SiftDescriptorExtractor 导致内存泄漏

发布于 2024-11-27 19:09:12 字数 801 浏览 1 评论 0原文

我目前正在实施 SIFT 从图像中提取特征点，并注意到在获取描述符时出现内存泄漏。无论如何，我可以释放课堂上可能附加的内存吗？

编辑向代码块添加了更多详细信息

cv::SiftFeatureDetector* features = new cv::SiftFeatureDetector();
cv::SiftDescriptorExtractor* extractor = new cv::SiftDescriptorExtractor();

std::vector<cv::KeyPoint> KeyPoints;
cv::Mat Descriptors;

// Turn the image into a Mat
cv::Mat mImage = cv::Mat(iplImage);

printf("Searching for keypoints in: %s.\n", szName.c_str());

// Detect keypoints
features->detect(mImage, KeyPoints);

printf("Found %d keypoints.\n", KeyPoints.size());

// Extract descriptors
extractor->compute(mImage, KeyPoints, Descriptors);

printf("Found %d descriptors.\n\n", Descriptors.rows);

// Let my memory go!
delete extractor;
delete features;

非常感谢任何建议。谢谢。

原文

I am currently implementing SIFT to extract feature points from an image and noticed that I have a memory leak when I get the descriptors. Is there anyway I can free the memory that may be attached in the class?

EDIT
Added more details to the code block

cv::SiftFeatureDetector* features = new cv::SiftFeatureDetector();
cv::SiftDescriptorExtractor* extractor = new cv::SiftDescriptorExtractor();

std::vector<cv::KeyPoint> KeyPoints;
cv::Mat Descriptors;

// Turn the image into a Mat
cv::Mat mImage = cv::Mat(iplImage);

printf("Searching for keypoints in: %s.\n", szName.c_str());

// Detect keypoints
features->detect(mImage, KeyPoints);

printf("Found %d keypoints.\n", KeyPoints.size());

// Extract descriptors
extractor->compute(mImage, KeyPoints, Descriptors);

printf("Found %d descriptors.\n\n", Descriptors.rows);

// Let my memory go!
delete extractor;
delete features;

Any advice is greatly appreciated. Thanks.

分享到QQ

分享到微博

如果你对这篇内容有疑问，欢迎到本站社区发帖提问参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。

发布评论

需要登录才能够评论，你可以免费注册一个本站的账号。

无名指的心愿 2024-12-04 19:09:12

你是对的。我刚刚在 Linux 上使用 OpenCV 2.3 进行了测试， compute() 确实存在内存泄漏。这会影响 SiftDescriptorExtractor，也可能影响其他类型，例如 SurfDescriptorExtractor、OrbDescriptorExtractor 和 BriefDescriptorExtractor。

顺便说一句，不要忘记cvReleaseImage() 在此代码末尾调用iplImage 的图像。