Application Background and Detection Principle
Application Background
During transportation and storage, oranges are susceptible to bruising caused by impacts and compression. In the early stages, this damage often occurs beneath the peel without visible changes to the fruit’s external appearance, making it difficult to identify through conventional visual inspection.
If damaged oranges are not removed in time, the affected tissue may deteriorate and develop mold, leading to product losses and increased sorting costs.
This experiment evaluates the feasibility of using the AttosTek SWIR051AU short-wave infrared camera to detect and locate hidden bruising in oranges by comparing visible-light and SWIR imaging results.
Detection Principle
When an orange is bruised, its local tissue cells are damaged, allowing juice to accumulate in the affected area and increasing its local moisture content.
Water absorbs infrared light around the 1450 nm wavelength. This difference in water absorption can be used to distinguish healthy tissue from damaged areas through short-wave infrared imaging.
Under suitable imaging conditions, healthy regions produce stronger reflected signals and appear brighter, while bruised regions absorb more infrared light and appear darker.
Additional infrared illumination and optical filtering can further enhance the intensity differences between healthy and damaged areas, improving detection contrast.
Experimental Setup and Conditions
The experiment used the following equipment and conditions:
| Equipment / Component | Model / Specification |
|---|---|
| SWIR Camera | AttosTek SWIR051AU |
| Visible-Light Camera | AttosTek VIS028CU3.2-NC |
| Illumination Source | 1300 nm infrared light source |
| Optical Filter | 1300 nm infrared filter |
| Test Samples | One healthy orange and one bruised orange |
| Environment | Indoor environment, 25°C, with no additional light sources except the experimental illumination |
Imaging Results and Analysis
3.1 Qualitative Imaging Comparison
Four imaging conditions were compared to evaluate the detection of bruising in oranges.
Condition 1: Visible-Light Imaging

Images captured using the VIS028CU3.2-NC visible-light camera under natural light show the normal external appearance of the oranges. The subsurface bruising cannot be clearly identified from their visible appearance.
Condition 2: SWIR Imaging Under Natural Light

Images captured using the SWIR051AU reveal a noticeable grayscale difference. The damaged orange appears darker overall, with the bruised area showing an additional reduction in grayscale intensity.
Condition 3: SWIR Imaging with 1300 nm Illumination

After adding a 1300 nm infrared light source, the brightness difference between the healthy and bruised oranges becomes more pronounced. The boundaries of the damaged area are easier to distinguish.
Condition 4: SWIR Imaging with a 1300 nm Light Source and Optical Filter

With both the 1300 nm illumination source and optical filter, the image background becomes cleaner and the dark boundaries of the bruised area become clearer, providing the best imaging result among the four configurations.
3.2 Quantitative Contrast Analysis
To evaluate the distinction between healthy and bruised oranges, the normalized grayscale difference was calculated using the following formula:
C=(Imax−Imin)/255
Where:
- (Imax) represents the average grayscale value of the healthy orange.
- (Imin) represents the average grayscale value of the bruised orange.
- 255 represents the full grayscale range of an 8-bit image.
The measured results are shown below:
| Imaging Condition | Healthy Orange Grayscale | Bruised Orange Grayscale | Contrast |
|---|---|---|---|
| Direct SWIR051AU Imaging | 73.3 | 47.8 | 0.10 |
| With 1300 nm Illumination | 154.8 | 97.5 | 0.22 |
| With 1300 nm Illumination and Optical Filter | 206.7 | 103.7 | 0.40 |
The results show that adding 1300 nm illumination increases the normalized grayscale difference from 0.10 to 0.22.
When both the 1300 nm light source and optical filter are used, the contrast reaches 0.40, approximately four times the value obtained through direct SWIR imaging.
This demonstrates that appropriate illumination and optical filtering can significantly improve the visibility of subsurface bruising in oranges.
Conclusion
The experiment demonstrates the feasibility of using the AttosTek SWIR051AU short-wave infrared camera to detect and locate subsurface bruising in oranges that is difficult to identify through conventional visible-light imaging.
Among the tested configurations, combining a 1300 nm infrared light source with a matching optical filter provides the best imaging performance. The normalized grayscale difference increases from 0.10 to 0.40, reaching approximately four times the value obtained through direct SWIR imaging.
With its ability to distinguish healthy and damaged fruit regions through infrared imaging, the SWIR051AU offers a potential imaging solution for non-destructive fruit quality inspection.
This approach can be further evaluated for fruit sorting applications in food processing facilities and fresh produce sorting lines, as well as quality inspection of other easily bruised fruits such as apples, pears, and peaches.
Looking for an SWIR camera for fruit quality inspection?
Contact AttosTek to learn more about the SWIR051AU camera and its potential applications in non-destructive food inspection.