gemini-count-in-video

1# Gemini Video Understanding Skill
2 
3## Purpose
4This skill enables video analysis and object counting using the Google Gemini API, with a focus on **counting pedestrians**, **detecting objects**, **tracking movement**, and **analyzing surveillance footage**. It supports precise prompting for differentiated counting (e.g., pedestrians vs cyclists vs vehicles).
5 
6## When to Use
7- Counting pedestrians, vehicles, or other objects in surveillance videos
8- Distinguishing between different types of objects (walkers vs cyclists, cars vs trucks)
9- Analyzing traffic patterns and movement through a scene
10- Processing multiple videos for batch object counting
11- Extracting structured count data from video footage
12 
13## Required Libraries
14 
15The following Python libraries are required:
16 
17```python
18from google import genai
19from google.genai import types
20import os
21import time
22```
23 
24## Input Requirements
25- **File formats**: MP4, MPEG, MOV, AVI, FLV, MPG, WebM, WMV, 3GPP
26- **Size constraints**:
27  - Use inline bytes for small files (rule of thumb: <20MB).
28  - Use the File API upload flow for larger videos (most surveillance footage).
29  - Always wait for processing to complete before analysis.
30- **Video quality**: Higher resolution provides better counting accuracy for distant objects
31- **Duration**: Longer videos may require longer processing times; consider the full video length for accurate counting
32 
33## Output Schema
34For object counting tasks, structure results as JSON:
35 
36```json
37{
38  "success": true,
39…