A traffic management system typically requires a camera sensor sensitive enough to capture images in all lighting conditions, regardless of sunny, cloudy, or rainy weather. The camera sensor should also be capable of adapting to changing day-and-night conditions.
Traditionally, the approach has been to choose a sensor with good NIR sensitivity. It can produce a good output in daylight conditions while enabling good low-light (or night) performance with the help of an IR LED. However, this requires an additional IR cut/pass switch in the lens holder setup. However, using an IR switch and IR LEDs is expensive, and it comes with its own set of design complications such as:
- Handing the IR cut/pass options through software algorithms
- Using ambient light sensors to switch between IR cut/pass options
Advent of low-light camera sensors
Nowadays, smart traffic management system manufacturers look for a camera solution that seamlessly delivers high-quality image outputs in all lighting conditions without mechanical IR switches or IR light sources. Today, with advancements in sensor technology, many have developed sensors specifically designed to work in extremely low-light conditions. Some of the most popular ones belong to the Sony STARVIS series – equipped with a minimum sensitivity of 2,000 mV/μm2.
While a low-light camera would be sufficient for majority of the smart traffic use cases, certain applications may require IR lighting and imaging. These include Automatic Number Plate Recognition (ANPR) and Facial Recognition of passengers inside a vehicle.
However, we will focus on using a low-light sensor without IR lighting or switch for this specific customer story (to be discussed later on in this blog).
Low-light camera module from e-con Systems for smart traffic management
One of the most popular sensors in the STARVIS series is IMX327. It comes with a resolution of 2MP and the HDR (High Dynamic Range) capability. e-con Systems offers a Sony STARVIS IMX327-based MIPI camera module called e-CAM220_CUMI327_MOD. This low-light camera module is compatible with various NVIDIA Jetson processors. A combination of these factors makes this product suitable for a vast majority of smart traffic applications such as vehicle counting, pedestrian tracking, vehicle detection, etc.
How e-CAM220_CUMI327_MOD empowered a leading European smart traffic manufacturer
Instead of talking about how great the product is, let’s look at the story of how e-con Systems successfully integrated the camera module into a customer’s product!
To start with, the customer was looking for a camera solution to capture quality images in daylight and low-light conditions for the NVIDIA Jetson Xavier NX platform. e-con Systems suggested e-CAM220_CUMI327_MOD, which is a raw Bayer camera module based on the SONY STARVIS IMX327 CMOS sensor with a MIPI CSI interface. We provided a driver to utilize the internal ISP available with Jetson Xavier NX to get the processed color data. In addition, this camera supported a frame rate of up to 60 fps in 1080p, making it the best-fit choice for the use case.
We initially used a high aperture lens (F#1.2) to utilize the ultra-low-light sensitivity offered by the SONY STARVIS IMX327 sensor. However, the customer faced overexposure issues due to the high sensitivity of this camera on sunny days as shown below.
A solution for this issue is to use a lens with a lower aperture (F#2.8), thereby avoiding overexposure in daylight conditions.
Have a look at the below comparison to see the difference between images taken using the large aperture and small aperture lenses.
Though this solution resolved overexposure issues during sunny conditions, using a lens with a low aperture affected the low-light performance of the camera, which is absolutely critical for smart traffic monitoring.
However, with e-con’s expertise in ISP tuning and a thorough understanding of the image sensor, we resolved the issue and provided a workable solution to the customer. Some of the highlights of the changes we made include:
- Changing the conversion gain mode from High Conversion Gain (HCG) to Low Conversion Gain (LCG), which reduced the gain of the overall frame
- Changing auto exposure target brightness, converge speed and slop factor in ISP configuration parameters
- Changing the minimum exposure value from 450 μs to 33 μs since the sensor can go only up to 450us (minimum) as per the BSP and V4L2 configuration
Below is a comparison of images taken pre and post-LCG and ISP finetuning, and changing minimum exposure to 33us.
On LCG mode, effective ISP tuning – along with minimum exposure of 33 microseconds – helped resolve overexposure issues while retaining the low-light capabilities of the camera module. With this customized solution, the customer was able to easily detect vehicles on bright sunny days as well as during night-time (low-light) conditions.
Below are images taken by the customer during day as well as night using e-con Systems’ camera solution.
Hope you got a fair understanding of how the IMX327 based custom camera solution from e-con Systems helped the smart traffic manufacturer achieve their imaging needs.
Are you looking for help in picking and integrating the right camera into your smart traffic system? If yes, please drop a note to email@example.com. Our camera experts will be happy to assist you.
Related e-con cameras:
Here is a list of camera solutions e-con offers with the Sony STARVIS IMX327 sensor.
In addition to the IMX327 based camera module, e-con Systems’ portfolio also includes cameras with a few other Sony STARVIS sensors such as IMX290, IMX415, and IMX462. Given below is a list of USB cameras and camera modules with these sensors:
- See3CAM_CU27 – Sony STARVIS IMX462 ultra low light USB NIR camera
- e-CAM82_USB – Sony STARVIS IMX415 4K Ultra-HD USB camera
- Hyperyon – Sony STARVIS IMX290 ultra low light USB camera
- e-CAM221_CUMI462_MOD – Sony STARVIS IMX462 low light camera module
- e-CAM83_CUMI415_MOD – Sony STARVIS IMX415 low light camera module
- e-CAM27_CUMI290_MOD – Sony STARVIS IMX290 low light camera module
To better understand where these sensors stand in comparison with each other, we have also written an article Similarities and differences between Sony STARVIS IMX290, IMX327, and IMX462.
If you have any questions around any of these products, or are looking for help in integrating cameras into your products, please write to us at firstname.lastname@example.org.
Vinoth Rajagopalan is an embedded vision expert with 15+ years of experience in product engineering management, R&D, and technical consultations. He has been responsible for many success stories in e-con Systems – from pre-sales and product conceptualization to launch and support. Having started his career as a software engineer, he currently leads a world-class team to handle major product development initiatives