficient image processing in resource-constrained environments
- ●OpenCV replacement — smaller footprint, faster performance, simpler integration
Contact Us
For any questions or inquiries, feel free to reach out:
Email: zangotech@163.com
Please include the inquiry type in your subject: [Technical Support], [Licensing], [Partnership]