Vole Hole Detection - Wortel AI annotation portfolio

Sample frames for this project are available under NDA. Write to bilal@wortel.ai and we will share the relevant before and after pairs.

Challenges

Identifying small vole holes against complex soil and vegetation backgrounds.
Telling genuine vole holes apart from natural soil patterns, rocks and shadows.
Handling holes that were only partially visible or obscured.
Soil conditions, lighting and image quality varied widely across the dataset.
Keeping bounding boxes tight and consistent around irregularly shaped holes.

Project overview

Agricultural field images were annotated for a computer vision dataset focused on detecting vole holes in agricultural environments. The work involved identifying every visible vole hole and producing accurate annotations to generate high-quality training data for an AI-based object detection model.

Vole holes are small, clean-cut openings in the ground measuring roughly one to two inches wide. They connect to surface runways and underground tunnels created by voles, small rodent pests that eat plant roots, bulbs and bark.

Dataset

Images: 1,000+
Classes: 1
Annotation: Bounding box detection
Source: Client-provided
Tool: CVAT
Client: Private client

Industry

Agriculture

Annotation type

Bounding Box

Labels

Vole_Hole

Related work

Ground Cover Quadrat Segmentation

2,275 cover-crop plot images annotated in Labelbox: plot corners boxed, then every pixel classed as groundcover, weed, residue, crop or soil.

Almond Tree Detection

Bounding boxes around individual almond trees in top-down drone imagery, where shadows and neighbouring vegetation mimicked real trees.

Apple Detection

Bounding boxes around individual apples on tree branches, where dense leaves and occlusion hid much of each fruit.