We study the faint end of the HI mass function (HIMF) in order to test the predictions of the CDM theory on the number density of objects with small (dark) masses. The neutral hydrogen is much better tracer of the underlying mass distribution compared to the luminous matter and can be used to test the existence of a population of small galaxies in which the star formation has been partially or completely suppressed during cosmic evolution. Due to technical limitations, the existing HI surveys are not very sensitive on masses below 10$^8$ HI $M_{\odot}$. We designed a blind HI survey to be sensitive for objects with small HI masses. The surveyed area is in the Canis Venatici groups of galaxies and covers in total $\approx$86 deg$^2$ of sky, with the observed velocities in the range $-350 < cz < 1400$ km s$^{-1}$. We detected 69 objects, 22 of them for the first time in HI. All new HI detections fall in the lower part of the mass-histogram, confirming our ability to detect galaxies with small HI masses. The calculated HIMF is flat in the faint end regime ($\alpha {\sim}{-}$1), different from the steep rise predicted by CDM models. Possible effects of the environment on the estimated HIMF parameters are discussed.