Creating those datasets the traditional way is a grind that consumes the patience of even the most dedicated entomologists. Each image captured by a light trap may contain dozens of insects at wildly different sizes, overlapping bodies, broken wings, and ambiguous orientations. An expert must draw a box around every insect, assign it to the correct species or category, and repeat this thousands upon thousands of times. The work is slow, expensive, and entirely dependent on scarce specialists whose time might otherwise be spent on actual research. A team of researchers from Zhejiang Sci-Tech University and collaborating institutions in China has now unveiled a solution that could dissolve this bottleneck, and their results, published in the Journal of Integrative Agriculture, suggest that the era of exhaustive manual insect annotation may be drawing to a close.

The team’s approach is a two-stage framework that separates the problem of dataset construction into detection and classification, tackling each with a purpose-built system. The first stage, called MLTIDD—short for multi-scale light-trapped insect detection and description—addresses one of the most technically thorny challenges in agricultural computer vision: the extreme disparity in scale between large and tiny insects within a single image. A large moth might occupy hundreds of pixels, while a minuscule midge may span barely a dozen. Standard object detectors, which rely on receptive fields tuned to typical object sizes, systematically miss the smallest specimens or fragment the largest ones.

To overcome this, the researchers fine-tuned Grounding DINO, a vision-language foundation model capable of open-vocabulary detection, meaning it can locate objects based on text descriptions rather than requiring exhaustive training on fixed categories. But foundation models alone were not enough. The team integrated SAM, the Segment Anything Model developed originally for general-purpose image segmentation, and SAHI, a slicing-aided hyper-inference technique that chops large images into overlapping tiles, runs detection on each tile, and then stitches the results back together. This combination allows the system to detect insects at multiple scales simultaneously: SAHI ensures that tiny insects magnified within tiles are not overlooked, while the fine-tuned Grounding DINO provides semantic understanding of what constitutes an insect even in cluttered, chaotic light-trap scenes.

The performance gains were substantial. In experiments conducted with only limited annotated samples—a few-shot setting that mirrors the real constraints of agricultural research labs—MLTIDD achieved 79.6 percent average precision at the strict AP50–95 threshold and 90.8 percent average recall. Compared against a standard DINO baseline, this represented improvements of 7.0 and 4.7 percentage points respectively. When the complete pipeline was evaluated against the combined DINO and iBOT setup, the detection stage improved mean AP50–95 by 10.91 percentage points and average recall by 8.26 percentage points. For a field where even a few percentage points can determine whether a monitoring system is practical or prone to dangerous miscounts, these are meaningful margins.

Detection, however, is only half the battle. Once thousands of insect sub-images have been sliced out of the original photographs, each one must be classified. This is where the second innovation, InsectSSRL, comes into play. Standing for insect self-supervised representation learning, InsectSSRL is built on iBOT, a self-supervised vision transformer framework that learns by masking parts of an image and training the network to predict what is hidden. Crucially, it requires no labels at all during pre-training—instead, it exploits the massive trove of unlabeled insect sub-images that MLTIDD itself generates, turning the detection stage’s output into fuel for the classification stage’s learning.

InsectSSRL enhances its feature extraction through three proxy tasks, teaching the model to capture the fine-grained visual signatures that separate insect species: wing venation patterns, body segmentation, coloration, and texture. The result is a feature extractor that, when evaluated with a k-nearest-neighbors classifier—a test of how well the learned features cluster similar insects together—achieved 85.87 percent top-1 accuracy. The true measure of its value emerged in few-shot classification experiments. When a Swin-T vision transformer pre-trained with InsectSSRL was fine-tuned on just 5 percent of the InsectID dataset, it reached 80.35 percent classification accuracy. That figure exceeded an iBOT baseline by 2.08 percentage points and surpassed a model trained with conventional transfer learning from COCO, a generic object-detection dataset, by a striking 11.3 percentage points. The message is clear: features learned from insect imagery, in a self-supervised manner, transfer to insect classification far better than features borrowed from unrelated domains.

The complete pipeline works as a virtuous cycle. MLTIDD detects and crops insects from raw light-trap images. InsectSSRL consumes those unlabeled crops to learn robust representations, then supports a classification model that pre-classifies each insect. Human experts then step in—but only to correct and verify the machine’s suggestions rather than to label everything from scratch. Finally, the corrected labels are traced back to the original images, completing the annotation of the entire dataset. This human-in-the-loop design preserves the quality guarantee that only expert knowledge can provide, while eliminating roughly 80 percent of the time experts would otherwise spend on manual labeling.

That 80 percent reduction is the headline number with the most immediate practical consequence. Light-trap monitoring networks can generate millions of images per season across wide geographic areas. If every image demanded full manual annotation, scaling these networks to national or continental coverage would be economically impossible. By slashing expert workload to a verification role, the new framework makes large-scale, multi-class insect datasets achievable by ordinary research teams. Better datasets, in turn, mean better AI models in the field—models that can distinguish pest outbreaks from benign insect activity with the reliability that precision agriculture demands.

The implications extend well beyond pest monitoring. Light traps are also used in biodiversity surveys, ecological monitoring, and studies of insect decline, where automated species identification is equally hampered by annotation scarcity. A framework that can bootstrap a high-quality dataset from a small number of labeled examples and a mountain of unlabeled images is exactly the kind of tool those fields need. The combination of open-vocabulary foundation models, segmentation models, tiled inference, and self-supervised learning represents a template that could be adapted to any domain where objects are small, numerous, and taxonomically diverse—from plankton imaging to medical cell counting.

There are, of course, caveats worth noting. The framework’s classification accuracy, while impressive, still falls short of the near-perfect reliability that fully autonomous pest-control decisions would require, which is precisely why expert correction remains embedded in the pipeline. And the approach’s reliance on fine-tuned foundation models assumes access to computational resources that some labs may lack. Yet the trajectory is unmistakable. As corresponding author Professor Qing Yao of Zhejiang Sci-Tech University puts it, the method improved label quality, decreased expert workload, and increased annotation efficiency simultaneously—a rare trifecta in dataset engineering, where improvements in one dimension typically come at the expense of another.

What began as a practical problem—too many insects, too few annotators—has yielded a demonstration of how modern vision foundation models can be orchestrated into something greater than the sum of their parts. Grounding DINO finds the insects, SAM delineates them, SAHI ensures the smallest are not lost, and InsectSSRL learns to name them, all while human experts supervise from a comfortable distance. The moths still fly into the light. But now, the machines that watch them are teaching themselves to see.

Mangoes, often dubbed the “king of fruits,” hold substantial economic significance, especially in tropical countries. Their market value is closely tied to their quality, which necessitates precise and objective grading methods to meet consumer standards. Traditionally, mango grading has relied heavily on manual labor, which is error-prone and inefficient. Masum and his team recognized the pressing need for a more robust system that could elevate the grading process to the next level through automation, hence their focus on machine vision technology.

The core of their research lies in the application of computer vision algorithms that are capable of analyzing a mango’s physical characteristics. Key metrics assessed include size, shape, color, and surface blemishes. The integration of these parameters allows the machine to make informed decisions regarding a mango’s quality. The research details how this technology employs high-resolution cameras and sophisticated software to capture and process images of mangoes on a conveyor belt, ensuring that the quality assessment occurs in real-time as the fruits move from processing to packaging.

One of the most compelling aspects of this technology is its versatility. The machine vision system can be fine-tuned to evaluate various mango varieties, detecting subtle differences that may be imperceptible to the naked eye. For instance, the research highlights the capability of the algorithm to classify mangoes into different grades such as top, medium, and low quality, which is critical for effectively managing inventory and meeting market demands. By automating this grading process, producers can better align their products with consumer preferences, thus maximizing profitability.

The researchers also addressed the potential challenges associated with implementing such a technology in existing supply chains. They examined the costs involved in integrating machine vision systems into traditional farming and packaging operations. Notably, their findings suggest that while the upfront investment may be significant, the long-term gains through enhanced efficiency and reduced labor costs could outweigh initial expenditures. This shift towards automation could potentially redefine how mango grading and sorting is approached globally.

Environmental sustainability was another pivotal aspect of the study. The researchers pointed out that by minimizing the number of discarded fruits due to grading errors, the machine vision system contributes to reducing waste in the agricultural sector. This aligns with global sustainability goals, as less food waste directly translates into a more responsible and efficient use of resources. Moreover, the reduction in labor requirements could free up human resources for other critical tasks within the supply chain, fostering a more balanced allocation of labor.

In practical applications, farmers and producers have already started to report noticeable changes in their grading processes after incorporating the newly developed automated systems. This practical deployment indicates a strong shift towards embracing technology to bolster agricultural productivity. Feedback from early adopters of the technology has revealed a marked improvement in sorting accuracy and speed, significantly impacting their operational efficiency.

A potential concern regarding machine vision systems lies in their reliability under varied conditions such as lighting and the presence of dust or obstructions. However, Masum’s team has conducted extensive testing in diverse environments to ensure the systems maintain their efficacy. Their research provides compelling evidence that these automated systems can function optimally even in less than ideal conditions, showcasing their robustness and adaptability.

Furthermore, the researchers explored the implications of utilizing artificial intelligence to enlarge the capabilities of machine vision systems. By incorporating machine learning models, the technology can continuously learn and adapt from new data, further refining its grading accuracy over time. This approach not only enhances the immediate usability of the system but also prepares it for future advancements in agricultural practices.

The methodology employed in the study presents a comprehensive framework that can be adapted for other fruits and agricultural products, suggesting a larger application for the discoveries made in mango grading. This transferable nature of the technology could herald a new age for agricultural automation, revolutionizing the way industries approach quality assessment across multiple types of produce.

As the study gains traction, the implications of this research extend beyond the agricultural sector. The integration of automated grading systems could inspire similar innovations in food processing industries, where efficiency and quality control are paramount. The cascading effects of this technology could contribute to enhancing food safety and standardization across borders, ensuring that consumers receive only the best quality produce.

Overall, the work conducted by Masum et al. offers promising prospects for the intersection of agriculture and technology. As the world grapples with food production pressures due to growing populations, automated solutions such as the proposed mango grader could play a pivotal role in meeting these demands. The successful implementation of machine vision technology stands to reshape the agricultural landscape, leading to increased output and sustainability in food systems.

In conclusion, the development of an automated real-time mango grader using advanced machine vision techniques represents a significant milestone in agricultural innovations. The potential to enhance efficiency, improve food quality, and contribute to sustainability makes this research noteworthy and inspiring. As technology continues to evolve, it is initiatives like this that illuminate the path towards a more efficient and sustainable agricultural future.

Subject of Research: Machine Vision Technology for Automated Mango Grading

Article Title: Development of automated real-time mango grader using machine vision technique

Article References:

Masum, A., Himel, M.M.H., Salehin, M.M. et al. Development of automated real-time mango grader using machine vision technique.
Discov Agric 3, 104 (2025). https://doi.org/10.1007/s44279-025-00281-w

Image Credits: AI Generated

DOI:

Keywords: Machine Vision, Agricultural Technology, Mango Grading, Automation, Sustainability