An explainable end-to-end computer vision pipeline for detection, segmentation, and reconstruction of…

AI / informaticsFrontiers in artificial intelligence· 2026

An explainable end-to-end computer vision pipeline for detection, segmentation, and reconstruction of occluded weapons in forensic imagery.

Vaibhav Rohella, Kumar Anurag, Aditya Kumar … Shanthi P

Medical Imaging AIAI / informaticsQ1IF 7.8

AI model detects partially concealed weapons in crime scene images with mAP@50 of 0.86

An explainable end-to-end computer vision pipeline for detection, segmentation, and reconstruction of occluded weapons in forensic imagery.

Vaibhav Rohella … Shanthi P

Frontiers in artificial intelligence · 2026 · Q1 in Artificial Intelligence

Background

Images from crime scenes often show partially concealed weapons due to obstructions such as hands and clothing, as well as surveillance camera limitations, which affect the efficacy of traditional detection methods.

Purpose

This work proposes an explainable forensic pipeline for occluded weapons detection, segmentation, and reconstruction.

Methods

The proposed framework integrates RT-DETR-L, a transformer-based weapon detection model; MobileSAM for zero-shot segmentation of visible weapon regions; Stable Diffusion Inpainting for reconstruction of occluded regions; and a bilateral filter with Canny edge detection for forensic sketch generation.

Results

0.86

model correctly flagged weapons in most images despite partial concealment

More results

Reconstruction quality meets minimum quality thresholds (SSIM > 0.54, PSNR > 16 dB, Region IoU > 0.80) up to approximately 50% occlusion.

An interactive Streamlite application demonstrates the pipeline's feasibility in 42-46 seconds as a controlled laboratory prototype.

“

Conclusion

The generated reconstructions and sketches are intended as visual aids for expert review and must undergo independent forensic validation before any operational use.

An explainable end-to-end computer vision pipeline for detection, segmentation, and reconstruction of occluded weapons in forensic imagery.

An explainable end-to-end computer vision pipeline for detection, segmentation, and reconstruction of occluded weapons in forensic imagery.

Related papers