How to Choose an Industrial Camera: Matching Logic for Area vs. Line Scan, Resolution/Frame Rate, and Lens Interfaces

What Is an Industrial Camera

An industrial camera is an image acquisition device purpose-built for machine vision applications. Compared to consumer-grade cameras, its core differentiators lie in superior imaging stability, stricter interface standards, and robust adaptation to harsh industrial environments. Industrial cameras convert optical signals captured by image sensors into digital image data through standardized interfaces such as GigE Vision, USB3 Vision, Camera Link, or CoaXPress, transmitting to industrial PCs or embedded processors for subsequent visual analysis tasks including defect detection, dimensional measurement, and character recognition. Unlike security surveillance cameras, industrial cameras prioritize precise image capture for static or high-speed motion scenarios rather than continuous video recording.

---

Area Scan vs. Line Scan: Applicable Scenario Differences

Sensor Type Determines Imaging Method

Industrial camera sensors fall into two primary categories—Area Scan and Line Scan—fundamentally differing in pixel array geometry and triggered acquisition mode.

Area scan cameras feature a rectangular two-dimensional pixel array, capturing a complete two-dimensional image frame in a single exposure. Core selection parameters include resolution (horizontal pixels × vertical pixels), sensor format (e.g., 1/1.8", 2/3"), pixel size (e.g., 3.45μm, 5.86μm), and frame rate. Area scan cameras suit stationary targets or scenarios with controllable motion speeds. Typical applications include electronic component appearance inspection, PCB solder joint examination, and pharmaceutical vial fill-level monitoring.

Line scan cameras feature a single row or limited number of pixel rows, acquiring one line of image data per trigger. Complete two-dimensional images require stitching with continuous target motion. The core advantage of line scan cameras lies in achieving extremely high line rates, making them suitable for high-speed continuous production scenarios. Typical applications include film/paper/textile surface defect detection, steel surface quality inspection, and wood grain scanning.

Selection Criteria: Motion Characteristics and Field of View Requirements

When choosing between area and line scan, apply the following decision logic:

Practical Steps: Assess target motion characteristics → Determine whether single-piece inspection or continuous material inspection applies → For continuous material inspection, further evaluate motion speed and defect detection precision requirements → Prioritize line scan for high-speed + large-area needs; otherwise, prioritize area scan.

---

Resolution vs. Frame Rate: Precision-Speed Trade-off

Mutual Constraints of Core Parameters

Resolution and frame rate are the two most frequently cited parameters in industrial camera selection, yet they exhibit physical-level constraints on each other. Camera data bandwidth is fixed, with interface protocols limiting maximum transmissible data volume per unit time. When resolution increases (more pixels), single-frame image data volume grows, and achievable frame rate under fixed bandwidth necessarily decreases; conversely, pursuing high frame rate requires reducing single-frame resolution.

Taking GigE Vision interface as an example, gigabit Ethernet's effective payload bandwidth is approximately 120MB/s (accounting for protocol overhead). A 2448×2048 pixel, 8-bit grayscale image occupies approximately 5MB, yielding a theoretical frame rate of approximately 24fps per interface under ideal conditions. Actual projects typically require reserving 20–30% bandwidth margin for system control and retransmission mechanisms.

Detection Precision Requirements Determine Resolution Floor

Resolution selection should start from detection precision requirements. The industry-standard empirical formula is:

Camera Resolution (pixels) = Field of View (mm) × Customer-Required Detection Precision (pixel/mm)

Detection precision (pixel/mm), also called the "resolution factor," represents the reciprocal of physical size per pixel. For instance, if field of view is 100mm × 80mm with a 0.1mm detection precision requirement (capable of identifying defects above 0.1mm), the horizontal direction requires at least 1000 pixels, vertical direction at least 800 pixels. Considering edge redundancy, selecting a camera with 1280×1024 or higher resolution is recommended.

Practical Checklist:

1. Clarify target field of view (FOV)
2. Determine customer or project requirements for minimum detectable defect size
3. Calculate required minimum pixel count = FOV size ÷ Defect precision
4. Select specifications that just meet or slightly exceed requirements from standard resolution series (e.g., VGA/1.3MP/2MP/5MP/12MP/20MP)
5. Verify whether frame rate under selected resolution meets production cycle requirements

Production Cycle Determines Frame Rate Floor

Frame rate selection should base on production line maximum motion speed and detection-required exposure time. Core calculation formula:

Minimum Frame Rate (fps) = Target Maximum Motion Speed (mm/s) ÷ Physical Size per Pixel (mm/pixel) ÷ Allowable Pixel Blur (pixel)

"Allowable pixel blur" is the key parameter controlling motion blur. For stationary targets, it can be set to 0 (no motion blur requirement); for moving targets, typically 1–2 pixels blur is acceptable to gain longer exposure time (thereby reducing light source brightness requirements).

Example: Conveyor speed is 500mm/s, pixel size is 5μm, allowable blur is 1 pixel. The correct understanding is: within 1 second, the target travels 500mm, and each pixel corresponds to 0.005mm, meaning the target traverses 100,000 pixel positions in 1 second. To ensure the target moves no more than 1 pixel during single-frame exposure, exposure time must be ≤ 1/100,000 second = 10μs. This imposes extremely high requirements on light source brightness.

In practical engineering, the more commonly used method calculates "equivalent line frequency" or directly uses the motion blur formula to estimate allowable maximum exposure time:

Maximum Exposure Time (ms) = Allowable Blur Pixels × Pixel Size (mm) ÷ Motion Speed (mm/s)

---

Interface Type Technical Characteristics and Selection Matching

Mainstream Interface Protocol Comparison

Industrial camera interface protocols affect not only image transmission bandwidth but also cable length, anti-interference capability, system cost, and integration complexity. Below is a horizontal comparison of three mainstream interfaces:

Interface Selection Scenario Matching

GigE Vision suits projects requiring long cable distances, flexible multi-camera system expansion, or cost sensitivity. In Southeast Asian industrial automation projects, distances from factory floor to control rooms often exceed 20 meters, with potential for subsequent multi-camera expansion. GigE interface represents the mainstream choice balancing flexibility and cost-effectiveness. Within HuaRay machine vision product lines, standard-resolution area scan cameras predominantly offer GigE Vision interface versions, facilitating compatibility with standard industrial Ethernet switches.

USB3 Vision outperforms GigE in bandwidth and latency, suitable for scenarios requiring higher frame rates but with transmission distances not exceeding 5 meters. Typical applications include high-speed pick-and-place machine inspection and rotating surface inspection. In compact vision systems, USB3's plug-and-play characteristics can also reduce integration complexity.

Camera Link targets precision measurement scenarios with extreme bandwidth and real-time requirements, such as high-speed motion analysis and sub-pixel-level measurement systems. Its hardware trigger synchronization's deterministic latency represents an advantage irreplaceable by other interfaces. However, dedicated frame grabber costs and system integration complexity are higher, recommending Camera Link only when genuinely necessary.

Practical Recommendations: Prioritize GigE Vision interface unless transmission distances are extremely short with stringent frame rate/latency requirements, then consider USB3 Vision; Camera Link applies only to precision measurement or high-speed inspection special requirement scenarios.

---

Lens and Light Source System Matching Logic

Core Parameter Chain for Lens Selection

Lens selection is not independent from the camera but requires joint calculation with camera sensor format, field of view, and working distance. Core parameters include:

Focal Length (f) = Working Distance (WD) × Pixel Size (requiring consideration of sensor format-to-FOV ratio)

Simplified formula: Focal Length = Working Distance × (Sensor Format / Field of View)

Sensor format requires using the camera image sensor's actual physical dimensions (e.g., 2/3" sensor diagonal approximately 11mm, 1/1.8" approximately 8.9mm), rather than direct resolution conversion.

Aperture (F-value) affects depth of field and light intake. Smaller F-values yield shallower depth of field and greater light intake. High-precision inspection typically requires larger depth of field to cover target height variation ranges; consider using smaller apertures or telecentric lenses in such cases.

Telecentric vs. Standard Lenses: When targets exhibit significant depth variation or when measurement precision requirements are stringent, telecentric lenses should be selected. Through special optical design, telecentric lenses eliminate perspective error, ensuring consistent imaging magnification at varying working distances—suitable for precision dimensional measurement scenarios. Standard lenses offer lower cost, suitable for appearance inspection with lower precision requirements.

Light Source Selection and Illumination Methods

Light sources provide stable, uniform, controllable illumination for target features, directly affecting image contrast and defect detection rates. Light source selection requires considering:

Light source type: Common types include backlights (coaxial), front lights (diffuse/bright field), dark field lights, ring lights, spot lights, bar lights, etc. Different light angles and methods produce varying surface reflection characteristics.

Wavelength selection: Different materials exhibit significantly varying reflectance across wavelengths. For example, red light (625nm) suits detecting scratches on metal surfaces; blue light (470nm) enhances contrast between organic materials and plastics; infrared light suits penetrating some semi-transparent materials. In food and pharmaceutical packaging inspection, wavelength selection often requires testing verification combined with actual material characteristics.

Light source controller: High-quality constant-current light source controllers ensure stable light output, eliminating power frequency flicker (50Hz/60Hz) effects on images. For high-speed inspection or precision measurement scenarios, configuring independent light source controllers rather than relying on simple camera GPIO switching is recommended.

Practical Checklist:

1. Clarify target feature (scratch/dent/character/dimension) and background contrast requirements
2. Determine target material and surface characteristics (flat/curved/reflective/diffuse)
3. Select light source type and wavelength; recommend initial comparative testing using standard light source kits
4. Calculate light source size matching with FOV (light source should cover FOV with margin)
5. Configure constant-current light source controller to eliminate light source fluctuation effects on images

---

Practical Recommendations: Five-Step Checklist for Industrial Camera Selection

Step 1: Clarify Inspection Tasks and Performance Indicators

Before engaging with specific products, clarify the following with end customers or project managers: Are detection targets static or dynamic? What defect types need detection? What is the minimum defect size? What is the production cycle requirement in pieces/minute or meters/minute? What are the allowable miss rate and false alarm rate? This information directly determines camera type, resolution, and frame rate selection starting points.

Step 2: Determine Camera Type and Resolution

Select area or line scan cameras based on inspection task motion characteristics. For area scan cameras, prioritize specifications slightly exceeding calculated resolution values for margin; for line scan cameras, select appropriate line width models (e.g., 2K/4K/8K) based on FOV width and line rate requirements.

Step 3: Verify Frame Rate and Interface Bandwidth

Calculate minimum frame rate requirements using production cycle and motion speed; verify whether frame rate under selected resolution meets requirements. If frame rate is insufficient, consider reducing resolution or upgrading to higher-bandwidth interfaces (e.g., from GigE to 5-GigE bonding or USB3 Vision).

Step 4: Match Lens and Light Source

Calculate lens focal length based on working distance and FOV; select lens models supporting corresponding sensor formats. If detection precision requirements are high or depth variation exists, prioritize telecentric lenses. Light source selection should combine material characteristics and contrast requirements; illumination testing to verify effectiveness is recommended.

Step 5: Reserve Expansion and Maintenance Space

Reserve 20–30% performance margin during selection to prevent system insufficiency under extreme conditions. Simultaneously confirm selected brand's spare parts supply cycles, technical support response speed, and localized service capabilities. For the Southeast Asian market, integrators should prioritize collaborating with suppliers maintaining localized service networks in both China and Southeast Asia.

---

FAQ

Q1: Which is more suitable for continuous production line inspection, area scan or line scan cameras?

This depends on production line motion characteristics and inspection requirements. If targets pass through the FOV in a continuous, uniform manner (e.g., roll materials, film, paper), line scan cameras are the more suitable choice, with line rates reaching dozens of kHz or higher. If targets enter the FOV as individual pieces (e.g., PCB boards, vials, electronic components), area scan cameras are more appropriate, with simpler installation and debugging.

Q2: How to calculate camera resolution required for detection precision?

Core formula: Required pixels = FOV size ÷ Minimum detectable defect size. Recommend multiplying by 1.5–2× redundancy factor to accommodate actual factors such as lens distortion and motion blur. For example, with FOV of 100mm × 80mm and 0.1mm defect precision requirement, the horizontal direction requires at least 1000 pixels; recommending selection of 1280 or higher resolution cameras.

Q3: What is the maximum stable transmission distance for GigE Vision interface in actual projects?

Standard gigabit Ethernet single cable maximum transmission distance is 100m—this is the physical layer specification limit. In industrial environments, recommending actual usage distances controlled within 80m, with shielded network cables, proper grounding measures, and industrial-grade Ethernet switches to ensure signal stability. For greater distances, fiber transceivers or switch cascading can extend reach.

Q4: When industrial camera frame rate is insufficient, what solution approaches exist?

Common solution approaches include: reducing single-frame resolution (e.g., from 5MP to 2MP) to decrease per-frame data volume; upgrading to higher-bandwidth interfaces (e.g., from GigE to USB3 Vision or 5-GigE bonding); employing Region of Interest (ROI) reading to transmit only necessary image regions; or selecting higher frame rate camera models.

Q5: How to match lens focal length with camera sensor format and working distance?

Simplified calculation formula: Focal Length = Working Distance × (Sensor Format / FOV). Sensor format requires using the camera image sensor's actual physical dimensions (e.g., 2/3" sensor corresponding diagonal approximately 11mm). Recommending use of industrial lens manufacturer selection tools or consulting technical support for more precise matching recommendations.