This guide compares the Sony IMX291, IMX335 and IMX415 STARVIS sensors for industrial USB cameras, explaining their low-light performance, HDR behavior and real-world suitability for robotics, smart retail, security and embedded AI vision systems.

IMX291, IMX335 and IMX415 offer three performance paths—starlight sensitivity, balanced resolution and high-detail 4K. This comparison helps engineers choose the right sensor for industrial USB cameras and embedded AI systems..

 

Quick Takeaway:

• Best for Absolute Darkness: Choose IMX291 (2.9µm pixel size) for starlight conditions < 0.01 Lux.

• Best Balanced 5MP: Choose IMX335 for standard industrial inspection requiring 2K resolution.

• Best for 4K Detail: Choose IMX415 for city surveillance and OCR where resolution beats sensitivity.

• New Upgrade Path: For 2025 projects, consider upgrading IMX291 to IMX462 (NIR Enhanced) and IMX415 to IMX678 (STARVIS 2).

Low-light imaging is one of the most demanding fields in embedded vision. Selecting the right night vision USB camera or starlight camera module is no longer just about resolution—it is about balancing sensor size, signal-to-noise ratio (SNR), dynamic range, and system integration requirements. In this article, we will examine three of Sony’s most widely used STARVIS sensors—IMX291, IMX335, and IMX415—and provide a side-by-side technical comparison to help engineers and system integrators select the optimal module for industrial, surveillance, and robotic applications.

 

Why These Three Sensors Matter in 2025–2030 Industrial Vision

These sensors dominate the market because they serve three different engineering priorities:

• IMX291 → Ultra-low-light champion (large pixels, excellent SNR)

• IMX335 → Balanced clarity & night performance

• IMX415 → High resolution for bright-light applications

Across robotics, AI retail, manufacturing and surveillance, engineers often struggle to balance:

• Low-light requirements

• Resolution needs

• HDR capability

• Computational load on AI models

• Lens FOV and optical compatibility

This guide resolves these trade-offs with clear, engineering-driven comparisons.

Key Parameters in Low-Light Imaging

When evaluating night vision sensors, engineers must consider several critical parameters:

• Pixel Size & Sensitivity
  Larger pixels collect more photons, which improves low-light sensitivity. Smaller pixels favor resolution but at the expense of per-pixel signal strength.
• Signal-to-Noise Ratio (SNR)
  In near-darkness, SNR determines whether the captured signal rises above noise. A high SNR ensures usable images in 0.01 lux or even starlight conditions.
• Dynamic Range (DR)
  A sensor’s ability to capture details in both highlights and shadows simultaneously is critical for real-world night surveillance where streetlights and dark alleys coexist.
• Frame Rate & Latency
  Surveillance and robotics require real-time imaging with minimal lag. Latency is influenced by interface type (USB vs HDMI) and encoding method (MJPEG vs H.264).
• Optical Adaptability
  Lens compatibility—F/number, wide-angle vs telephoto, IR sensitivity—directly impacts night vision performance.

Side-by-Side Comparison of Sony STARVIS Sensors

The table below summarizes the core technical specifications of the IMX291, IMX335, and IMX415 sensors.

Parameter	IMX291	IMX335	IMX415
Sensor Type	1/2.8" STARVIS CMOS	1/2.8" STARVIS CMOS	1/2.8" STARVIS CMOS
Resolution	2.13 MP (1920×1080)	5 MP (2592×1944)	8.3 MP (3840×2160, 4K)
Pixel Size	2.9 µm × 2.9 µm	2.0 µm × 2.0 µm	1.45 µm × 1.45 µm
Minimum Illumination	0.005 lux (F1.2)	0.1 lux (F1.2)	0.09 lux (F1.6)
Dynamic Range	~120 dB (with WDR)	~72 dB	~72 dB
Frame Rate	1080p @ 60 fps	5MP @ 30 fps	4K @ 30 fps
Interface Options	USB2.0/3.0, AHD, CVBS	USB3.0, HDMI, AHD	USB3.0, HDMI
Typical Application	Extreme low-light starlight cameras	Balanced resolution vs sensitivity	High-resolution 4K surveillance

 

Sony STARVIS Sensor Tech Specs: IMX291 vs IMX335 vs IMX415

Parameter	Sony IMX291	Sony IMX335	Sony IMX415
Resolution	2.1 MP (1080p)	5.0 MP (2K)	8.3 MP (4K)
Optical Format	1/2.8"	1/2.8"	1/2.8"
Pixel Size	2.90 µm (Huge)	2.00 µm	1.45 µm (Small)
SNR1s (Low Light Index)	0.23 lx	0.59 lx	0.23 lx
Real-World Low Light	⭐⭐⭐⭐⭐ (Cleanest)	⭐⭐⭐ (Standard)	⭐⭐⭐⭐ (High Res)
Max Frame Rate	60 fps	30 fps	30 fps (at 4K)
Dynamic Range (HDR)	~120 dB (DOL-HDR)	~72 dB	~72 dB
Best For (Application)	Robot Nav / Extreme Dark	Industrial Inspection	Face Rec / OCR / City
Goobuy Solution	View IMX291 Module >>	View IMX335 Module >>	View IMX415 Module >>

 

Engineering Note on SNR1s & Pixel Physics:

You might notice that the IMX291 and IMX415 share a similar SNR1s value (0.23 lx). Does this mean they perform identically in the dark? No.

• The Physics: The IMX291 has a much larger pixel surface area (2.9µm vs 1.45µm), allowing it to physically collect 4x more photons per pixel than the IMX415.

• The Reality: In real-world analog environments, the IMX291 produces a "cleaner" image with less electronic gain noise in pitch blackness. The IMX415 achieves its high score through advanced sensor tech, but for absolute starlight conditions without auxiliary lighting, the large-pixel IMX291 (or its successor, the IMX462) is still the king of sensitivity.

• The Choice: Choose IMX415 if you need Resolution (4K). Choose IMX291 if you need Pure Night Vision.

 

Key Insights:

• IMX291 remains the most sensitive in ultra-low light due to its large pixel pitch, making it the go-to choice for starlight camera modules.
• IMX335 strikes a balance between detail and sensitivity, widely used in automotive, industrial, and mid-range surveillance.
• IMX415 delivers true 4K resolution, optimized for environments where detail is more critical than extreme low-light performance.

For engineers evaluating Sony STARVIS sensors, IMX291 is the best option for ultra-low-light starlight environments, IMX335 delivers the most balanced clarity and sensitivity for AI vision systems, and IMX415 provides high-resolution performance for bright, controlled lighting conditions. 

Night Surveillance Case – Great Lakes Region, USA

A real-world deployment highlights the trade-offs between these sensors.

• Scenario: A logistics hub near the Great Lakes required outdoor night surveillance with variable lighting—from bright floodlit docks to near-dark parking zones.
• Challenge: Standard 1080p IP cameras produced excessive noise in shadowed regions, while 4K cameras captured detail but struggled in low-light.
• Solution:
  • IMX291 modules were deployed at perimeter fences and parking zones where illumination drops below 0.01 lux. The starlight performance allowed clear identification of moving personnel and vehicles.
  • IMX335 modules covered medium-light areas like warehouse entrances, offering higher resolution for license plate recognition without compromising sensitivity.
  • IMX415 modules monitored floodlit docks, capturing high-detail 4K video streams for cargo verification and forensic evidence.

The combination ensured optimized coverage across lighting conditions while minimizing bandwidth and storage costs by selecting the right sensor for each zone.

 

How to Optimize Performance with USB/HDMI Outputs

Selecting the right sensor is only part of the engineering decision. Output interface and module design significantly affect system performance:

• USB Night Vision Modules
  Offer plug-and-play flexibility with UVC compliance across Windows and Linux. Ideal for prototyping, machine vision testbeds, and portable inspection devices.
• HDMI Outputs
  Provide low-latency, uncompressed video suitable for real-time monitoring (e.g., security control rooms, tunnel inspection systems).
• AHD Interfaces
  Still relevant for industrial retrofit projects where legacy coaxial infrastructure dominates. An AHD module for industry can deliver 1080p without requiring IP upgrades.

Optimizing latency involves selecting the right codec (MJPEG for rapid frame handling vs H.264 for compressed storage) and ensuring adequate host processing power.

Application-Specific Recommendations

Industrial USB3 Cameras

• Best low-light USB camera: IMX291

• Best general-purpose USB camera: IMX335

• Best 4K inspection camera: IMX415

Smart Retail, AI Kiosks & Analytics

• IMX335 wins due to clarity + low-light balance

• IMX415 only suitable for bright indoor stores

• IMX291 only for specialty dark environments

Warehouse Robots & AMRs

• IMX291 is the top choice

• IMX335 works in medium-light warehouses

• IMX415 typically not recommended

Security & Monitoring

• Starlight → IMX291

• Mixed lighting → IMX335

• Bright indoor → IMX415

Recommendation Framework for Engineers

Based on project requirements, here is a structured framework:

1. Ultra-Low Light (<0.01 lux, outdoor, rural, tunnels):
   → Choose IMX291 starlight camera module (best SNR and sensitivity).
2. Balanced Detail vs Sensitivity (indoor, logistics, robotics):
   → Choose IMX335 modules (ideal for 5MP inspection and LPR systems).
3. High Detail, Well-Lit Environments (urban, smart city, ports):
   → Choose IMX415 USB/HDMI modules (true 4K clarity).
4. Integration Considerations:
   • USB vs HDMI camera: Choose USB for software integration, HDMI for real-time video feeds.
   • AHD module for industry: Cost-effective solution for existing coaxial cabling systems.

 

Technical Selection Chart

Project Requirement	Recommended Sensor	Interface	Notes
Rural night surveillance	IMX291	USB2.0/3.0	Best for <0.01 lux
Smart factory inspection	IMX335	USB3.0	Balance of detail & sensitivity
4K city monitoring	IMX415	HDMI + USB3.0	True UHD clarity
Legacy coaxial upgrade	IMX291/IMX335	AHD	Industry retrofit option

Real-World Performance Comparison（Night Vision Focus）
IMX291 – Best for Starlight & Dark Factories

You can expect:

Minimal noise

Clear shapes and edges

Detectable human features at near darkness

Stable exposure even with small light sources

 

IMX335 – Best All-Around Performer

You get:

Clearer details than IMX291

Good low-light without overexposure

Balanced dynamic range

Suitable for multi-purpose embedded AI systems

 

IMX415 – Best for High Resolution in Bright Scenes

If lighting is sufficient:

Highest detail and clarity

Good for 4K analytics or displays

Not recommended for dark warehouses or outdoor night scenes

 

Compact Projects Note — Consider UC-501 (15×15 mm)

If your device enclosure is very space-constrained and cannot accommodate a full-size STARVIS module (IMX291 / IMX335 / IMX415), the Goobuy UC-501 USB camera offers a practical alternative.

This 15×15 mm 2 MP micro USB camera provides stable 1080p imaging with simple plug-and-play USB integration, making mechanical design and installation much easier in compact embedded or robotic systems.

It’s not intended to match STARVIS low-light performance, but for projects where size, simplicity, and integration feasibility matter more, Goobuy UC-501 can be an effective choice.  read its technical articles  15×15mm Goobuy UC-501 Micro USB Camera for tight spaces

 

Conclusion

Choosing the right night vision USB camera sensor requires careful consideration of application environment, illumination, and integration constraints. The IMX291 excels in starlight conditions, the IMX335 provides a balanced mid-tier solution, and the IMX415 offers high-resolution 4K imaging for controlled lighting scenarios.

 

Engineering FAQ
Q1. Which sensor performs best below 0.01 lux for starlight-level night vision?
IMX291. Its 2.9 µm pixel architecture provides the highest sensitivity and lowest noise among the three.

 

Q2. Why is IMX335 often preferred for AI boxes and retail analytics?
It delivers a rare balance of 5MP clarity + strong low-light performance, ideal for people detection, faces, objects, and analytics in indoor environments.

 

Q3. Is IMX415 suitable for industrial night-vision applications?
No. IMX415 requires bright illumination to perform well due to its small 1.45 µm pixels. It is designed for high-resolution daylight or controlled-light environments.

 

Q4. Which sensor is best for USB3 industrial cameras used in manufacturing?
IMX335 for most cases.
IMX291 when low-light is critical.
IMX415 for bright, high-resolution inspection.

 

Q5. How do these sensors impact neural-network inference accuracy?
Higher SNR (IMX291/IMX335) → better edge detection, fewer false positives, and cleaner feature maps for CNN/transformer models.

 

Q6. Does IMX415 offer any advantage over IMX291/IMX335 in night-vision pipelines?
No. IMX415’s advantage is resolution, not low-light performance.