appcell / OverwatchDataAnalysis

A video analyzer for getting player/chara(name, health, ult etc)/killfeed info in an Overwatch game video replay.

Overwatch Replay Analyzer (ORA) v0.1.1

Introduction

Overwatch Replay Analyzer (ORA) v0.1.1 is a free software developed by OWDATA.ORG, a free organization devoted to competitive Overwatch data analysis. ORA can extract a timeline of events from a third-person perspective replay video. Currenly our effort is limited to non-Asian custom games(*) and OWL videos only. In v0.1.1, a timeline includes:

1. Elimination(Killer, killed player, ability, assist players, critical shot)
2. Resurrection (Resurrector, resurrected player)
3. Ult charged/ult used
4. Ult charge percentage
5. Chara switching

The released binaries & source code of ORA are published under GPL v3 license. It's free for all to redistribute and modify under the terms of the GPL v3.

Feel free to join us on Discord: https://discord.gg/hnvc3yG

(*): Since Asian Overwatch game UI uses a different font from non-Asian ones, issues might rise during recognition. Hopefully we will fix it in the next release.

Change log

• Added support for custom games replay videos
• Added support for OWL stage 3-4 replay videos
• Updated elimination ability & assist recognition
• Added ult charge percentage recognition
• Added support for Brigitte, Moira and new Hanzo
• Added automatic update
• Added JSON output for further analysis and sharing
• Bugs fixed

Environment requirements

OS: Windows 7, 8 & 10, MacOS & Linux

Usage

1. Download & extract to a directory, then double click on main.exe to run
2. Choose the video file to be analysed in Video path
3. Choose the output path in Save to
4. Input team names: team on left side in Team A, team on right side in Team B
5. Input player names (1-6, 7-12 from left to right)
6. Input start & end time, game type, analysis FPS and stage number (for OWL replays only)
7. Click on Analyze and wait

Video file requirements

The video has to be full-screened without watermark during the match.

Resolution should be no less than 720p.

All events in the same video will be outputed to the same timeline. Thus it's strongly recommended not to include more than 1 matches in one run.

Bug report

If the program doesn't run, or the output isn't what you expect, please rise a new issue in: https://github.com/appcell/OverwatchDataAnalysis/issues

An issue usually includes:

1. problem description
2. name & version of your operating system
3. specify the match or the video file you analyze, and when the problem happens
4. the outputed .xlsx file and '.zip' file (if exists)
5. contact info (if you don't check your email that often)

Join us

You can join us on Discord: https://discord.gg/hnvc3yG

Or for contacting, send an email to: [e-mail]

Donation

ORA is a free project, but we would still appreciate donation to maintain our server. You can donate by,

Paypal: https://www.paypal.me/yzhuappcell

Or,

All donations will be used on server maintainance and further development. Thank you for your help & support!

---

守望先锋录像分析器 v0.1.1

Overwatch Replay Analyzer (ORA) v0.1.1

软件简介

守望先锋录像分析器v0.1.1（以下简称 ORA）是由OWDATA.ORG开发的自由软件。该软件可以根据守望先锋的比赛 ob 视角录像 （目前仅限 非亚洲语言的自定义比赛视频与OWL 比赛视频） 提取事件时间轴，并将时间轴输出到 Excel 表格。当前版本（v0.1.1）中，时间轴包括以下内容：

1. 击杀事件（击杀者，被击杀者，击杀所用技能，助攻，是否暴击）
2. 复活事件（复活者，被复活者）
3. 大招充能完毕及大招使用事件
4. 大招能量
5. 当前玩家所用英雄，及英雄切换事件

ORA的发行版及其源码遵从 GPL v3 许可，免费供所有爱好者获取和使用。商业性使用请参阅源码中附带的 GPL v3 许可协议以获取法律信息。

欢迎加入我们的Discord群组： https://discord.gg/hnvc3yG

更新日志

• 加入对自定义比赛视频的支持
• 加入对OWL第3-4阶段比赛视频的支持
• 改进了技能和助攻的识别
• 增加了大招能量的识别和记录
• 增加了对布里吉塔，莫伊拉和新版半藏的支持
• 增加了程序自动更新
• 增加了JSON格式输出以便后续分析和数据共享
• 修复了一些已知bug

运行环境

操作系统： Windows 7，8，10， MacOS， Linux

使用方法

1. 双击运行程序后，在Video path中选择需要分析的视频文件
2. 在 Save to 中选择分析结果的保存路径
3. 在“A 队”文本框中输入左边一队的队名，“B 队”输入右边一队的队名
4. 在红色玩家名输入框内按照从左至右的顺序输入左边一队的玩家名
5. 在蓝色玩家名输入框内按照从左至右的顺序输入右边一队的玩家名
6. 输入视频中比赛的开始/结束时间（以秒为单位），按提示输入视频类型。OWL视频需输入阶段数。
7. 点击 “Analyze” 按钮，等待分析结果生成

视频文件要求：

录像需为全屏且左右上角无水印，清晰度不小于720p，长宽比为16：9。

一个视频只能包含一张地图的比赛，暂不支持攻防互换。

Bug 反馈

如遇到程序无法运行，或分析结果与预期不符，请在 https://github.com/appcell/OverwatchDataAnalysis/issues 提交新的issue。

Issue应包括：

1. 问题描述
2. 所用操作系统
3. 所用视频文件（或OWL比赛场次信息）及错误发生所在视频时间
4. （如有）生成的.xlsx和.zip分析结果
5. 联系信息

联系我们

您可以加入我们的Discord群组: https://discord.gg/hnvc3yG

或发邮件至: [e-mail]

赞助我们

ORA is a free project, but we would still appreciate donation to maintain our server. You can donate by, 尽管ORA是一个免费项目，我们仍然需要捐款来维持服务器的运行。您可以通过以下方式对ORA进行赞助：

Paypal: https://www.paypal.me/yzhuappcell

或者：

所有捐款将被用于服务器维护以及未来研发。感谢您的使用和测试！