You will get a reliable people and vehicle tracking and counting tool
Project details
Need to count people or vehicles crossing a defined line in recorded fixed-camera footage? I will configure and deliver a Python/OpenCV tracking and counting solution around your video, target classes, direction rules, required outputs, and hardware.
Depending on the selected tier, deliverables can include an annotated MP4 with track IDs and detector confidence, per-class directional totals, timestamped CSV/JSON events, reusable configuration, source code, setup documentation, and a measured performance report.
Before implementation, I review a representative clip and define exactly what creates one count. Outputs are checked for readable video, processed/output frame totals, reproducible settings, and incomplete-run errors. The public demo processed 647 frames at 78.2 end-to-end FPS on the test CPU, but performance is measured again for your footage and target hardware.
Formal accuracy is reported only against human-verified ground truth. Client footage remains private and is never published without written permission.
Depending on the selected tier, deliverables can include an annotated MP4 with track IDs and detector confidence, per-class directional totals, timestamped CSV/JSON events, reusable configuration, source code, setup documentation, and a measured performance report.
Before implementation, I review a representative clip and define exactly what creates one count. Outputs are checked for readable video, processed/output frame totals, reproducible settings, and incomplete-run errors. The public demo processed 647 frames at 78.2 end-to-end FPS on the test CPU, but performance is measured again for your footage and target hardware.
Formal accuracy is reported only against human-verified ground truth. Client footage remains private and is never published without written permission.
Machine Learning Tools
NumPy, OpenCV, PythonWhat's included
| Service Tiers |
Starter
$35
|
Standard
$85
|
Advanced
$180
|
|---|---|---|---|
| Delivery Time | 3 days | 5 days | 7 days |
Number of Revisions | 1 | 2 | 2 |
Number of Model Variations | 1 | 1 | 1 |
Number of Scenarios | 1 | 1 | 1 |
Number of Graphs/Charts | 0 | 0 | 0 |
Model Validation/Testing | |||
Model Documentation | - | ||
Data Source Connectivity | - | - | - |
Source Code | - |
Optional add-ons
You can add these on the next page.
Additional Revision
+$15Frequently asked questions
About Lin
Computer Vision Engineer | OpenCV, Image Processing & Deep Learning
Xiamen, China - 9:07 am local time
I can help with:
Building image classification pipelines for defect detection
Developing image preprocessing and data preparation workflows
Creating computer vision prototypes with OpenCV and deep learning models
Training and evaluating binary image-classification models
Developing semantic segmentation solutions with U-Net
Preparing datasets and reporting model performance with accuracy, precision, and recall
In my current Computer Vision Software Engineer internship, I contributed to a pharmaceutical bottle defect-detection prototype. The project used OpenCV and a MobileNet-based approach to distinguish normal and defective bottle images. I prepared, preprocessed, trained, and evaluated a dataset of approximately 2,000 industrial images.
I approach projects by first clarifying the image source, target defect or class, expected output, and evaluation criteria. I then build a practical prototype that can be tested with real images and documented for further development.
If you need help with an image-inspection task, dataset preparation, image classification, or a computer vision prototype, feel free to share a sample image and the outcome you need.
Steps for completing your project
After purchasing the project, send requirements so Lin can start the project.
Delivery time starts when Lin receives requirements from you.
Lin works on your project following the steps below.
Revisions may occur after the delivery date.
Configure the counting pipeline
I freeze the agreed classes, line position, directions, thresholds, output format, and target environment from the supplied requirements.
Process and inspect the video
I run detection, tracking, and line-crossing logic and generate the annotated video and event data included in your tier.