Abstract
In a previous study, a fresh chili destemming system has been proposed, and the TCS3200 color sensor was used as low-cost equipment in the arrangement section of the system. However, the working and installing parameters have not yet been investigated in detail. This study focused on investigating the optimal working conditions of TCS3200 on the fresh chili destemming system. The study included three experiments including the effect of the red and green color channels, the effect of the height between the sensor and fruits, and the effect of the running velocity on the received signal level of the sensor. A total of 260 ripen fruits were chosen to test. The results showed that the red channel generated a stronger signal than the green one. The optimal height from the sensor to the fruits was obtained at 25 mm, and the maximum conveyor velocity is limited to 100 mm/s. This study helps to determine the optimal operating parameters and enhance the working ability of a low-cost sensor.
Introduction
Chili is a vital agricultural product in the Mekong Delta, yet its processing—specifically destemming—remains largely manual and labor-intensive. While automated solutions exist, they often rely on expensive computer vision hardware. A previously proposed mechanical system uses a "clamping belt" to pull stems, but its success depends on the precise arrangement of chilies on a conveyor. This arrangement phase utilizes the low-cost TCS3200 color sensor to trigger a pusher mechanism.
The effectiveness of this arrangement module depends on several critical parameters: the sensor's installation height, the color channel used for detection, and the conveyor's operating speed. This study systematically evaluates these factors to establish optimal working conditions for industrial application.
Methodology
The experimental rig consists of a V-shaped groove conveyor, a TCS3200 color sensor, and an Arduino Uno R3. The TCS3200 is a light-to-frequency converter that outputs a square wave proportional to light intensity. The Arduino measures the signal period ($T$) to detect the presence of a chili against the black background of the conveyor.
A sample of 260 fully ripe red chilies (Chanh Phong F4 variety) was used for testing. The researchers conducted three primary experiments:
- Color Channel Evaluation: Comparing Red vs. Green channels for signal clarity.
- Height Optimization: Testing sensor heights ($h$) from 15 to 40 mm.
- Velocity Testing: Evaluating conveyor speeds ($v$) from 10 to 180 mm/s.
Results and Discussion
Color Channel and Height
The Red channel provided a significantly stronger signal than the Green channel when detecting red chilies against a black background. This higher signal-to-noise ratio allows for more reliable thresholding in the control software, reducing false negatives.
Regarding installation height, a "sweet spot" was identified at 25 mm. At lower heights (15-20 mm), the sensor's field of view is too narrow, reducing radiation intensity. Conversely, at higher positions (>30 mm), the signal becomes too weak and diffused. At 25 mm, the contrast between the empty conveyor and the fruit is maximized.
Conveyor Velocity
As speed increases, the detection window for each chili narrows. The study found that while the system can function at 120 mm/s, signal degradation becomes problematic at 140 mm/s. A recommended operating limit of 100 mm/s ensures a safety margin for industrial stability, enabling a throughput of approximately 2.5 fruits per second.
Conclusion
This research successfully identifies the optimal operating parameters for the TCS3200 sensor in a chili destemming system: a 25 mm installation height using the red color channel at a maximum speed of 100 mm/s. These findings demonstrate that low-cost sensors can achieve performance levels comparable to expensive vision systems when properly calibrated. Future work will address automated feeding and mechanical stability at high speeds.
Summary updated