z轴飞了

[JiangWeiHn]

请问我在采集的较为平整的公园的数据上，建图效果很好，测试的每个点的位置和rtk的坐标对比，经纬度误差都在10cm内。用同样的参数，相同小车采集的公司园区数据，有斜坡，也有一些较窄的长道。跑的结果如下图所示


z轴明显漂移，有什么建议么？谢谢。

[woshinidiesb]

FAST-LIO-Multi-Sensor-Fusion 使用效果和这个对比有什么差异

[JiangWeiHn]

gnss有抑制效果，但横向也有漂移。现在正尝试把他们结合起来。

[woshinidiesb]

First at all,thank you very much for your reply.l found a few method about how to  combine gnss to fast_lio at last week.lt's named that FAST_LIO_SAM(https://github.com/kahowang/FAST_LIO_SAM) and FAST-LIO-Multi-Sensor-Fusion(https://github.com/zhh2005757/FAST-LIO-Multi-Sensor-Fusion) .the fast_lio_samcombine gnss to fast_lio in gtsam .however, the FAST-LIO-Multi-Sensor-Fusion that even use the odometry to constraint pose combine gnss to fast_lio in ieskf , but the both code is in ros1,however,l can't transform it to ros2 and use it.May be the both code can give you some idea ★往事“曾经 ***@***.***  

…

------------------ 原始邮件 ------------------ 发件人: "hku-mars/LTAOM" ***@***.***>; 发送时间: 2024年8月26日(星期一) 下午3:04 ***@***.***>; ***@***.******@***.***>; 主题: Re: [hku-mars/LTAOM] z轴飞了 (Issue #25) gnss有抑制效果，但横向也有漂移。现在正尝试把他们结合起来。 — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you commented.Message ID: ***@***.***>