ES-Alexander / ES-Alexander/pythonic-cv

Investigate performance effects of current structure

未关闭
#18 1 条评论 0 个 reaction 已指派 0 人 在 GitHub 查看
question
主要语言
Python
星标
45
派生
3
PR 合并指标
30 天内没有已合并 PR

描述

## Motivation:
Every convenience comes at a cost, but quantifying the cost makes it possible to determine whether it's "too expensive".

## Suggested Method
Currently use a combination of inheritance and composition, which make modification and new base classes really simple to implement and use, but may come at a performance cost. Would be good to quantify the difference between the current situation and e.g. the 'best-case' performance of `VideoReader`, `LockedCamera`, and `LockedScreen` being implemented with `cv2.VideoCapture` as their direct parent class.

## Possible Resulting Changes
If performance difference is significant, it may be worth
- making a code-generator that can trade off some program memory for the benefit of fewer function calls at each step
- changing code structure (hard to recommend - very versatile at the moment, and repetition is a 'bad smell')
- recommending this library for rapid development, and more optimised code once development is complete (e.g. lower level looping, or just C++ implementation)

贡献指南

这个仓库没有索引到贡献指南

评估

这个 Issue 还没有评估数据。

把新 issue 发到你的邮箱

精选适合新手参与的 GitHub issue 摘要。