Merger problem of inv Variation

[SunWinner01]

I would like to ask, I am currently using several software to detect SV. I chose to use svimmer software to first merge the vcf result files of different software for each individual, and then merge the result files of all samples into a population file for subsequent analysis. . But I found that the final result showed very few INVs. Was it lost during the merge? Regarding the merging of Variation like inv, does it need to be merged separately? Or has svimmer already excluded related issues.